EVALUATION OF PASSIVE UHF RADIO FREQUENCY IDENTIFICATION TRANSPONDER PERFORMANCE USING DIFFERE NT PACKAGING MATERIALS By Yuanchenxi Zhang A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Packaging Ð Master of Science 2018 ABSTRACT EVALUATION OF PASSIVE UHF RADIO FREQUENCY IDENTIFICATION TRANSPONDER PERFORMANCE USING DIFFERENT PACKAGING MATERIALS By Yuanchenxi Zhang Radio Frequency Identification (RFID) technology plays an important role in supply chains by providing possibilities of improved security, efficiency and visibility of item tracking and management. In order to achieve the expected function ality of RFID technology, it is critical to understand factors that influence RFID transponder performance. The objective of this research was to determine the effect of packaging materials on the performance of passive UHF RFID transponders in a simulated manufacturing environment. Three general -purpose passive transponders were tested when they were attached to four packaging materials. Performance pa rameters of read range, which is the maximum distance the transponder can be detected by the interrogator, and orientation read rate, which is the percentage of orientations in which the transponder was read, out of the randomly chosen set of orientations evaluated based on rotation, tilt, or incline of the transponder within three -dimensional space, were quan tified and analyzed. The results showed that packaging materials had a consistent effect on read range and orientation read rate across different antenna designs of passive dipole antenna transponders. Transponder antenna designs had a significant effect o n read range and orientation read rate. Interrogator antenna polarizations had a significant effect on transponder read range but did not have a consistent effect on transponder orientation read rate. !!iii ACKNOWLEDGEMENTS I would like to thank my committee: Dr. Selke, Dr. Bix of the School of Packaging, and Dr. Steve Melnyk of the Eli Broad College of Business . I would like to thank Dr. Clarke for getting me involved in the field of RFID and generously giving his knowledge and patience. I would also like to thank Dr. Burgess for designing and building the wooden transponder stand used in my study. I would like to thank the CANR Statistical Consulting Center, especially, Jessica Fry and Moslem Ladoni. Thank you for all of the time you dedicated through the statistical analysis process of my research. Thanks to Impinj Inc. and Mr. Gregory Robinson for generously providing me the transponders in my research. I would like to thank all my friends in the School of Packa ging, especially, Rui Chen, Tony Trier, Fatin Sabbagha, Paula Perez , Alyssa Harben, and Eric Estrada . Thank you for all of your encouragement and friendship. Thanks to all of the professors, staff members in the School of Packaging for supporting me when I needed it. And finally, I would like to thank my family and friends outside of school for your support and encouragement. A special thank s to Mom and Dad, for giving me life and loving me endlessly. !iv TABLE OF CONTENTS LIST OF TABLES ...................................................................................................................... v LIST OF FIGURES ................................................................................................................... ix KEY TO ABBREVIATIONS ................................................................................................... xvi CHAPTER 1 INTRODUCTION ................................................................................................. 1 1.1 Overview .......................................................................................................................... 1 1.2 Research Goals .................................................................................................................. 3 1.3 Thesis Summary ................................................................................................................ 4 CHAPTER 2 LITERATURE REVIEW ....................................................................................... 5 2.1 Automatic Identification and Data Capture Technologies ................................................... 5 2.2 History of RFID ................................................................................................................ 5 2.3 Components of RFID ........................................................................................................ 6 2.3.1 Transponder ................................................................................................................ 6 2.3.2 Interrogator ................................................................................................................. 7 2.3.3 Host Computer and Middleware ................................................................................. 9 2.4 RFID Standards, Regulations and Protocols ...................................................................... 9 2.4.1 RFID Standard Bodies ................................................................................................ 9 2.4.2 Frequency Allocation .................................................................................................. 9 2.4.3 RFID Protocols ......................................................................................................... 11 2.4.4 The Electronic Product Code .................................................................................... 12 2.5 Applications of RFID ...................................................................................................... 12 2.5.1 RFID in the Transport and Logistics Industry ........................................................... 13 2.5.2 RFID -based Electronic Article Surveil lance (EAS) System in the Retail Store .......... 14 2.5.3 Successful Case of Benefiting from RFID in Supply Chain ...................................... 15 2.6 Principles of RFID .......................................................................................................... 16 2.6.1 Near Field and Far Field ........................................................................................... 16 2.6.2 Forward Link Communication .................................................................................. 17 2.6.3 Reverse Link Communication ................................................................................... 18 2.7 Transponder Performance in the Real World .................................................................... 20 2.7.1 Product and P ackaging Materials .............................................................................. 21 2.7.2 Research on the Effect of Packaging and Product on Transponder Performance ........ 24 !v 2.7.3 Operati ng Environment Effect on Transponder Performance ..................................... 28 2.8 Research on Testing Methodology ................................................................................... 30 2.9 Relevancy ....................................................................................................................... 36 CHAPTER 3 METHODOLOGY .............................................................................................. 37 3.1 Test Environment ............................................................................................................ 37 3.2 RFID Equipment ............................................................................................................. 39 3.3 Packaging Materials ........................................................................................................ 42 3.4 Procedures ....................................................................................................................... 46 3.4.1 Read Range Test ....................................................................................................... 46 3.4.2 Orientation Read Rate ............................................................................................... 48 3.5 Data Management ........................................................................................................... 49 CHAPTER 4 RESULTS ............................................................................................................ 51 4.1 Results of the Read Range Test ........................................................................................ 51 4.2 Results of the Orientation Read Rate Test ........................................................................ 76 CHAPTER 5 ANALYSIS AND DISCUSS ION ......................................................................... 85 5.1 Overview ........................................................................................................................ 85 5.2 Analysis of the Read Range Test ...................................................................................... 86 5.3 Analysis of the Orientation Read Rate Test ...................................................................... 89 CHAPTER 6 CONCLUSION S AND FUTURE RESEARCH ................................................... 96 APPENDICES .......................................................................................................................... 99 Appendix 1 Pilot Test 1 ....................................................................................................... 100 Appendix 2 Pilot Test 2 ....................................................................................................... 110 Appendix 3 Figures of TotCnt and RSSImx of Each Read Point in the Rea d Range Test ..... 119 Appendix 4 Figures of TotCnt and RSSImx or Each Read Orientation in the O -RDRate Test ........................................................................................................................................... 129 REFERENCES ....................................................................................................................... 139 !vi LIST OF TABLES Table 1 UHF Regulations for Passive RFID Within the 860 to 960 MHz Band in Different Countries and Regions ....................................................................................................... 11 Table 2 Dielectric Properties of Materials Tested at Certain Frequencies and Temperatures ....... 24 Table 3 Transponder Descriptions by the Manufacturer ............................................................. 41 Table 4 60 Sets of Random Angle Combination ........................................................................ 49 Table 5 Summary of the Read Range Test of Tag 1 with Four Materials Using LP Antenna ....... 52 Table 6 Summary of the Read Range Test of Tag 1 with Four Materials Using CP Antenna ....... 52 Table 7 Summary of the Read Range Test of Tag 2 with Four Materials Using LP Antenna ....... 53 Table 8 Summary of the Read Range Test of Tag 2 with Four Materials Using CP Antenna ....... 54 Table 9 Summary of the Read Range Test of Tag 3 with Four Materials Using LP Antenna ....... 54 Table 10 Summary of the Read Range Test of Tag 3 with Four Materials Using CP Antenna ..... 55 Table 11 Summary of the Orientation Read Rate Test of Tag 1 Using LP Antenna ..................... 77 Table 12 Summary of the Orientation Read Rate Test of Tag 1 Using CP Antenna .................... 77 Table 13 Summary of the Orientation Read Rate Test of Tag 2 Using LP Antenna ..................... 77 Table 14 Summary of the Orientation Read Rate Test of Tag 2 Using CP Antenna .................... 78 Table 15 Summary of the Orientation Read Rate Test of Tag 3 Using LP Antenna ..................... 78 !vii Table 16 Summary of the Orientation Read Rate Test of Tag 3 Using CP Antenna .................... 79 Table 17 Test of Equality over Material (M1, M2 and M3) ........................................................ 87 Table 18 Test of Equality over Antenna ..................................................................................... 87 Table 19 Test of Equality over Tag ............................................................................................ 87 Table 20 Result of the Cox Proportional Hazard Regression of the Survival Analysis ................ 88 Table 21 Fit Statistics of the Model with Equal Variance ........................................................... 91 Table 22 Fit Statistics of the Model with Unequal Variance ....................................................... 91 Table 23 Effect of Antenna, Tag and Material on the TotCnt in the Orientation Read Rate Test . 91 Table 24 Least Squares Means of TotCnt Tested with the LP and CP Antennas .......................... 92 Table 25 Least Squares Means of TotCnt with Different Packaging Materials ........................... 92 Table 26 Least Squares Means of TotCnt with Different Tags .................................................... 93 Table 27 Least Squares Means of TotCnt with Antenna by Tag .................................................. 94 Table 28 Test of Equality over Procedures for Critical Distance and Acquisition Distance ....... 104 Table 29 Test of Equality over Angle ! .................................................................................... 105 Table 30 CP Antenna, TotCnt as a Function of ! , " and r ......................................................... 107 Table 31 CP Antenna, Response Surface for TotCnt ................................................................ 107 Table 32 RSSImx as a Function of ! , " and r ........................................................................... 108 !viii Table 33 CP Antenna, Response Surface for RSSImx .............................................................. 108 Table 34 LP Antenna TotCnt as a Function of ! , " and r .......................................................... 108 Table 35 LP Antenna, Response Surface for TotCnt ................................................................. 109 Table 36 LP Antenna, RSSImx as a Function of ! , " and r ....................................................... 109 Table 37 LP Antenna, Response Surface for RSSImx .............................................................. 109 !ix LIST OF FIGURES Figure 1 Side View of the Test Set Physical Block Diagram for Tests in the EPCglobal Tag Performance Parameters and Test Methods ........................................................................ 31 Figure 2 The Machinery Lab in the School of Packaging at Michigan State University ............. 37 Figure 3 Layout of the Machinery Lab and the Test Location in the Room ................................ 38 Figure 4 Vaisala HMI41 Humidity and Temperature Indicator ................................................... 39 Figure 5 Impinj IPJ -R1000 Reader ............................................................................................ 39 Figure 6 Circularly Polarized Antenna on the PVC Pipe. ........................................................... 40 Figu re 7 Transponders used in this t hesis .................................................................................. 42 Figure 8 Packaging Material Samples ....................................................................................... 42 Figure 9 Transponder Support, Front View ................................................................................ 43 Figure 10 Transponder Support, Front View .............................................................................. 44 Figure 11 Parameter Definitions ................................................................................................ 45 Figure 12 Read Range of Tag 1 with M1 Measured Using LP Antenna ...................................... 59 Figure 13 Read Range of Tag 1 wi th M2 Measured Using LP Antenna ...................................... 59 Figure 14 Read Range of Tag 1 with M3 Measured Using LP Antenna ...................................... 60 Figure 15 Read Range of Tag 1 with M1 Measured Using CP Antenna ..................................... 60 !x Figure 16 Read Range of Tag 1 with M2 Measured Using CP Antenna ..................................... 61 Figure 17 Read Range of Tag 1 with M3 Measured Using CP Antenna ..................................... 61 Figure 18 Read Range of Tag 2 with M1 Measured Using LP Antenna ...................................... 62 Figure 19 Read Range of Tag 2 with M2 Measured Using LP Antenna ...................................... 62 Figure 20 Read Range of Tag 2 with M3 Measured Using LP Antenna ...................................... 63 Figure 21 Read Range of Tag 2 with M1 Measured Using CP Antenna ..................................... 63 Figure 22 Read Range of Tag 2 with M2 Measured Using CP Antenna ..................................... 64 Figure 23 Read Range of Tag 2 with M3 Measured Using CP Antenna ..................................... 64 Figure 24 Read Range of Tag 3 with M1 Measured Using LP Antenna ...................................... 65 Figure 25 Read Range of Tag 3 with M2 Measured Using LP Antenna ...................................... 65 Figure 26 Read Range of Tag 3 with M3 Measured Using LP Antenna ...................................... 66 Figure 27 Read Range of Tag 3 with M1 Measured Using CP Antenna ..................................... 66 Figure 28 Read Range of Tag 3 with M2 Measured Using CP Antenna ..................................... 67 Figure 29 Read Range of Tag 3 with M3 Measured Using CP Antenna ..................................... 67 Figure 30 Tag 1 with M1, M2, and M3, TotCnt versus Distance when " =0¡ .............................. 68 Figure 31 Tag 1 with M1, M2, and M3, RSSImx versus Distance when " =0¡ ............................ 68 Figure 32 Tag 2 with M1, M2, and M3, TotCnt versus Distance when " =0¡ .............................. 69 !xi Figure 33 Tag 2 with M1, M2, and M3, RSSImx versus Distance when " =0¡ ............................ 69 Figure 34 Tag 3 with M1, M2, and M3, TotCnt versus Distance when " =0¡ .............................. 70 Figure 35 Tag 3 with M1, M2, and M3, RSSImx versus Distance when " =0¡ ............................ 70 Figure 36 Each Tag with M1, TotCnt versus Distance when " =0¡ .............................................. 71 Figure 37 Each Tag with M1, RSSImx versus Distance when " =0¡ ........................................... 71 Figure 38 Each Tag with M2, TotCnt versus Distance when " =0¡ .............................................. 72 Figure 39 Each Tag with M2, RSSImx versus Distance when " =0¡ ........................................... 72 Figure 40 Each Tag with M3, TotCnt versus Distance when " =0¡ .............................................. 73 Figure 41 Each Tag with M3, RSSImx versus Distance when " =0¡ ........................................... 73 Figure 42 LP Antenna, Each Tag with M1, M2, and M3, TotCnt versus Distance when " =0¡ .... 74 Figure 43 LP Antenna, Each Tag with M1, M2, and M3, RSSImx versus Distance when " =0¡ .. 74 Figure 44 CP Antenna, Each Tag with M1, M2, and M3, TotCnt versus Distance when " =0¡ .... 75 Figure 45 CP Antenna, Each Tag with M1, M2, and M3, TotCnt versus Distance when " =0¡ .... 75 Figure 46 Orientation Read Rate of Tag 1 with Four Materials Tested Using LP Antenna .......... 82 Figure 47 Orientation Read Rate of Tag 1 with Four Materials Tested Using CP Antenna .......... 82 Figure 48 Orientation Read Rate of Tag 2 with Four Materials Tested Using LP Antenna .......... 83 Figure 49 Orientation Read Rate of Tag 2 with Four Materials Tested Using CP Antenna .......... 83 !xii Figure 50 Orientation Read Rate of Tag 3 with Four Materials Tested Using LP Antenna .......... 84 Figure 51 Orientation Read Rate of Tag 3 with Four Materials Tested Using CP Antenna .......... 84 Figure 52 Residuals for TotCnt in the Orientation Read Rate Test ............................................. 90 Figure 53 Tag AD -338m5 and the High -Density Fiberboard .................................................... 100 Figure 54 Comparison of the R !-"-cr and R !-"-ac Measured by the LP Antenna .................... 103 Figure 55 Comparison of the R !-"-cr and R !-"-ac Measured by the CP Antenna .................... 104 Figure 56 LP Antenna, Tag AD -381m5 with HDF, ! =90¡ and " =0¡, TotCnt versus # ............... 112 Figure 57 LP Antenna, Tag AD -381m5 with HDF, ! =120¡ and " =315¡, TotCnt versus # ......... 112 Figure 58 LP Antenna, Tag AD -381m5 with HDF, ! =150¡ and " =345¡, TotCnt versus # ......... 113 Figure 59 CP Antenna, Tag AD -381m5 with HDF, ! =90¡ and " =0¡, TotCnt versus # ............... 113 Figure 60 CP Antenna, Tag AD -381m5 with HDF, ! =120¡ and " =315¡, TotCnt versus # ......... 114 Figure 61 CP Antenna, Tag AD -381m5 with HDF, ! =150¡ and " =345¡, TotCnt versus # ......... 114 Figure 62 the Medium Density Fiberboard .............................................................................. 115 Figure 63 the Low Density Fiberboard .................................................................................... 116 Figure 64 LP Antenna, Rmax of Tags with Different Materials while ! =90¡, " =0¡ and # =0¡ ... 118 Figure 6 5 CP Antenna, Rmax of Tags with Different Materials while ! =90¡, " =0¡ and # =0¡ ... 118 Figure 66 LP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Point ..................... 119 !xiii Figure 67 LP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Point ..................... 120 Figure 68 LP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Point ..................... 120 Figure 69 CP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Point ..................... 121 Figure 70 CP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Point ..................... 121 Figure 71 CP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Point ..................... 122 Figure 72 LP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Point ..................... 122 Figure 73 LP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Point ..................... 123 Figure 74 LP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Point ..................... 123 Figure 75 CP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Point ..................... 124 Figure 76 CP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Point ..................... 124 Figure 77 CP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Point ..................... 125 Figure 78 LP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Point ..................... 125 Figure 79 LP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Point ..................... 126 Figure 80 LP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Point ..................... 126 Figure 81 CP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Point ..................... 127 Figure 82 CP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Point ..................... 127 Figure 83 CP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Point ..................... 128 !xiv Figure 84 LP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Orientation ............ 129 Figure 85 LP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Orientation ............ 130 Figure 86 LP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Orientation ............ 130 Figure 87 CP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Orientation ........... 131 Figure 88 CP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Orientation ........... 131 Figure 89 CP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Orientation ........... 132 Figure 90 LP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Orientation ............ 132 Figure 91 LP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Orientation ............ 133 Figure 92 LP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Orientation ............ 133 Figure 93 CP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Orientation ........... 134 Figure 94 CP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Orientation ........... 134 Figure 95 CP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Orientation ........... 135 Figure 96 LP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Orientation ............ 135 Figure 97 LP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Orientation ............ 136 Figure 98 LP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Orientation ............ 136 Figure 99 CP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Orientation ........... 137 Figure 100 CP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Orientation ......... 137 !xv Figure 101 CP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Orientation ......... 138 !xvi KEY TO ABBREVIATIONS !AIC Akaike Information C riterion AIDC Automatic Identification and Data Capture AIRTC Auto -Identification Research and Testing Center AM-EAS Acousto -magnetic Electronic Article Surveillance ASTM American Society for Testing and Materials CANR College of Agriculture & Natural Resources CP Circularly Polarized CW Continuous Wave dBi decibel isotropic dBm decibel milliwatts DC Distribution Center DoD Department of Defense DSB -ASK Double -sideband Amplitude Shift Keying EIRP Effective Isotropic Radiated Power EM Electromagnetic EPC Electronic Product Code ERP Effective Radiated Power !xvii ETSI European Telecommunications Standards Institute FCC Federal Communications Commission HF High Frequency IC Integrated Circuit IEC International Electrotechnical Commission ISO International Organization for Standardization IT Information Technology LF Low Frequency LP Linearly Polarized M1 C-flute uncoated corrugataed paperboard M2 Polyethylene film M3 Polyethylene corrugated board M4 PET/Al/polyethylene film MF Medium Frequency MSU Michigan State University NPV Net Present Value OCR Optical Character Recognition O-RDRate Orientation Read Rate PR-ASK Pulse -interval Amplitude Shift Keying !xviii RCS Radar Cross Section RF Radio Frequency RF-EAS Radio Frequency Electronic Article Surveillance RFID Radio Frequency Identification RH Relative Humidity RSSImx Maximum Response Signal Strength Indicator SAP Superabsorbent Polymer SAS Statistical Analysis System SCC Statistical Consulting Center SHF Super High Frequency SSB-ASK Single -sideband Amplitude Shifting Keying TIPP Tagged -Item Performance Protocol TotCnt Total Count of Reads UHF Ultra High Frequency RCS Delta Radar Cross Section Three Dimensional !1 Chapter 1 Introduction 1.1 Overview This research topic began with an evaluation of a passive ultra -high frequency (UHF) radio frequency identification (RFID) transponderÕs three -dimensional read range in a simulated manufacturing environment within the Michigan State University Auto -Identification Research and Testing Center (MSU AIRTC). The MSU AIRTC, dire cted by Dr. Robb Clarke, focused on research topics related to applications of automatic ide ntification and data capture (AIDC) technology in process control, logistics and transportation, and inventory management, as well as the effects of packaging and products on the performance of AIDC technology [1] Ð[8] . Results from testing the passive UHF RFID transponderÕs three -dimensional read range led to an interest in evaluating passive UHF RFID transpondersÕ performance while the transponder is attached to different packagin g materials. RFID, as one type of AIDC technology, is a means to identify an object using radio frequency (RF) transmission. RFID consists of an interrogator, transponder and host computer installed with middleware. The interrogator receives a request for information from the host c omputer through the middleware, and sends the request to the transponder within its interrogation zone via RF signals. The transponder processes the reque st and sends a response containing the requested information to the interr ogator. The interrogator then sends the collected information to the host computer [9] . RFID technology acquires an itemÕs information from a relatively far distance !2 compared to the short distance possible with the commonly known barcode technology. With RFID , the communication can be achieved even without a line -of-sight to the object. RFID technology enables visibility, efficiency and security of automatic data collection processes within supply chain s [9] . There are many factors that influence the performance of a UHF RFID system, including the sensitivity of the interrogator and the transponder, th e transponderÕs orientation, interference from the operation al environment, a productÕs packaging materials, a productÕs material and the operation processes [2], [3] . Many researchers have developed different methodologies to evaluate the performance of RFID transponders from varying approaches: some methodologies requ ired anechoic chambers to conduct tests (e.g. [10], [11] ); some focused on transponder performance test s in real world environments (e.g. [12], [13] ); and some explored the effect of packaging materials and product contents (e.g. [1], [2] ), environment al noise [3] , and practical conditions of operati on processes (e.g. [7], [8] ) on the performance of a RFID transponder. Although the effects of packaging materials on transponder performance have been somewhat documented [2], [8], [14], [15] , there are still many questions in this area that are worth exploring. This thesis focuses on the evaluation of RFID transponder performance with different packaging materials in a simulated manufacturing environm ent. The transponder performance was quantified and analyzed by means of read range and orientation read rate. In this research, the read range was interpreted as the maximum distance from which a transponder can be read by the interrogator . A new term nam ed orientation read rate was introduced and defined as the percentage !3 of orientations in which the transponder was read, out of the randomly chosen set of orientations evaluated based on rotation , tilt, or incline of the transponder within three -dimensiona l space. This research provides an unbiased testing methodology to measure RFID transpondersÕ read range and orientation read rate. It thus contributes insights into the packaging materialsÕ effects on RFID transponder performance. 1.2 Research Goals The p rimarily goal of this research was to investigate the effect of packaging materials on the performance of passive UHF RFID transponders in a simulated manufacturing environment. To fulfill this goal, three general -purpose passive UHF RFID transponders, cou rtesy of Impinj, were evaluated when attached to four commonly used packaging materials, including uncoated corrugated paperboard, polyethylene film, polyethylene corrugated board and PET /aluminum/polyethylene laminated film. The three transponders were al l dipole antenna transponders. These evaluations were conducted in a machinery lab at the MSU School of Packaging. A testing methodology consisting of two sub -tests was developed based on the results of pilot testing. In the first test, the transponderÕs r ead range at given orientations was measured. In the second test, the transponderÕs orientation read rate at a given distance from the interrogator was evaluated. Both linearly polarized and circularly polarized interrogator antenna e were used in each test . Statistical analysis for each test was conducted using Statistical Analysis System (SAS) software. The hypotheses of this study are as below: !4 ¥!Hypothesis 1: Packaging materials will NOT have a consistent effect on read range across different antenna desig ns of general -purpose dipole antenna transponders. ¥!Hypothesis 2: Packaging materials will NOT have a consistent effect on orientation read rate across different antenna designs of general -purpose dipole antenna transponders. 1.3 Thesis Summary Chapter 2, Literature Review , presents an RFID primer, application examples of RFID technology in supply chains, and principles of RFID communication , as well as relevant research about evaluation of RFID system performance and development of methodologies for evaluation. Chapter 3, Methodology , describes the testing equipment, specimen preparation, and testing location. Then the procedures for each sub -test contained in the methodology are present ed. Chapter 4, Results , presents the data collected during this research. Chapter 5, Analysis and Discussion , explains the statistical model for each test and discusses the results. Chapter 6, Conclusion s and Future Research , summarizes the new lea rned res ults, proposes avenues for future study, and discusses the limitations of this work. !5 Chapter 2 Literature Review 2.1 Automatic Identification and Data Capture Technologies AIDC technologies come in a wide variety of functionalit ies , all targeted towards automatically identifying a physical object with associated data contained within an information -technology (IT) system [16] . Barcodes, optical character recognition (OCR), biometric procedures (e.g. voice identification), smart cards, and RFID are some of the common forms of AIDC [17] . AIDC technologies ar e often combined together in practical applications to provide a more reliable base of information flow [2] . To date, among all the methods of AIDC, by far t he most commonly used one is barcod e technology ---- a machine -decipherable bar code on an object, scanned using an opti cal laser scanner to extract identifying information [18] . RFID is the use of radio communication to identify a physical object [9] . Compared to barcode s, RFID operates at a much greater d istance with higher rates, can carry more information than just the ID, and may also identify an object without line -of-sight [19] . In a ddition, it provides faster processing , increased security and accu racy of information, as well as better anti -counterfeiting function with minimal or no manual intervention [16], [20] . 2.2 History of RFID RFID has existed for more than half a century, but its widespread application has had to wait for inexpensive integrated circuits to enable small, low -cost transponders. Since the mid -1990Õs, the application of RFID in supply chain management had drawn a lot of attention. In 1999, MIT !6 launched the Auto -ID Center, supported by various parties in the industry. After research on various alternative approaches, the Auto -ID Center concluded that 900 -MHz regime is the best operating frequ encies for RFID transponders in consideration of cost, read range and capability [18] . In 2003, EPCglobal Inc. was found ed as an independent and non -profit organization to develop industry -driven standards to support the adoption and implementation of RFID and the Ele ctronic Product Code (EPC, a code system) technology. In 2004, the Generation 2 protocol was developed by EPCglobal and in 2006 was endorsed by the International Organazation for Standardization (ISO) as ISO 18000 -6 [18] . At the same period of time, Wal -Mart and the US Department of Defense (DoD) began to mandate the use of RFID to track shipping cases, containers and pallets. By 2010, it was fair to con clude that RFID technology is an appropriate data carrier technology for tracking goods [18] . 2.3 Components of RFID In a basic backscatter RFID system, three fundamental components are required for data to travel: transponder, interrogator with antenna , and host computer with middleware. 2.3.1 Transponder A transponder, also called a tag, generally consists of a chip, which is an integrated circuit (IC) made of silicon, an antenna that is made of metal or metal -based material, and a substrate that houses the chip and the antenna. A transponder is programmed with information that can identify !7 itself uniquely [9] . Tags/transponders fal l into three broad categories based on power source : passive tag s, active tag s, and semi -passive tag s. Passive tags have no independent source of electrical power. They depend on the power transmitted from a n interrogator to drive the circuit and modify th eir interaction with the transmitted power in order to send information to the interrogator. Depend ing on the operation frequency, passive tags can be read up to 20 meters (about 65 feet). They are inexpensive, with a cost of a few cents to a dollar , and can be used for tracking consumer goods [18] . Passive tags dominate supply chain s for cost reason s [1] . Active tags have a local power source such as a battery , as well as a conventiona l transmitter. They are able to initiate communication by sending their own signal. Active tags have a much greater operation al distance compared to pass ive tags, and are priced from 20 to over 100 dollars . They are used for expensive objects with long use ful lifetimes [18]. Semi -passive tags have a local battery but only for turning on the tag circuit. Similar to the passive tags, semi -passive tags require energy transmitted from a n interrogator for the tag -to-reader communication . They can achieve ranges as much as 10 0 meters (up to 320 feet), and are often used in tracking high -value reusable assets such as airplane parts. A semi -passive tag usually cost s about 20 to 30 US dollars [18] . 2.3.2 Interrogator An interrogator, also called a reader, is the RFID component that transmits RF signals to the tags, receives information from the tags , and sends information to a host system. It is composed of !8 an RF module, a signal processing and control unit, and a coupling element, which is essentially an antenna. Interrogators can be categori zed by the location of use as fixed -mount interrogator s, handheld interrogator s, and vehicle -mount interrogator s [9] . Antennas are often classified by their polarization ---- either linearly polarized (LP) antenna s or circularly polarized (CP) antenna s. Fro m the basic principles of the oscillation and propagation of electromagnetic (EM) field s, one can know that the electric field direction is perpendicular to the wave propagation direction. If the direction of the electric field is constant in time, the wav e is linearly polarized , and if the electric field rotates around the propagation direction and keeps constant magnitude, the wave is circularly polarized [18] . LP antenna s can be further classified into two kinds: horizontally polarized and vertically polarized, depen ding on the plane on which the wave travels relative to the earth [9] . In theory, if the LP transmitting antenna that generates a horizontal electric field, and a receiving antenna are vertically polarized, the power transfer will degrade considerably. Howeve r, with the same LP transmitting antenna, a CP receiving antenna will receive some radiation no matter what angle the receiving antenna inclines within the plane perpendicular to the EM wave propagation direction [9], [18] , but in every case only half the transmitted power can be rec eived [18] . T he reader and the tag antennas should have the same polarization in order to achieve the maximum power transfer efficiency [9] . !9 2.3.3 Host Computer and Middleware The host computer installed with middleware provides two functions: 1) to send request s and receive response s from the interrogator, and 2) to filter, store and forward the received data to other systems [9], [21] . 2.4 RFID Standards, Regulations and Protocols 2.4.1 RFID Standard Bodies RFID standards help to ensure products from different manufacturers operate to gether and provide guidelines for manufacturers to develop complementary products [22] . There are four main types of RFID standards: technology standards, data standards, conformance standards, and application standard s [23] . ISO and EPCglobal are two main international RFID standards bodies. ISO set up a joint committee with the International Electrotechnical Commission ( IEC ) to look at standardization for RFID technology since 1996 [24] , and EPCglobal was founded in 2003 [18] , as mentioned previously in this chapter . In the US, the organization that develops and issues RFID standards is the Federal Communication Commission (FCC). In Europe, it i s the European Telecommunications Standards Institute (ETSI) [22] . 2.4.2 Frequency Allocation The frequency used by a n RFID system determines many of the characteristics of the system [25] . Regulatory bodies allocate different frequency bands as follows, and RFID systems are !10 available in all the radio frequency ranges [9], [25] : ¥!125 -134.2 kHz and 140 -148.5 kHz , Low Frequency (LF) ¥!6.765 -6.795 MH z, M edium Frequency (MF) ¥!13.56 MHz, High Frequency (HF) ¥!433 MHz, UHF ¥!860 -960 MHz, UHF ¥!2.45 GHz, Super High Frequency (SHF) . LF is often used for vehicle identification. HF is typically used for contactless payment, access control , and electronic ticketing. UHF is the only spectral region that does not have uniformly allocated frequency bands around t he world. Common use of UHF includes container tracking and asset management. SHF is usually used by an active system for long range tracking [25] . Generally speaking, higher frequency system s provide a larger read range but less tolerance to interference [2] . Table 1 shows an overview of UHF regulations for passive RFID system s within the 860 to 960 MHz band in several countries and regions, along with the maximum power allowed [26] . The Tags discussed hereinafter refer to passive UHF RFID tags if not specified. In Table 1, the maximum power was expressed as EIRP (equivalent isotropic radiated power) or ERP (effective radiated power). EIRP and ERP are measured by different methods, and approximately, 2 Watts ERP is equal to 3.2 Watts EI RP [26] . !11 Table 1 UHF Regulations for Passive RFID Within the 860 to 960 MHz Band in Different Countries and Regions Country Frequency in MHz Maximum Power Brazil 902-907.5 4 W EIRP 915-928 4 W EIRP Canada 902-928 4 W EIRP China 920.5 -924.5 2 W ERP Europe 865.6 -867.6 2 W ERP India 865-867 4 W ERP Japan 916.7 -920.9 4 W EIRP 916.7 -923.5 0.5 W EIRP Korea 917-920.8 4 W EIRP 917-923.5 200 mW EIRP Mexico 902-928 4 W EIRP South Africa 865-867.6 2 W ERP 915.4 -919 4 W EIRP 919.2 -921 4 W EIRP United States 902-928 4 W EIRP 2.4.3 RFID Protocols Protocols are of importance for the communication of RFID systems. They define the physical layer and the tag -identification layer of the communication between RFID components [27] . For p assive RFID systems which operate at 860 - 960 MHz, the most common and globally accepted protocol is the EPCTM Radio Frequency Identity Protocols, Class -1 Generation -2 UHF RFID Protocol for Communications at 860 MHz Ð 960 MHz . It is often referred to as the Gen 2 protocol. In a passive RFID system compliant with the Gen 2 protocol, the reader sends a modulated RF signal that contains information to one or more tags. The tag is powered up by this modulated RF signal. The reader receives a response from the tag by transmitting a continuous -wave (CW) !12 RF signal that does not contai n any information to the tag while listening for a backscattered reply. The tag responds by backscatter modulating the amplitude and/or phase of the RF signal. The modulation methods used in the Gen 2 compliant systems include double -sideband amplitude shift keying (DSB -ASK), single -sideband amplitude shift keying (SSB -ASK), and phase -reversal amplitude shift keying (PR -ASK) [27] . In addition to the Gen 2 protocol, EPCglobal also issued documents about conformance requireme nts and interoperability test system s for certifying UHF RFID devicesÕ compliance to the Gen 2 protocol [28], [29] . 2.4.4 The Electronic Product Code The Electronic Product Code ( EPC ) is syntax for unique identifiers assi gned to any entity that is identifiable in business processes and applications [30] . RFID is a main carrier technology for the use of EPCs [31] . When used with RFID tags, EPCs are encoded in binary forms that carry information within RF signals . By accessing the database, EPCs can be represent ed in text form and are suitable for data sharing a mong enterprise information system s [30] . 2.5 Applications of RFID The use of RFID in supp ly chain covers a large part of RFID applications. RFID in the supply chain delivers better security, increased efficiency and visibility, increased process and data accuracy, be tter counterfeit detection, and easier product recall. As a result, members of the supply chain can potentially benefit from the implementation of RFID systems [20] . In this section, examples of RFID application s in the tran spo rt and logistics industry, as well as in the retail store environment , will be discussed, followed by a successful RFID application case. !13 2.5.1 RFID in the Transport and Logistics Industry A cargo consists of various layers: item level (layer 0), package l evel (layer 1), case level (layer 2), pallet level (layer 3), container level (layer 4), and vehicle level (layer 5). S hipper s will focus more on layer s 0, 1 and 2, and warehouse operator s will pay more attention to layer 3. Port terminal operators will fo cus on layer 4. Different RFID readers and tags should be used for different layers to achieve optimum results. However, for the majority of applications within the logistics industry, passive RFID system s are adequate [32] . For different layers of cargo, tags are read at different points. First, for warehouse s and shippers, items and cartons are usually read at the tag commissioning. Pallet tags are read when the cartons have been loaded in the case that the pallet has already been equipped with a tag (e.g., a returnable pallet embedded with RFID tags). If the pallet does not have a tag, the tag will be attached and read at the tag commissioning. Pallets are also read at the dock door when being loaded onto the vehicle. Typicall y, the conveyance is read at gate in and gate out, as well as when it arrives at the dock door. Second, for carriers, tags are read at the loading and unloading process. And last, for terminal operators, tags are read at gate in and gate out. [32] . The use of returnable pallets, or other forms of returnable transport items (e.g. returnable totes and bins) is a trend in the modern supply chain. It offers reduced costs, increased handling efficiency, as well as improved env ironmental sustainability. Integrating RFID tags into returnable pallets brings benefits from both product logistics and asset management point s of view. To assume a minimal level of process security, at least two tags are suggested to be attached on a pal let. For !14 wooden pallets, one tag should be attached on the longer side and one on the shorter side of the pallet. For plastic pallets, it is recommended that the two tags be placed on either pair of the diagonal corners [33] . Cargo container security and tracking revenues were expected to grow at a c ompound annual growth rate (CAGR) of 27 percent from 212 million in 2011 dollars to 690 million dollars in 2016 [34] . However, the customs inspection process at international borders, especially in maritime ports, is inefficient , time consuming, costly , and only manually operated [35] . Electronic seals (e -seals) are a solution to increase efficiency, security and visibility, as well as decrease cost s of importing and exporting goods. Cargo containers with e -seals can be automatically inspected without opening the container doors, since all the co ntent information can be accessed through reading the e -seal. RFID based e -seals, compared to the e-seals based on other technologies, have been estimated to provide greater Net Present Value (NPV), higher Return On Investment (ROI) and a lower payback period [36] . 2.5 .2 RFID -based Electronic Article Surveilla nce (EAS) System in the Retail Store EAS at retail stores is used for loss prevention and dete ction to reduce theft including shoplifting. Compared to RF based (RF -EAS) and Acousto -magnetic based (AM -EAS) system s, which are two dominant EAS technologies currently, an RFID based system adds additional benefits by reducing out -of-stock merchandise, a utomating replenishment, eliminating manual processes and enhancing the consumer buying experience [37], [38] . A typical implementation of an RFID -based EAS system in the retail store environment !15 would be: 1) items are received in the back room, and the EPC number that is encoded on the RFID tag for each item is read at the door and added to the EAS database; 2) when the merchandise is brought to the retail area, the RFID tags can be read at the entry from the back ro om to the retail area; 3) staff are able to use handheld readers to locate and manage merchandise in the retail area as well as in the back room; 4) at the point of sale, the EPC numbers are removed from the EAS database after purchase, and consumer notifica tion occurs; 5) at the point of exit, the RFID -based EAS system automatically verifies sales and activates alarms if unsold items are detected [39] . 2.5.3 Successful Case of Benefiting from RFID in Supply Chain Levi Strauss & Co. started a pilot test of integrating EPC -enabled RFID technology into operations in its Mexico facilities in 2005. With the in itial success of the pilot test, in 2007, the company decided to sequentially implement the EPC -enabled RFID solution in its remaining stores. Highlight s of the positive business benefits that Levi Strauss experienced from the integration of EPC/RFID technology are listed below: ¥!Reduced inventory in stores from a four -month to a two -month supply; ¥!Improved inventory accuracy and better inventory details; ¥!Increased sales by 11 percent; ¥!Reduced lost sales by 40 percent because of reduction in out -of-stock merchandise; ¥!Increased efficiency in distribution center (DC) and reduced logistics costs; ¥!Enabled fully automated replenishment management [38] . !16 2.6 Principles of RFID In order to investigate pas sive UHF RFID tagsÕ performance, understanding the physics of passive UHF RFID system s is critical. This section will illustrate communication principles of a passive UHF RFID system in free space under ideal condition. The system is compliant to the EPCgl obal Gen 2 protocol. 2.6.1 Near Field and Far Field In an EM field, there is a distance from the antenna that defines the near field and the far field . Inside this distance, the zone is called the near field, and systems operating in the near field depend on inductive coupling between the reader and tag antennas. Outside of this distance, the EM field separates from the antenna , and propagates into free space as a plane wave. Typically, LF and HF systems operate in the near field, while UHF systems operate in both the near field and far field, depending on the distance between the transponder and the interrogator [40] . For small antennas in which the maximum dimension D of the radiating structure is considerable smaller than the wavelength $ (!"#), the approximate distance where the far field starts is $%#&' (1) where r is the distance from the antenna. For any antenna, the zone inside $%()* can be defined as the reactive near field , where the electric field E and magnetic field H are not orthogonal. In the case when !+#, the far field is estimated to start from !17 $%&!)# (2) The region between a reactive near field and a far field is called the radiating near field, as given below [40] : #&',$,&!)# (3) The signal power produced by an antenna within the near field region, according to the solution of MaxwellÕs equations, drops off with distance -.$/ [15] . 2.6.2 Forward Link Communication The forward link in an RFID system is defined as the signaling from the reader to the tag [41] . Assuming it has an isotopic antenna that radiates power in all directions uniformly , the power flux density S , at a distance r from the antenna can be expressed as follow s: 0%123456'$) (4) where 12345 is the power produced by the readerÕs transmitting antenna . However, for a n antenna in the real world , the power flux density is directional with a gain of 72, so the actual power flux density, which is called the directional power flux density , transmitted by a reader antenna is [40] 08%123456'$)972 (5) Then the amount of power received by a tagÕs receiving antenna when the tag is placed at a distance r from the readerÕs transmitting antenna can be expressed as 14325%09:;%123459726'$)9#)6'974%123457274<#6'$=) (6) where the :; is the effective area of the tag antenna given by an isotropic antennaÕs aperture multiplied by the gain of the actual tagÕs antenna 74 [42] , !18 :;%#)6'974 (7) When the 14325 exceeds the minimum required power ( 14>?@ AB@ ) for the tag IC to be switched on, the tag is activated. Thus, the maximum distance to activate the tag in the forward link $CDE AF can be calculated by substituting 14325 with 14>?@ AB@ in equation 6 and is expressed as $CDE AF%#6'G12345727414>?@ AB@ (8) 2.6.3 Reverse Link Communication The reverse link in an RFID system is defined as the signaling from the tag to the reader [41] . From the tag designÕs point of view, the factors that affect the tagÕs pe rformance include: 1) the impedance match between the tag antenna and the chip [43] ; 2) the minimum required power for the IC to be turned on, i.e. 14>?@ AB@; 2) the capability of the tag antenna to collect wireless energy, i.e., the tagÕs energy harvesting ability; and 4) the differential radar cross section ( % RCS) of the tag that determines the clarity of the tagÕs backscattered signal [44] . Typically, HIJKL AML of a tag is -10 dBm to -15 dBm [44] . When the tag is turned on by the readerÕs transmitting signal, the tag sends data back by modulating its reflection coefficient, then backscattering the signal containing information to the reader [41] . A tagÕs power reflection coefficient NON) is interpreted as NPN)%QRSTRURSVRUQ) (9) where the WX%YXVZ[X is the complex antenna impedance, and W\%Y\VZ[\ is the complex IC impedance [45] , and s is defined as the power wave reflection coefficient [43] . When !19 modulating the backscattered signal, the tag switches its input impedance between two states. In order to have a clearly modulated backscatter, usually one state is high impedance and the other one is low [42] . Antenna impedance is typically made to match the high impedance state of the chip for maximizing the power transfer efficiency [42] . The power transfer efficiency, or the power transfer coefficient , can be expressed in IC and antenna impedance as ]%6^S^UNRSVRUN) (10) If YX%Y\, and WX%W\, then & =1 and the maximum power is transferred. The sum of the power transfer coefficient and the powe r reflection coefficient is 1 [43] . ]VNPN)%- (11) The radar cross section (RCS) is defined as Òa measure of the ratio of backscatter power per steradian in the direction of the radar (from the target) to the power density that is intercepted by the target Ó [46] . Each of the tagÕs impedance states represents a n RCS [42] , and the difference between the two RCSs, i.e. % RCS, identifies the quality of the ASK modulation of the tag [44] . The % RCS can be expressed as [44] , _`%74)#)-a'NPbTP)N (12) where Pb and P) are the power wave reflection coefficients of the two impedance states, and cI is the ta g antenna gain. By understanding the factors of the tag, the amount of power the reader antenna receives 12325 can be calculated as !20 12325%143259d-TNPN)e7472<#6'$=) (13) Substituting equation (6) in equation (13), then 12325 is 12325%123459d-TNPN)e74)72)<#6'$=f (14) Finally, by defining the minimum signal power for demodulation at the reader as 12ACg@ , the maximum range in the reverse link $CDE A2 can be obtained [18] : $CDE A2%#6'G1234572)74)d-TNPN)e12ACg@ 6 (15) Generally, the read range of a passive UHF RFID system depends on the forward link, as the reader sensitivity is much higher than the tag [47] , i.e., 12ACg@ "14>?@ AB@. 2.7 Transponder Performance in the Real World An ideal RFID system, in which the reader and tagÕs performance perfectly and strictly follows its designed functionality, has the following characteristics: 1)!Orientation of tags does not affect the tag performance; 2)!The object to which the tag is attached does not affect the tag performance; 3)!The environment in which the system is operated does not affect tag performance; 4)!The interrogation zone produced by the reader has a clear boundary; 5)!All and only the tags in the readerÕs interrogation zone can be read; 6)!Relative motion between the tag and the reader does not affect the tag performance; 7)!The system performance is not affected by the presence of multiple readers and/or multiple tags [19] . !21 However, an ideal RFID system is unrealizable. The performance of a tag while in a readerÕs interrogation zone can be influenced by the object the tag is attached to, the orientation and relativ e motion of the tag, as well as the environment al interference. 2.7.1 Product and Packaging Materials The packaging and the product inside of the packaging have a significant effect on the performance of RFID systems [1], [2], [8] . When a RFID tag is attached to a packaged product, the tagÕs readability may degrade considerably, depending on what materials the packaging and the pr oduct are made of. There are five main categories of packaging materials: wood, paper, plastic, glass , and metal [15] . Materials can be also classified into three categories: dielectric, conductor , and magnetic materials [43] . Since magnetic materials are relatively rare, and barely used for packaging, they will not be discussed in this section . Dielectric materials can be characterized through two propert ies: the dielectric constant, also called the permittivity, and the loss tangent [43] . The permittivity ' is a complex, frequency -dependent quantity expressed as h%ibjZi) (16) The loss tangent is the ratio of the imaginary part Im( ') to the real part Re( '), and is usually written as tan ( [15] . Dielectric materials have a tan ( less than 0.01, and conductors have tan ( larger than 100. Materials that have a tan ( between 0.01 and 100 are defined as semi -conductor s [48] . Dielectric materials are basically electrical insulators. They resist the electric field when in the presence of an electric field. The permittivity of a material can be interpreted as the amount of !22 counteraction the material has to the electric field. The relative permittivity iK is often used to characterize a material. iK of a material is defin ed as h.ik, where ' is the materialÕs permittivity and hk is the permittivity of a vacuum. When a wave travels in a dielectric material with relative permittivity iK, the velocity of the propagating wave is l%mniK (17) where c is the speed of light in a vacuum. T he wave slows down in the material, and its wavelength decreases by a factor of -.nh? [43] . If a n RFID tag is completely immers ed in a dielectric material that has a relative permittivity h?, the tag antenna is larger by a factor of nh?, compared to the wavelengt h in the material, and the resonant frequency of the tag antenna will be decreased by a factor of -.nh?. Chang es of the res onant frequency normally result in a considerable impedance mismatch between the tag ante nna and IC, therefore influencing the tagÕs p erformance [43] . The interference that affects a tagÕs performance by changing its impedance is called detuning [47] . In reality, RFID tags are typically on some objects rather than imm ersed in materials. Thus , the tag can be regarded as partially surrounded by the object and partially by the air, if the object is relatively thick compa re to the size of the antenna , then the effective relative permittivity can be estimate d as oipqq riKVistK & (18) Although this estimation may be not appropriate for substance s with large h?, it works well for small h? [43] . The h? of commonly used packaging material s, such as polypropylene, !23 polyethylene and polyethyleneterephthalate (PET) , is relatively small (see Table 2). Paper and corrugated board also have small h? when they are dry, but the h? will change if moisture is absorbed. The loss tangent tan ( determines how much energy is lost to the material per wavelength. The larger the loss tangent of a material, the poorer the wave propagates in the material , and the more energy is converted into heat [43] . Some of the common materialsÕ relative dielectric constant s and loss tangent s are listed in Table 2 [49], [50] . It sho uld be mentioned that the dielectric property of many substance chang es not only w ith frequency and temperature, i t changes even with specimens made from the same material but with different manufacturing processes, different amounts of oxidation, and many other factors. Thus, the numbers listed in Table 2 only serve as an indication, and are not precise data that are repeatable for a particular specimen [49] . The dielectric property of water, specifically, when tested at 20 ¡C with frequency of 1 MHz and 3 GHz, shows a relative permittivity about 80 at both frequencies, and a loss tangent 0.04 at 1 MHz, while the loss tangent is 0.157 at 3 GHz [50] . The challenge of RFID around water comes from the resulting change in the antenna impedance, and the modification to the radiation pattern, as the antenna tends to radiate more energy into the water [43] . !24 Table 2 Dielectric Properties of Materials Tested at Certain Frequencies and Temperatures Material Temperature ¡C Frequency f !r 104" tan # Glass (borosilicate) 20 1 kHz /1 MHz 5.3 50 /40 Paper (kraft, tissue, unimpregnated, dry) 20/90 1 kHz 1.8~3.0 10~35 Wood (balsa, 0% water content) 20 50 Hz/3 GHz 1.4 /1.2 40 /140 Wood (scots pine, 15% content) 20 1 MHz/100 MHz 8.2 /7.3 590 /940 Polyethylene 20 50 Hz/1 GHz 2.3 2 /3 Polypropylene 20 50 Hz/1 MHz 2.2 5 Polystyrene 20 50 Hz/1 GHz 2.6 2 /5 Polyethyleneterephthalate (PET) 20 50 Hz/100 MHz 3.2 /2.9 20 /150 Polyvinylchloride (PVC) 20 50 Hz/100 MHz 3.2 /2.8 200 /100 Polytetrafluoroethylene (PFTE, Teflon) 20 50 Hz/3 GHz 2.1 2 Water 20 1 MHz/3 GHz 80 0.04/0.157 Metals are RF reflective material s and result in distortions of the field homogeneity [47] . If a passive tag is directly attached to a metal, it is expected to have poor performance. The easiest solution for reading tags on the metal is to try to provide some spacing from the metal objects [43] , usually by an offset of 3 mm to 5 mm from the material surface [15] . However, this solution still performs poorly [43] . Tags that are specific ally made for metals are typically based on microstrip antenna s. Compared to common passive tags that are usually ba sed on dipole antenna s, microstr ip antenna tags tend to be relatively expensive [43] . 2.7.2 Research on the Effect of Packaging and P roduct on Transponder Performance Although the effect of materials on RFID systems is already complicated, it becomes more !25 complicated when taking a packaging/product system into account. The materials used for different layers of packaging, the compounds of the packaged product, the water content and water forms of the product, the operating processes of the packaging/product system may all affect the performance of the RFID system [15] . While the e ffects of packaging and product on the RFID are still not completely known [15 ], many researchers have made efforts to investigate the effects: Tazelaar evaluated the effect of tag orientation and package content on the readability of an RFID system and concluded that orientation and product type have a significant effect on tag readability. In TazelaarÕs research, RFID tags were attached to empty cases and cases that were filled with foams, empty PET bottles, rice or water bottles. 48 of the tagged cases with the same content s were stacked onto a pallet with the RFID tags facing in a certain orientation. The tags were then read by running the pallets through a simulated portal installed with reader antennas. 100% of the tags were read in experiments using empty cases, foam -in-place filled cases, and cases filled with empty PET bottles while the tags were facing outward, forward, and upward. For cases filled with rice and with water bottles , certain orientations of the tags resulted in sign ificantly degraded tag performance. In particular, in the test of downward tag orientation using water bottle filled cases , no tag s were read at all. In addition, TazelaarÕs research showed that case location on the pallet played a role in the tagÕs perfor mance, but the effect of case location on performance was not consistent across all product types. Thus, in order to maximize the tagÕs performance for a product/package system, the pallet patterns, and tag orientation and location should all be tested [2] . !26 Falls investigated the effect of conveyor speed, packaging materials, and product on the tagÕs performance. In FallsÕ research, the tag on a case of potato chips in plastic tubs and that on a case of chips in a metalized spiral wound fiberboard container were tested to evaluate the effect of the package type; the tag on a case of metal cans, a case of metal bottles and that on a case of metal tins were tested to evaluate the effect of the pac kage shape; and the tag on a case of bottled ketchup and that on a case of bottled motor oil were tested to compare the effect of the product type. Tags were read when the cases were on a conveyor with a speed of 300 feet per minute and of 600 feet per min ute to investigate the effect of conveyor speed. Additionally , two generations of tags were used in the research. The results showed that conveyor speed, package type, package shape and product type all had a significant effect on the average amount of tag reads per trial. The two tag types used in the research were found to have a significant effect on the average number of reads per trial for product effect and package shape effect, but did not have a significant effect when testing the package type effec t [8] . Onderko evaluated tagÕs performance on refrigerated and frozen beef loin muscle packages. The result showed that RFID systems in 13.56 MHz frequency range performed well with no loss of data while reading the tag on a package of refrigerated or frozen beef. However, systems operating in the 915 MHz frequency range could read tags on frozen beef packages, but experienced difficulty with the refrigerated beef package [6] . Zhang investigated the influence of water content in products on dielectric properties, antenna radiated power, and tag readability. Hydrated superabsorbent (SAP) polymer was used in ZhangÕs !27 research to simulate products with water content from 45% to 90%. Permittivity of the hydrated SAP was measured, and then the effect of water content on tag performance was evaluated. The result s showed that tags could barely be read when evaluated with the hydrated SAP sample that had water content more than 75%. The conclusion was that product content and packaging operations are important factors to the success of employing RFID system in supply chain s [1] . Crawforth investigated the effect of antennae configuration, product and tag type on the tag performance when tagged cases were stretch wrapped on a stretch wrapper. Results of CawforthÕs research showed that the antenna configurations with a c ombination of antennas both on top of the pallet load facing downward and on the side of the pallet load resulted in the greatest read rate of tags, while the configuration of only antennas on the side had a lower read rate, and that of antennas mounted on ly above the pallet had the worst read rate. No significant difference was found between 1 and 2 antennas, or 3 and 4 antennas. In addition, Cawforth found that empty cases had no effect on the readability of tags, rice filled jars degraded the tag perform ance, and water bottle filled cases decrease d the tag readability considerably. The tag type was not found to have a significant effect on the tagÕs total reads, but differences were observed with specific combinations of variables [7] . Ukkonen and Syd−nheimo developed a testing methodology to measure a passive UHF RFID tagÕs radiation pattern based on the tagÕs 14>?@ AB@ and compared a tagÕs radiation patterns when the tag was attached on a case of six -pack metallic cans to that in free space. Results of the measurements showed that the tagÕs in -free -space radiation patterns of the E plane and the H plane are both symmetrical and agreed with the expected antenna radiation pattern, while distortions and !28 change in direction were observed on the tagÕs radiation patterns tested with the case of metallic cans [14] . Last but not least, Suwalak and Phongcharoenpanich found that when a UHF passive tag was attached to an elliptical curved surface, the angles of the curvature of the tag antenna affected the tag performance. The tag performance change was interpre ted as th e tagÕ s input impedance changed with changing elliptical curved surface Õs axis dimensions [51] . Yang et.al tested UHF RFID tag performance in various simulated dielectric background s. The results showed that the tag antennaÕs resonant frequency was proportional to -.nh? when attached on a material with a relative permittivity of h?, and the resonant frequency decreased with t he increasing thickness and increasing size of the material [52] . Bolton et.al performed a study to test 10 different multisurface tags on three mediums (air, metal, and cardboard) at 6 distance s (5 ft, 10 ft, 15 ft, 20 ft, 25 ft, and maximum distance) with four tag orientations (0, 45, 90, and 180 degrees). The results showed that no tag had 90% or higher readability or precision. Blind spots were observed at various distance s and orientations in different medium s, and the author s conclude d that this wa s mainly because of the polarity of the antenna used in the study [53] . All of the research above demonstrated that understanding and evaluating the effects of packaging, products, and packaging related processes on RFID system performance is critical to the success of implementing a RFID system. 2.7.3 Operating Environment Effect on Transponder Performance The practical environment in which a n RFID system is deployed affects the systemÕs !29 performance considerably [3] . A pre -installation site analysis is necessary before implementing RFID technology at a location, as the surrounding physical environment may deviate the performance of a RFID system from the expected [2] . RF wave s travel from the transmitter to the receiver can be affected by a variety of factors related to the practical environment, including absorption, dielectric effects, diffraction, reflection, refraction, scattering, noise and int erference [9] . Absorption occurs if a part of the RF energy is absorbed when the RF wave strikes an object in the field. The amount of energy absorbed by an object depends on the dielectric property of the object, as mentioned in the Product and Packaging Ma terial s section. Water and liquid products absorb RF waves. Wood, paper and paperboard, and food that contain s a certain amount of water may also absorb RF waves [9] . Dielectric effects, also mentioned in the Product and Packaging Material section, refer to t he materialÕs capability to resist the electric field. As a result, the velocity of a wave is reduced, and the signal is detuned. Diffraction happens when a wave strikes sharp edges or when it passes through narrow gaps. Reflection occurs when a wave strik es a surface that has much larger dimension s compared to the wave. Refraction is the phenomenon of the wave changing its direction when it travels through two substance s. The wave direction changes at the boundary of the two mediums. Scattering refers to an object absorbing a wave and then reradiating it in a different direction [9] . Noise is defined as an unwanted electrical wave or energy present in a signal [9] . Electronic !30 motors, light switches, and thermostats are some of the common sources of noise [3] . Interference is the interaction of two or more waves while they are traveling along the same medium. There are two forms of interference: constructive interference and destructive interference. Constructive interference occurs when the resultant wave has a larger amplitude compared to the original waves, and destructive interference is the case when the resultant wave has a sma ller amplitude compared to the original waves [9] . 2.8 Research on Testing Methodology Besides the research mentioned in the section of Research of the Effect of Packaging and Product on Transponder Performance , testing methodologies for evaluating various as pects of RFID tag performance have also been developed by many other researchers and organizations. EPCglobal issued a document titled Tag Performance Parameters and Test Method s to provide a systematic means to evaluate tag performance. The current version of this document was released in 2008. The document addresses testing procedures for determining read range, orientation tolerance, frequency tolerance, interference tolerance, backscatter range, write time and tag proximity of a passive UHF RFID tag. Procedures for testing read range and orientation tolerance are the most releva nt to this research . "#$%&'! ( shows a side view of the test set physical block diagram for tests in the EPCglobal Tag Performance Parameters and Test Method s [41] . !31 !"#$%&'! (!)#*'!+#',!-.!/0'!1'2/!)'/!3042#567!87-59!:#6$&6;!.-&!1'2/2!#$7-?67!16$! 3'&.-&;6<5'!36&6;'/'&2!6<*!1'2/!@'/0-*2 ! All tests should be conducted in a controlled environment, such as an anechoic chamber. The material on which the tag is attached should be tailored to the approximate dimensions of the tag. An RF -inert, styrene foam block is desired while evaluating the t agÕs free space performance. Other materials may be used to evaluate the tagÕs performance on the material [41] . Specific ally , the read range of a tag is measured by positioning the tag in the far field of the transmitting antenna, and then varying the transmitting power until the tag reaches a 50% read rate. The free space read range is reported under the assumption of a 35 dBm transmitting EIRP. The orientation tolerance is defined as the percentage of tag positions where the range is at least half !32 of its maximum read ra nge. It is measured by rotating the tag from ) 90¡ to +90¡ on its horizontal centerline and on its vertical centerline, and varying the transmitting power until the tag reaches a 50% read rate. The result of the orientation tolerance of a tag is again repor ted under the assumption of a 35 dBm transmitting EIRP [41] . EPCglobal also developed a testing methodology for measuring the readability of a n RFID tag when the tag is applied to a case and the case travels through convey portals. In this test methodology, one antenna on ea ch side of the portal is required, and an antenna on the top or bottom of the portal is optional, but no more than four antennas shall be used. The suggested conveyor speed is 625 feet/minute [54] . In addition, EPCglobalÕs Static Test Method for Applied Tag Performance T esting acts as a low -cost alternative compared to tests such as the convey portal test that was mentioned above. The main goal of the static test method is to evaluate the tagÕs read rate when the tag is placed on an item, case, or pallet. An anechoic cham ber is preferred, but when not available , an open area test site with well -defined RF characteristics is acceptable [55] . Last but not the least, GS1 issued a document defining the testing methodology guideline used for the Tagged -Item Performance Protocol (TIPP) tagged -item grading. The TIPP prensents a grading system to classify the RFID performance of a tagged item. This document provides the method and criteria for validating if a TIPP tagged -item meets a specified grade level [56] . ASTM International issued three standard test method s for tag performance eva luation. They are intended for use in laboratory settings that simulate the practical distribution environment of !33 the product being tested [57] Ð[59] . ASTM D7435 -08 Standa rd Test Method for Determining the Performance of Passive Radio Frequency Identification (RFID) Transponders on Loaded Containers is the standard that gave enlightenment to the development of the test methodology used in this research . Two types of transpo nder distance are defined by the ASTM standard , depend ing on how they are tested. The critical transponder distance is Òthe distance between the transponder and the interrogator antenna at which a transponder becomes undetectable by an RFID system, when moving the RFID transponder out of the read field Ó, and the transponder acquisition distance is defined as Òthe distance between the transponder and the interrogator antenna at which a transponder is first detected by an RFID system, when moving the transp onder into the read field Ó [57] . The ASTM D7435 -08 method first determines the read performance of a RF system by measuring the critical distance read field and the tran sponder acquisition distance read field with a transponder attached to an empty container. Both types of transponder distance shall be measured along every 15¡ radian within two of the two -dimensional quadrants in front of the antenna. So a total of 13 rad ii should be tested at a horizontal plane. The height of the interrogator antenna shall be fixed, and the height of the transponder shall be adjusted from 1 foot to 6 feet from the floor with an increment of 1 foot, and be adjusted to the height of the int errogator antenna, if this height is not included in every 1 foot from 1 to 6 feet. All the 13 radii on each horizontal plane shall be tested. After all the tests , the radiation pattern of the interrogator antenna in a three -dimensional space in front of t he interrogator is given. Then the transponder readability in the established read !34 field of the RF system while the tag is attached to a product -filled container is determined by following the same procedures of measuring critical distance and acquisition distance at each plane. In the end, the performance of the transponder when it is attached to the product -filled container is compared to that of the empty container [57] . ASTM D7434 -08 evaluates RFID system performance while the tag is attached to a unit load. The testing procedures are similar to ASTM D7435 -08 with measuring the critical distance and acquisition distance of a transponder affixed to a pallet load by movi ng the pallet load towards or away from the interrogator antenna. The transponderÕs height is fixed but not changed compared to the ASTM D7435 -8 procedures, as ASTM D7434 -08 is intended to simulate the use of a transponder on a pallet load in practical pro cesses [58] . ASTM D7580/D7580M -09 is a rotary stretch wrapper test method for determining transponder performance on homogenous unitized loads. A unitized load is assembled with RFID -tagged cases, and then the tags on the load are read in a stationary RF field while the load being stretch wrapped by an automated rotary stretch wrapping machine [59] . Other than the test methods issued by standards bodies, many researchers have developed different test methods to evaluate transponder performance under different settings. DÕMello, Mathew, McCauley and Markham studied the effects of tagÕs orientation on a n asset on the tag performance in a highly controlled environment, and the effect of tagÕs relative position to the interrogator antenna on the tag performance in a three -dimensional real world environment. Results of DÕMello et al.Õs experiment indicated that orientation of a tag has an enormous impact !35 on readability. The authors questioned the applicability of RFID technology in the real world based on the generally poor performance of the tested RFID tags in both tests, and emphasized that understanding the tag performance under different orientations is critical for real world application s of RFID [13] . Derbek, Steger, Weiss, Preishuber -Pf, Pistauer presented an UHF RFID measurement based on a modular hardware system for evaluating designs of tags and ICs, and provided means to quantitatively anal yze tag performance in various applications. With tests of 20 tag types, the authors made conclusions including: different ICs have huge differences on tag performance; different tag types have large variance in performance consistency; not all tags are ap plicable to global operations; tag types have large difference between orientation sensitivity; read range s of more than 10 meters were measured, and the read range of different tag types is limited by the forward link; some tags have problems at higher po wer levels [47] . Choi, Kim, Cho, Joo and Lee performed a study to d etermine the minimum requirements for ISO/IEC 18046 -3 Information technology - Radio frequency identification device performance test methods test methods for tag performance Ð Part 3: Test Methods for tag performance [11] . Luh and Liu developed a test method based on ICÕs 14>?@ AB@ to meas ure the read distance of UHF RFID passive tags in a mini anechoic chamber [10] . Poque conducted performance analysis of UHF passive RFID tags with respect to read range and orientation in a real world environment setting. The tag was placed on a cardboard box. Little or no read points were found within the tagÕs read range [60] . Ayyer developed methodology to evaluate tag performance in a trauma !36 resuscitation bay setting that is dynamic and time critical [12] . Lu, Chen and Ye created a 3D model to estimate the interference between tags, while each tag was attached to a box and the boxes stacked in a 3 * 3* 3 pile [61] . Zou, Wu and Zhao proposed an automatic testing system with hardware and software architectures for investigating tag performance [62] . Huo, He, Li, She, Zuo and Zhu presented a software -defined test method for measuring RFID system signal transmission and evaluating tag performance [63] . Kosuru and Devours developed a theoretical mod el for evaluation of the tag performance when the tag is immersed in a dielectric medium [64] . Colella, Catarinucci, Coppola and Tarricone presented a UHF RFID tag performance evaluation platform that can automatically calcu late a tagÕs sensitivity, working range, and radiation pattern based on a theoretical formulation of the tagÕs minimum 14>?@ AB@, which was derived from testing with varied frequencies and tag orientations [65] .Stasa, Svub and Benes developed a RFID measurement chamber that can automatically evaluate the performance of tags on objects, and greatly reduced human labor involvement as well as manual ope ration errors [66] . 2.9 Relevancy Literature discussed in this chapt er provide d essential knowledge and insights to prepare, design, and conduct this research . This thesis focused on the evaluation of the performance of general -purpose transponders in a working environment using packaging materials. Performance parameters of read range and orientation read rate were quantified by procedures described in Chapter 3 Methodology. Creative analysis models will be provided in further chapters. !37 Chapter 3 Methodology 3.1 Test Environment The testing took place at th e machinery lab in the School of Packaging at Michigan State University. This room has a large amount of space filled with equipment made from metal ( Figure 2). Robinson of MSU AIRTC evaluated the RF noise of this room and concluded that this ro om showed a clear disturbance in communication capabilities of passive tags due to the RF noise and the reflective surfaces present in this room. Possible signs of both con structive and dest ructive interference were present in this room [3] . This room was chosen to investigate RFID transponder performance in a simulated manufacturing environment. Tests were conducted along an open aisle close to the centerline of the room. Figure 3 shows the layout of the machinery lab and the test location in the room. !Figure 2 The Machinery Lab in the School of Packaging at Michigan State University !38 Figure 3 Layout of the Machinery Lab and the Test Location in the Room !10'!&--;!/';A'&6/%&'!6<*!/0'!&'76/#B'!0%;#*#/4!CDEF !,'&' !;'62%&'*!6<*!&'5-&*'*! /0& -%$0/-%/!/0'!'L!,#/0!/0'!;6M#;%;! B67%'!-.!KNGOP! ¡>L!/0'!;#<#;%;!B67%'!-.!(QGPR! ¡>L!6<*!6!2/6<*6&*!*'B#6/#-G!10'! 6B'&6$'!DE!/0-%$0-%/!/0'!/'2/#<$!A'&#-*!,62!RJGKR SL!,#/0!/0'!;6M#;%;!-.!NNGPR!SL!/0'! ;#<#;%;!-.!(OGTPSL!6<*!6!2/6<*6&*!*'B#6/#-?@ AB@ as low as -20 dBm with a dipole antenna, according to Impinj [68] . One tag for each tag design was chosen from three packs of sample tags to use in this research. Figure 7 shows the actual tags. Table 3 Transponder Descriptions by the Manufacturer Name Size (in) Chip Memory Common Applications AD-227m5 3.74 * 0.33 Impinj ¨ Monza ¨ 5 128 bits ¥!Supply chain, inventory and logistics ¥!Apparel and other item -level retail ¥!Returnable transport units AD-233m5 2.76 * 0.57 Impinj ¨ Monza ¨ 5 128 bits ¥!Supply chain, inventory and logistics ¥!Apparel and other item -level retail AD-381m5 1.97 * 1.18 Impinj ¨ Monza ¨ 5 128 bits ¥!Supply chain, inventory and logistics ¥!Apparel and retail ¥!Library, media, documents and files [Source: Avery Dennison RFID Solutions] !42 !Figure 7 Transponders used in this thesis 3.3 Packaging Materials Packaging materials used in this research were C -flute uncoated corrugated paperboard (M1), polyethylene film of 0.5 mil thickness (M2), polyethylene corrugated board (M3) and PET/Al/polyethylene film (M4). Materials were cut to 7 * 7 inch square samples ( Figure 8). Uncoated corrugated paperboard was used to frame the films. One sample for each material type was used in this research . !Figure 8 Packaging Material Samples !43 To fu lfill the purpose of evaluating tag performance using packaging materials, a transponder support was made specifically for this research ( Figure 9 and Figure 10). Material samples could be plugged into the slot of the transponder su pport. The tag attached to the material sample could be rotated, tilted and inclined to any angle in three -dimensional space using this transponder support. The center of the material sample was 4 feet 4 in ches above the ground when the sample was placed v ertically on the transponder support. Tags were attached to the center of the material samp le. !Figure 9 Transponder Support, Front View !44 Figure 10 Transponder Support, Front View In addition, distance marks were glued on the floor as Figure 9 and Figure 10. These distance marks were made from wood sticks and have thickness, in order to position the transponder support accurately each time. Distance marks were 6 inch es apart. The parameters for the transponder support are shown in Figure 11. Written descriptions follow . !45 !Figure 11 Parameter Definitions ¥!Angle ! : the angle achieved by turning the material sample clockwise around the horizontal axis located at the bottom of the sample. When ! =0¡, the packaging material sample is horizontal and the tag on the sample is facing downward. When ! =90¡, the sample is vertical. uvwx,-yuv. ¥!Angle " : the angle achieved by spinning the material sample clock wise along the vertical axis. When ! =90 ¡ and " =0¡, the tag is facing the antenna. uvwz,{auv. ¥!Angle # : the angle achieved by spinning the tag on the material sample counter clockwise around the axis that is perpendicular to the material sample and through the center of both the tag and the material sample. When ! =9 0¡, " =0¡ and # =0¡, the tag is in its preferred orientation to the antenna. ¥!Distance r: the distance between the center of packaging material sampleÕs bottom edge to !46 the vertical plane that the antenna surface is on. The transponder support was always moved along the line that ran across the center of the antenna and perpendicular to the antenna surface. 3.4 Pr ocedures The testing procedures used in this research were developed from two pilot tests. Details of the pilot tests are provided in Appendix 1 and 2 . Conclusions of the pilot tests that contributed to the development of the testing methodol ogy in this ch apter include : 1) critical distance and acquisition distance of a transponder are statistically the same, in other words, the maximum read distance of a tag measured by moving the tag towards or outwards from the readerÕs interrogation zone does not signif icant ly differ; 2) angle ! does not have a significant effect on the read range of a tag; 3) angle # may influence the read range of a tag, whether the tag communicates with a CP reader antenna or a n LP reader antenna; and 4) material may have inconsistent effect s on a tagÕs performance, that is to say, tag AÕs read range tested with material A is greater than that of with material B, but tag BÕs read range tested with Material A is smaller than that with material B. Based on the above conclusions, tests for this research were conducted using with following procedures. 3.4.1 Read Range Test The purpose of this test is to measure the transponderÕs read range using different packaging material s. Angle ! and angle # were at fixed value s of ! =90¡ and # =0¡ throughout the entire read range test. Packaging material samples and transponders were conditioned in the machinery room for at least 48 hours prior to the test. After conditioning, procedures of the read range test were as follows : !47 1)!Assemble the LP antenna to the reader, and configure the reader to a transmitting power of 20 dBm, and a receiving sensitivity of -70 dBm; 2)!Assemble the packaging material to the transponder support and set the angle ! t o 90¡, keep ! =90¡ for all the procedures; 3)!Attach Tag 1 to the material sample M1 at # =0¡ using a small piece of masking tape. T he center of the tag is aligned with the center of the material sample. Angle # is kept as 0¡ for all the procedures; 4)!Rotate angl e " to 0¡, and move the transponder support to the distance r=0.5 foot; 5)!Turn on the reader to automatically read the tag 3 times, 5 seconds each time. Delay between two runs is 5 seconds . Each 5 -second run was considered as one trial of the measurement ; 6)!Save the data file with information of reader antenna polarization, tag number, material number, angle " and distance r in the filename; 7)!Move the transponder support to 1 foot and repeat step s 5 and 6; 8)!Move the transponder support outwards from the antenna w ith a n increment of 0.5 fo ot each time, until the tag cannot be detected by the reader for 5 consecutive increases in distance ; 9)!Record the last distance r the tag could be read as R 0. R0 is defined as the the maximum read distance for the tag at " =0¡; 10)!Move the transponder support back to r=0.5 foot, turn the angle " =30¡ and repeat step s 5 to 9; 11)!Increase the angle " by 30¡ each time until " =330¡; 12)!Repeat step 2 to step 11 for all three tags using samples of four packaging materials; !48 13)!Assemble the CP antenna to the reader and configure the reader the same, repeat step 2 to step 12. 3.4.2 Orientation Read Rate The purpose of this test is to determine how well the tag is read when it is rotated away from its preferred orientation, which is ! =90¡, " =0¡ , and # =0 ¡ for the trans ponders evaluated in this research . 60 sets of random angle combination s !-"-# were generated using Python and listed in Table 4. Angles used in generating the random !-"-# were integer multiple s of 15¡ within th e ranges shown in Figure 11. Packaging material samples and transponders were conditioned in the machinery room for at least 48 hours prior to t he test. The test procedures to determi ne the orientation read rate were: 1)!Assemble the LP antenna to the reader, and configure the reader to a transmitting power of 20 dBm, and a receiving sensitivity o f -70 dBm; 2)!Locate the transponder support at the distance r=2 feet for the entire test; 3)!Attach Tag 1 to material M1 at # =345¡ using a small piece of masking tape. T he center of the tag is aligned with the center of the material sample; 4)!Assemble the materia l sample with the tag to the transponder support and turn the angle ! to 0¡ and angle " to 60¡; 5)!Read the tag 3 times, 5 seconds each time, and 5 seconds delay between each run; 6)!Save the data file with information of reader antenna polarization, tag number, material number, angle combination !-"-# in the filename; !49 7)!Repeat step 3 to 6 for the remaining !-"-# in Table 4; 8)!Repeat step 3 to step 7 for all thre e tags using four packaging material samples; 9)!Assemble the CP antenna to the reader and configure the reader the same, repeat step 2 to step 8. Table 4 60 Sets of Random Angle Combination $¡ %¡ &¡ $¡ %¡ &¡ $¡ %¡ &¡ $¡ %¡ &¡ $¡ %¡ &¡ 3.5 Data Management Test data saved each tim e with the Impinj MultiReader in a comma separated values (.csv) file . The two tests conducted above generated a large amount of .csv files that were difficult to handle for analyzing. Desired information from both the filename and th e content of the .csv files was extracted and written into Excel worksheets usi ng Python. Selected outputs were used in this thesis to draw graphs, conduct analysis and compare tag performance : ¥!Total count of reads (TotCnt) : number of times the interrogator read th e tag in each 5 second run ; !50 ¥!Maximum received signal strength indicator (RSSImx) in dBm: the maximum signal strength the interrogator received from the transponder; ¥!R" and R max in feet: for a given tag and a given material tested with a given polarized antenna in the read range test, the maximum distance the tag could be read at angle " , and the maximum distance the tag could be read among all " s, respectively; ¥!Orientation read rate (O -RDRate): the ratio of the amount of angle combinations a tag could be read out of all 60 sets of !-"-# in the orientation read rate test. !51 Chapter 4 Results 4.1 Results of the Read Range Test In the read range test, the tag with the m aterial sample was read at a distance r from the reader, and with an orientation of fixed ! =90¡ and #=0¡, but a changing " every 30¡ from 0¡~330¡. The measurement at each distance r with angle " was performed as described in section 3.4.1, and the TotCnt a nd RSSImx were recorded. Since each measurement consisted of 3 trails, and each trial was a 5 -second run of the reader, there were three TotCnt and three RSSImx values recorded for each measurement. The integer of the average TotCnt of three trials, and th e value of the ma ximum RSSImx of three trials were used to represent the TotCnt and RSSImx value of this measurement. T hen the following terms were defined: A Read Point was defined as a measurement with an average TotCnt value of at least 1. A No-read Point was defined as a measurement with an average TotCnt less than 1. The No -read Points did not have any RSSImx value, since the reader could not detect any signal from the tag. Tables Table 5 to Table 10 are summaries of the read range test s of the three tags with the four packaging materials using the LP an d CP antenna. In the tables, the count of Read Points for a tag with each material is first given, then the maximum, minimum, mean, and standard deviation values of TotCnt and RSSImx were calculated from all the Read Points of the tag with a given material. The R max , the Angle " at where the R max appeared, and the average value of R"s are !52 also provided. Finally , the table shows whether a No -read Point was found within the tagÕs read range as outlined by the R " value and shown in figures from Figure 12 to Figure 29. In this chapter, general dis cussion of the observed results will be given. Statistical analysis of the results are given in depth in Chapter 5. Table 5 Summary of the Read Range Test of Tag 1 with Four Materials Using LP Antenna LP Antenna, Tag 1 Material M1 M2 M3 M4 Count of Read Points 199 209 211 0 TotCnt of Read Points Max 431 430 428 Not Applicable Min 2 5 2 Mean 337 344 329 Deviation 134.98 124.97 137.67 RSSImx of Read Points Max -24 -27 -23 Min -62 -64 -62 Mean -51 -53 -51 Deviation 8.53 8.24 8.50 Rmax (ft) 14.0 14.5 15.0 Angle % of R max 0¡, 150¡, 180¡, 330¡ 0¡ 0¡, 180¡ Average of R % (ft) 9.0 9.1 9.2 No-read Point in the Read Range Yes Yes Yes Table 6 Summary of the Read Range Test of Tag 1 with Four Materials Using CP Antenna CP Antenna, Tag 1 Material M1 M2 M3 M4 Count of Read Points 124 125 127 0 TotCnt of Read Points Max 430 430 431 Not Applicable Min 8 5 3 Mean 338 332 339 Deviation 140.09 139.95 138.33 !!53 Table 6 (contÕd) !CP Antenna, Tag 1 Material M1 M2 M3 M4 RSSImx of Read Points Max -31 -34 -31 Not Applicable Min -62 -64 -62 Mean -52 -54 -52 Deviation 7.97 7.75 8.05 Rmax (ft) 9.0 9.0 9.0 Angle % of R max 0¡, 180¡ 0¡, 180¡ 0¡, 180¡ Average of R % (ft) 5.4 5.6 5.7 No-read Point in the Read Range Yes Yes Yes Table 7 Summary of the Read Range Test of Tag 2 with Four Materials Using LP Antenna LP Antenna, Tag 2 Material M1 M2 M3 M4 Count of Read Points 142 157 158 0 TotCnt of Read Points Max 430 431 431 Not Applicable Min 2 3 1 Mean 335 328 328 Deviation 150.20 147.98 148.63 RSSImx of Read Points Max -30 -32 -29 Min -64 -64 -63 Mean -53 -55 -53 Deviation 7.95 7.69 8.02 Rmax (ft) 13.5 13.5 14.0 Angle % of R max 0¡ 0¡, 180¡ 0¡ Average of R % (ft) 6.6 7.0 7.0 No-read Point in the Read r ange Yes Yes Yes !54 Table 8 Summary of the Read Range Test of Tag 2 with Four Materials Using CP Antenna CP Antenna, Tag 2 Material M1 M2 M3 M4 Count of Read Points 95 96 95 0 TotCnt of Read Points Max 430 430 431 Not Applicable Min 3 2 4 Mean 326 322 331 Deviation 150.87 150.84 145.18 RSSImx of Read Points Max -36 -39 -36 Min -64 -65 -64 Mean -55 -57 -54 Deviation 7.23 6.94 7.31 Rmax (ft) 6.0 6.0 6.0 Angle % of R max 0¡, 30¡, 150¡, 180¡, 210¡, 330¡ 0¡, 30¡, 150¡, 180¡, 210¡, 330¡ 0¡, 150¡, 180¡, 210¡, 330¡ Average of R % (ft) 4.0 4.0 4.0 No-read Point in the Read Range No No No Table 9 Summary of the Read Range Test of Tag 3 with Four Materials Using LP Antenna LP Antenna, Tag 3 Material M1 M2 M3 M4 Count of Read Points 83 94 95 0 TotCnt of Read Points Max 430 430 430 Not Applicable Min 3 5 10 Mean 341 337 340 Deviation 147.55 137.23 139.11 RSSImx of Read Points Max -35 -37 -34 Min -62 -64 -62 Mean -53 -55 -53 Deviation 7.17 7.03 7.37 Rmax (ft) 5.5 6.0 6.0 !!55 Table 9 (contÕd) LP Antenna, Tag 3 Material M1 M2 M3 M4 Angle % of R max 0¡, 150¡, 180¡, 330¡ 0¡, 180¡ 0¡, 180¡ Not Applicable Average of R % (ft) 3.5 4.0 4.0 No-read Point in the Read Range No Yes No Table 10 Summary of the Read Range Test of Tag 3 with Four Materials Using CP Antenna CP Antenna, Tag 3 Material M1 M2 M3 M4 Count of Read Points 55 56 57 0 TotCnt of Read Points Max 430 430 430 Not Applicable Min 4 8 10 Mean 323 308 315 Deviation 153.13 160.27 156.31 RSSImx of Read Points Max -41 -44 -41 Min -63 -65 -63 Mean -55 -57 -55 Deviation 6.15 5.96 6.27 Rmax (ft) 3.5 3.5 3.5 Angle % of R max 0¡, 180¡, 330¡ 0¡, 150¡, 180¡, 210¡, 330¡ 0¡, 30¡, 150¡, 180¡, 210¡, 330¡ Average of R % (ft) 2.3 2.3 2.4 No-read Point in the Read Range No No No Comparing within each table to evaluate the effect of materials on the tagÕs performance in the read range test, the observations include: 1)!There was no read point for M4 for any tags using either antenna ; !56 2)!Material s M1, M2 , and M3 result ed in similar values of count of Read Points. The greatest difference was obtained in Table 6 between M1 and M3 , with M3 having 16 more Read Points than M1; 3)!Material s M1, M2 and M3 resulted in similar values of TotCnt. For a given tag with M1, M2, and M3, the maximum values of TotCnt were around 430, the minimum values of TotCnt were in the range of 1 to 10, the mean values of TotCnt were from 308 to 344, and the deviation values were from 124 to 160; 4)!M2 showed smaller maximum, minimum and mean values of RSSImx than M1 and M3. The differences were within -3dBm; 5)!The Rmax of a tag with M1, M2, and M3 were either identical or no more than 1 foot difference; 6)!The angle " where the Rmax was measured had subtle difference between materials, but all the " angles of Rmax with material s M1, M2 and M3 were within the ranges of 0¡±30¡ and 180¡±30¡ (0¡ ) 30¡ is equal to 330¡); 7)!Material s M1, M2 and M3 had little effect on the average value of R" with a difference that was no more than 0.4 foot for each tag; 8)!Only Tag 3 tested with LP antenna (Table 9) had a difference in the No -read Point in the read range between M1, M2 and M3. Comparing between Table 5, Table 7, and Table 9, or comparing between Table 6, Table 8, and Table 10, tags with different anten na designs showed the following in the read range t est: 1)!Tag 1 had more Read Points, larger R max and larger average R " than Tag 2, and Tag 3 had the !57 least Read Points, smallest Rmax and smallest average R"; 2)!All three tags had very similar value s for TotCnt; 3)!The maximum and mean values of RSSImx of Tag 1 were greater than those of Tag 2, and Tag 3 had the lowest values of maximum and mean RSSImx. However, the minimum values of RSSImx of each tag were very similar . 4)!Different tags did not have much difference in the angle " of Rmax . For all the tags, the angle " of Rmax were within the ranges of 0¡±30¡ and 180¡±30¡ (0¡ ) 30¡ is e qual to 330¡); 5)!Tag 1 had No -read Points in its read ranges with material s M1, M2 and M3 measured by both the LP and CP antenna, while Tag 2 had No -read Points with material s M1, M2 and M3 using the LP antenna. Tag 3 showed No -read Points only with M2 while using the LP antenna. Comparing Table 5 to Table 6, and Table 7 to Table 8, as well as Table 9 to Table 10, the effect of different reader antennas on the tagÕs performance in the read range test was: 1)!The tag tested with the LP antenna always had more Read Points, larger maximum RSSImx, larger mean RSSImx, as well as greater R max and greater ave rage R", compared to the results tested using the CP antenna ; 2)!The values in the TotCnt category and the minimum RSSImx were similar between measurements using the LP antenna and the CP antenna; 3)!CP antenna readings of the tag resulted in more " angles where the Rmax was obtained; 6)!All three tags read by the LP antenna had No -read Points within their read ranges, while only Tag 1 had No -read Points in the read ranges tested using the CP antenna. !58 The following figures are created to visually present the data in a detailed and straightforward manner. Figure 12 to Figure 17 show the read ranges of Tag 1 with each material tested by the LP antenna and CP antenna. Figure 18 to Figure 23 show the read ranges of Tag 2. Figure 24 to Figure 29 show the read ranges of Tag 3. No-read Points were marked as ! in the read range figures. The trends of TotCnt versus distance and RSSImx versus distance are also shown below. However, only t he test results at " =0¡ are provided in this document as representative examples. Figure 30 to Figure 35 show the TotCnt v ersus distance and RSSImx versus distance at " =0¡ with the same tag; Figure 36 to Figure 41 show the TotCnt or RSSImx versu s distance at " =0¡ with the same material; and Figure 42 to Figure 45 show the TotCnt or RSSImx versus distance at " =0¡ using the same reader anteena. Since there was no read point for M4 for any tags using either antenna , M4 results are all excluded in the figures . TotCnt versus RSSImx figures are provided in Appendix 3 to show how the results were clustered. !59 Figure 12 Read Range of Tag 1 with M1 Measured Using LP Antenna Figure 13 Read Range of Tag 1 with M2 Measured Using LP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 1 with M1, Read Range LP-T1-M1-Read Range (ft) No-read Point 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 1 with M2, Read Range LP-T1-M2-Read Range (ft) No-read Point !60 Figure 14 Read Range of Tag 1 with M3 Measured Using LP Antenna Figure 15 Read Range of Tag 1 with M1 Measured Using CP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 1 with M3, Read Range LP-T1-M3-Read Range (ft) No-read Point 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 1 with M1, Read Range CP-T1-M1-Read Range (ft) No-read Point !61 Figure 16 Read Range of Tag 1 with M2 Measured Using CP Antenna Figure 17 Read Range of Tag 1 with M3 Measured Using CP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 1 with M2, Read Range CP-T1-M2-Read Range (ft) No-read Point 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 1 with M3, Read Range CP-T1-M3-Read Range (ft) No-read Point !62 Figure 18 Read Range of Tag 2 with M1 Measured Using LP Antenna Figure 19 Read Range of Tag 2 with M2 Measured Using LP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 2 with M1, Read Range LP-T2-M1-Read Range (ft) No-read Point 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 2 with M2, Read Range LP-T2-M2-Read Range (ft) No-read Point !63 Figure 20 Read Range of Tag 2 with M3 Measured Using LP Antenna Figure 21 Read Range of Tag 2 with M1 Measured Using CP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 2 with M3, Read Range LP-T2-M3-Read Range (ft) No-read Point 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 2 with M1, Read Range CP-T2-M1-Read Range (ft) !64 Figure 22 Read Range of Tag 2 with M2 Measured Using CP Antenna Figure 23 Read Range of Tag 2 with M3 Measured Using CP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 2 with M2, Read Range CP-T2-M2-Read Range (ft) 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 2 with M3, Read Range CP-T2-M3-Read Range (ft) !65 Figure 24 Read Range of Tag 3 with M1 Measured Using LP Antenna Figure 25 Read Range of Tag 3 with M2 Measured Using LP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 3 with M1, Read Range LP-T3-M1-Read Range (ft) 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 3 with M2, Read Range LP-T3-M2-Read Range (ft) No-read Point !66 Figure 26 Read Range of Tag 3 with M3 Measured Using LP Antenna Figure 27 Read Range of Tag 3 with M1 Measured Using CP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡LP Antenna, Tag 3 with M3, Read Range LP-T3-M3-Read Range (ft) 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 3 with M1, Read Range CP-T3-M1-Read Range (ft) !67 Figure 28 Read Range of Tag 3 with M2 Measured Using CP Antenna Figure 29 Read Range of Tag 3 with M3 Measured Using CP Antenna 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 3 with M2, Read Range CP-T3-M2-Read Range (ft) 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0¡30¡60¡90¡120¡150¡180¡210¡240¡270¡300¡330¡CP Antenna, Tag 3 with M3, Read Range CP-T3-M3-Read Range (ft) !68 Figure 30 Tag 1 with M1, M2, and M3, TotCnt versus Distance when " =0¡ Figure 31 Tag 1 with M1, M2, and M3, RSSImx versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) Tag 1 with M1, M2, and M3, % =0 ¡, TotCnt vs Distance r LP-T1-M1-TotCnt LP-T1-M2-TotCnt LP-T1-M3-TotCnt CP-T1-M1-TotCnt CP-T1-M2-TotCnt CP-T1-M3-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) Tag 1 with M1, M2, and M3, %=0 ¡, RSSImx vs Distance r LP-T1-M1-RSSImx LP-T1-M2-RSSImx LP-T1-M3-RSSImx CP-T1-M1-RSSImx CP-T1-M2-RSSImx CP-T1-M3-RSSImx !69 Figure 32 Tag 2 with M1, M2, and M3, TotCnt versus Distance when " =0¡ Figure 33 Tag 2 with M1, M2, and M3, RSSImx versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) Tag 2 with M1, M2, and M3, % =0 ¡, TotCnt vs Distance r LP-T2-M1-TotCnt LP-T2-M2-TotCnt LP-T2-M3-TotCnt CP-T2-M1-TotCnt CP-T2-M2-TotCnt CP-T2-M3-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) Tag 2 with M1, M2, and M3, % =0 ¡, RSSImx vs Distance r LP-T2-M1-RSSImx LP-T2-M2-RSSImx LP-T2-M3-RSSImx CP-T2-M1-RSSImx CP-T2-M2-RSSImx CP-T2-M3-RSSImx !70 Figure 34 Tag 3 with M1, M2, and M3, TotCnt versus Distance when " =0¡ Figure 35 Tag 3 with M1, M2, and M3, RSSImx versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) Tag 3 with M1, M2, and M3, %=0 ¡, TotCnt vs Distance r LP-T3-M1-TotCnt LP-T3-M2-TotCnt LP-T3-M3-TotCnt CP-T3-M1-TotCnt CP-T3-M2-TotCnt CP-T3-M3-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) Tag 3 with M1, M2, and M3, %=0 ¡, RSSImx vs Distance r LP-T3-M1-RSSImx LP-T3-M2-RSSImx LP-T3-M3-RSSImx CP-T3-M1-RSSImx CP-T3-M2-RSSImx CP-T3-M3-RSSImx !71 Figure 36 Each Tag with M1, TotCnt versus Distance when " =0¡ Figure 37 Each Tag with M1, RSSImx versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) Each tag with M1, % =0 ¡, TotCnt vs Distance r LP-T1-M1-TotCnt LP-T2-M1-TotCnt LP-T3-M1-TotCnt CP-T1-M1-TotCnt CP-T2-M1-TotCnt CP-T3-M1-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) Each tag with M1, % =0 ¡, RSSImx vs Distance r LP-T1-M1-RSSImx LP-T2-M1-RSSImx LP-T3-M1-RSSImx CP-T1-M1-RSSImx CP-T2-M1-RSSImx CP-T3-M1-RSSImx !72 Figure 38 Each Tag with M2, TotCnt versus Distance when " =0¡ Figure 39 Each Tag with M2, RSSImx versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) Each tag with M2, % =0 ¡, TotCnt vs Distance r LP-T1-M2-TotCnt LP-T2-M2-TotCnt LP-T3-M2-TotCnt CP-T1-M2-TotCnt CP-T2-M2-TotCnt CP-T3-M2-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) Each tag with M2, % =0 ¡, RSSImx vs Distance r LP-T1-M2-RSSImx LP-T2-M2-RSSImx LP-T3-M2-RSSImx CP-T1-M2-RSSImx CP-T2-M2-RSSImx CP-T3-M2-RSSImx !73 Figure 40 Each Tag with M3, TotCnt versus Distance when " =0¡ Figure 41 Each Tag with M3, RSSImx versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) Each tag with M3, % =0 ¡, TotCnt vs Distance r LP-T1-M3-TotCnt LP-T2-M3-TotCnt LP-T3-M3-TotCnt CP-T1-M3-TotCnt CP-T2-M3-TotCnt CP-T3-M3-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) Each tag with M3, % =0 ¡, RSSImx vs Distance r LP-T1-M3-RSSImx LP-T2-M3-RSSImx LP-T3-M3-RSSImx CP-T1-M3-RSSImx CP-T2-M3-RSSImx CP-T3-M3-RSSImx !74 Figure 42 LP Antenna, Each Tag with M1, M2, and M3, TotCnt versus Distance when " =0¡ Figure 43 LP Antenna, Each Tag with M1, M2, and M3, RSSImx versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) LP Antenna, each tag with M1, M2, and M3, %=0 ¡, TotCnt vs Distance r LP-T1-M1-TotCnt LP-T1-M2-TotCnt LP-T1-M3-TotCnt LP-T2-M1-TotCnt LP-T2-M2-TotCnt LP-T2-M3-TotCnt LP-T3-M1-TotCnt LP-T3-M2-TotCnt LP-T3-M3-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) LP Antenna, each tag with M1, M2, and M3, %=0 ¡, RSSImx vs Distance r LP-T1-M1-RSSImx LP-T1-M2-RSSImx LP-T1-M3-RSSImx LP-T2-M1-RSSImx LP-T2-M2-RSSImx LP-T2-M3-RSSImx LP-T3-M1-RSSImx LP-T3-M2-RSSImx LP-T3-M3-RSSImx !75 Figure 44 CP Antenna, Each Tag with M1, M2, and M3, TotCnt versus Distance when " =0¡ Figure 45 CP Antenna, Each Tag with M1, M 2, and M3, TotCnt versus Distance when " =0¡ 0501001502002503003504004500.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Count of Reads Distance r (ft) CP Antenna, each tag with M1, M2, and M3, %=0 ¡, TotCnt vs Distance r CP-T1-M1-TotCnt CP-T1-M2-TotCnt CP-T1-M3-TotCnt CP-T2-M1-TotCnt CP-T2-M2-TotCnt CP-T2-M3-TotCnt CP-T3-M1-TotCnt CP-T3-M2-TotCnt CP-T3-M3-TotCnt -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 RSSImx (dBm) Distance r (ft) CP Antenna, each tag with M1, M2, and M3, %=0 ¡, RSSImx vs Distance r CP-T1-M1-RSSImx CP-T1-M2-RSSImx CP-T1-M3-RSSImx CP-T2-M1-RSSImx CP-T2-M2-RSSImx CP-T2-M3-RSSImx CP-T3-M1-RSSImx CP-T3-M2-RSSImx CP-T3-M3-RSSImx !76 4.2 Results of the Orientation Read Rate Test In the orientation read rate test, the reader read the tag while the tag was located at the distance r=2 feet, and rotated to 60 orientations characterized by !-"-# . The 60 sets of !-"-# were randomly generated, where ! , " , # were integer multiples of 15¡, and 0¡ +! <180¡, 0¡ +" <360¡, 0¡ +#+ 360¡. The measurement at each orientation of !-"-# consisted of three trials, 5 seconds run for e ach trial. The reader recorded the TotCnt and RSSImx of each trial. Then, the integer of the average TotCnt of three trials, and the value of the maximum RSSImx of three trials were calculated and used to represent the TotCnt and RSSImx value of this measu rement. The following terms were defined: A Read Orientation was defined as a measurement of three trials at the orientation !-"-# that had an average TotCnt value of at least 1 . A No-read Orientation was defined as a measurement that had an average TotCnt less than 1. The No-read Orientations did not have any RSSImx value, since the reader could not detect any signal from the tag . Table 11 to Table 16 are summaries of the orientation read rate test of the three tags with the four packaging materials using the LP and CP antenna. These tables first provide the count of Read Orientations for a tag with each material, then the maximum, minimum, mean, deviation values of the TotCnt and RSSImx was calculated from all the Read Orientations of the tag with a material. The TotCnt and RSSImx of the measurement at ! =90¡, " =0¡, # =0 and r=2 feet collected from the read range test are provided at the end of each table to serve as a reference. The orientation of 90¡ -0¡-0¡ is the preferred orientation of each tag. !77 Table 11 Summary of the Orientation Read Rate Test of Tag 1 Using LP Antenna LP Antenna, Tag 1 Material M1 M2 M3 M4 Count of Read Orientations 48 48 48 0 TotCnt of Read Orientations Max 430 429 429 Not Applicable Min 138 22 333 Mean 407 401 422 Deviation 53.76 80.33 13.93 RSSImx of Read Orientations Max -39 -42 -38 Min -59 -63 -58 Mean -45 -48 -44 Deviation 6 6 6 Orientation of 90¡-0¡-0¡ TotCnt 429 412 426 0 RSSImx -38 -42 -38 Not Applicable Table 12 Summary of the Orientation Read Rate Test of Tag 1 Using CP Antenna CP Antenna, Tag 1 Material M1 M2 M3 M4 Count of Read Orientations 50 47 48 0 TotCnt of Read Orientations Max 430 430 429 Not Applicable Min 4 2 34 Mean 371 392 390 Deviation 126.96 88.55 93.79 RSSImx of Read Orientations Max -44 -46 -44 Min -63 -64 -61 Mean -51 -54 -51 Deviation 5 5 5 Orientation of 90¡-0¡-0¡ TotCnt 428 428 429 0 RSSImx -45 -48 -44 Not Applicable Table 13 Summary of the Orientation Read Rate Test of Tag 2 Using LP Antenna LP Antenna Tag 2 Material M1 M2 M3 M4 Count of Read Orientations 45 45 45 0 !!78 Table 13 (contÕd) LP Antenna Tag 2 Material M1 M2 M3 M4 TotCnt of Read Orientations Max 430 429 430 Not Applicable Min 1 47 1 Mean 361 367 368 Deviation 134.88 120.84 121.36 RSSImx of Read Orientations Max -45 -47 -44 Min -63 -64 -62 Mean -51 -54 -51 Deviation 5 5 5 Orientation of 90¡-0¡-0¡ TotCnt 428 431 430 0 RSSImx -44 -46 -44 Not Applicable Table 14 Summary of the Orientation Read Rate Test of Tag 2 Using CP Antenna CP Antenna, Tag 2 Material M1 M2 M3 M4 Count of Read Orientations 43 44 45 0 TotCnt of Read Orientations Max 430 430 430 Not Applicable Min 7 2 14 Mean 348 335 336 Deviation 137.43 142.64 143.91 RSSImx of Read Orientations Max -49 -51 -49 Min -64 -67 -64 Mean -56 -58 -56 Deviation 4 4 4 Orientation of 90¡-0¡-0¡ TotCnt 429 429 427 0 RSSImx -50 -52 -50 Not Applicable Table 15 Summary of the Orientation Read Rate Test of Tag 3 Using LP Antenna LP Antenna Tag 3 Material M1 M2 M3 M4 Count of Read Orientations 35 36 36 0 TotCnt of Read Orientations Max 428 428 428 Not Applicable Min 45 47 4 !79 Table 15 (contÕd) LP Antenna Tag 3 Material M1 M2 M3 M4 TotCnt of Read Orientations Mean 405 400 393 Deviation 81.20 80.83 107.43 RSSImx of Read Orientations Max -48 -50 -49 Min -61 -63 -62 Mean -53 -55 -52 Deviation 3 3 3 Orientation of 90¡-0¡-0¡ TotCnt 428 427 428 0 RSSImx -49 -51 -49 Not Applicable Table 16 Summary of the Orientation Read Rate Test of Tag 3 Using CP Antenna CP Antenna, Tag 3 Material M1 M2 M3 M4 Count of Read Orientations 31 32 32 0 TotCnt of Read Orientations Max 428 429 428 Not Applicable Min 7 3 10 Mean 345 339 341 Deviation 137.94 148.61 147.88 RSSImx of Read Orientations Max -53 -55 -53 Min -63 -65 -63 Mean -59 -60 -59 Deviation 3 3 3 Orientation of 90¡-0¡-0¡ TotCnt 428 426 426 0 RSSImx -54 -56 -54 Not Applicable Comparing within each table to evaluate the effect of materi als on the tagÕs performance in the orientation read rate test, the observat ions include : 1)!There was no Read Orientation for M4 (PET/Al/polyethylene film) for any tags using either antenna ; !80 2)!Material M1, M2 and M3 resulted in similar count s of Read Orientations. The counts were either identical, or no more than 3 counts in difference; 3)!Material M1, M2 and M3 had little difference on the maximum and mean values of TotCnt. The minimum TotCnt with M1, M2 and M3 showed difference in Table 11, Table 12, Table 13, and Table 15, but the difference was not consistent; 4)!Material M2 showed smaller maximum, minimum and mean values of RSSImx than those of M1 and M3 in each table. The differences were within -5dBm; Comparing between Table 11, Table 13, and Table 15, or comparing betwe en Table 12, Table 14, and Table 16, tags with different antenna designs showed the following in the orient ation read rate test : 1)!Tag 1 had more Read Orientations than Tag 2, and Tag 3 had the least counts of Read Orientations; 2)!All three tags had very similar value s of the maximum TotCnt. Tag 1 had the greater mean values of TotCnt than the other two; 5)!The maximu m and mean values of RSSImx of Tag 1 were greater than those of Tag 2, and Tag 3 had the lowest values of the maximum and mean RSSImx. The minimum values of RSSImx of each tag were very similar . Comparing Table 11 to Table 12, Table 13 to Table 14, and Table 15 to Table 16, the effect of the LP and CP antenna on the tagÕs performance in the orientation read rate test are shown below: 1)!Tag 1 and Tag 2 showed very similar values of the count of Read Orientations between !81 measurements using the LP antenna and CP antenna. Tag 3 had the more Read Orientations while using the LP antenna; 2)!Polarizations of the antennas did not have effect on the maximum TotCnt value in the tables. However, in general, the measurements with the LP antenna had greater v alues of mean TotCnt. The effect of antennas on the minimum TotCnt was not consistent across different tags; 3)!Measurements with the LP antenna resulted in larger values in the RSSImx categories than those with the CP antenna ; The following figures are creat ed to visually present the data. Figure 46 to Figure 51 show the O -RDRate of each tag with each material tested by the LP antenna and CP antenna. Figures of TotCnt versus RSSImx of each tag are provided in Appendix 4 to show how the results were clustered. M4 are excluded in the figures of TotCnt versus RSSImx in the appendix, as there was no Read -Orientation for M4 for any tags using either antenna. !82 Figure 46 Orientation Read Rate of Tag 1 with Four Materials Tested Using LP Antenna Figure 47 Orientation Read Rate of Tag 1 with Four Materials Tested Using CP Antenna 80.00% 80.00% 80.00% 0.00% M1M2M3M4LP Antenna, Tag 1, Orientation Read Rate with Packaging Materials O-RDRate 83.33% 78.33% 80.00% 0.00% M1M2M3M4CP Antenna, Tag 1, Orientation Read Rate with Packaging Materials O-RDRate !83 Figure 48 Orientation Read Rate of Tag 2 with Four Materials Tested Using LP Antenna Figure 49 Orientation Read Rate of Tag 2 with Four Materials Tested Using CP Antenna 75.00% 75.00% 75.00% 0.00% M1M2M3M4LP Antenna, Tag 2, Orientation Read Rate with Packaging Materials O-RDRate 71.67% 73.33% 75.00% 0.00% M1M2M3M4CP Antenna, Tag 2, Orientation Read Rate with Packaging Materials O-RDRate !84 Figure 50 Orientation Read Rate of Tag 3 with Four Materials Tested Using LP Antenna Figure 51 Orientation Read Rate of Tag 3 with Four Materials Tested Using CP Antenna 58.33% 60.00% 60.00% 0.00% M1M2M3M4LP Antenna, Tag 3, Orientation Read Rate with Packaging Materials O-RDRate 51.67% 53.33% 53.33% 0.00% M1M2M3M4CP Antenna, Tag 3, Orientation Read Rate with Packaging Materials O-RDRate !85 Chapter 5 Analysis and Discussion 5.1 Overview The results of the experiments showed that M1, M2, and M3 had very little effect on the read range and O -RDRate of the tag, while with M4 the tag was not read by the reader in either the read range test or the orientation read rate test. The hypothesis of this research that were established in Chapter 1 were: ¥!Hypothesis 1: Packaging materials will NOT have a consistent effect on read range across different antenna designs of general -purpose dipole antenna transponders. ¥!Hypothesis 2: Packaging materials will NOT have a consistent effect on orientation read rate across different antenna designs of general -purpose dipole antenna transponders. Observation of the experiment results indicated that both the two hy potheses were false, in other words, packaging mater ials have a consistent effect on the read range and O-RDRate of general -purpose dipole antenna transponders. To support the observation, a stat istical model for each sub -test was created using SAS with the assistance of the MSU College of Agriculture and Natural Resources (CANR) Statistical Consulting Center (SCC). For the read range test, a survival analysis model was employed as a novel method to analyze the effects on the tagÕs read range. Survival analysis models are usually used to evaluate factors tha t influence the time to an event, for example, to examine influence of factors such as age, gender , and body mass index on survival time after heart attack [69] . In the read range test, the survival analysis model was used to examine !86 factors in cluding mater ial, antenna, and tag affects on the read range, i.e. the R ". For the orientation read rate test, a completely randomized design model was used to a nalyze the effects of material, antenna, and tag on the TotCnt in order to support the conclusio n of Hypothesis 2 . 5.2 Analysis of the Read Range Test In the survival analysis model for the read range test, material s M1, M2 , and M3 were used . M4 was excluded , since it did not have any Òsurvival timeÓ. First, assessment of whether different categories of variables have different survival functions was conducted. The survival function was originally interpreted as the probability of surviving past time t [69] . In this research, the survival function was described as the probability of read ing a tag past distance r.Table 17, Table 18, and Table 19 shows the test of equality over factors of material (M1, M2 and M3 ), an tenna (LP and CP antenna), and tag (Tag 1, Tag 2 and Tag 3) , respectively. In the survival analysis mod el, P value s smaller than 0.01 are considered as significant. Table 17 shows that P values of the Log -Rank, Wilcoxon and -2Log Chi -Squire tests were all greater than 0.01, which means that survival functions of M1, M2 and M3 showed no evidence of difference . In other words, material M1, M2 and M3 did not have significant effect on the tagÕs read range. Table 18 has P values all smaller than 0.01, suggest ing that the effect of reader antenna polarization on the tagÕs read range was significant. Similar to Table 18, P values in Table 19 are all smaller than 0.01, thus indicate that different tags have significantly different read ranges. !87 Table 17 Test of Equality over Material (M1, M2 and M3) Test of Equality over Material Test Chi -Square DF Pr > Chi -Square Log -Rank 0.4634 2 0.7932 Wilcoxon 0.2088 2 0.9009 -2Log(LR) 0.0766 2 0.9624 Table 18 Test of Equality over Antenna Test of Equality over Antenna Test Chi -Square DF Pr > Chi -Square Log -Rank 25.4413 1 <.0001 Wilcoxon 16.6251 1 <.0001 -2Log(LR) 13.9277 1 0.0002 Table 19 Test of Equality over Tag Test of Equality over Tag Test Chi -Square DF Pr > Chi -Square Log -Rank 71.9002 2 <.0001 Wilcoxon 49.5087 2 <.0001 -2Log(LR) 26.3705 2 <.0001 Second , a Cox proportional hazards regression was used to conf irm the above results. Table 20 shows the analysis of maximum likelihood estimates of the Cox proportional hazards regression model. The hazard ratio in the table indic ates the risk that the tag cannot be read at distance r. Table 20 shows that Tag 1 performed better than the reference tag (Tag 3) with a hazard ratio of 0.177. So, Tag 1 had a Ndu|--}T-eN~%y&| decrease in the risk of not being read by the reader compared to Tag 3. And the difference between Tag 1 and Tag 3 was significant with P<0.0001. Tag 2 also performed better than Tag 3. The hazard ratio of Tag 2 was 0.334 , and this !88 indicate d that Tag 2 had a Ndu|{{6T-eN~%aa| decrease in the risk of not being read by the reader compared to Tag 3: t he difference between Tag 2 and Tag 3 was significant as well, with a P value less than 0.0001. The CP antenna performed a lot worse than the LP antenna with the hazard ratio of 2.420. The CP antenna had Nd&|6&uT-eN~%-6&|u more risk of failing to detect a tag at distance r. The difference between the CP antenna and the LP antenna was significant with P <0.0001. Finally, M2 showed a ha zard ratio of 0.902 and M3 showed a hazard ratio of 0.863 compared to the reference material (M1) . However, the difference between M2 and M1, as w ell as that between M3 and M1 was not significant with the p=0.5385 and p=0.3818, respectively. Table 20 Result of the Cox Proportional Hazard Regression of the Survival Analysis Analysis of Maximum Likelihood Estimates Parameter DF Parameter Estimate Standard Error Chi -Square Pr > ChiSq Hazard Ratio Label Tag 1 1 -1.73423 0.21262 66.5269 <.0001 0.177 Tag 1 Tag 2 1 -1.09569 0.18885 33.6621 <.0001 0.334 Tag 2 Antenna CP 1 0.88357 0.15442 32.7392 <.0001 2.420 Antenna CP Material M2 1 -0.10274 0.16704 0.3783 0.5385 0.902 Material M2 Material M3 1 -0.14734 0.16846 0.7650 0.3818 0.863 Material M3 Hence, the Hypothe sis 1: Packaging materials will NOT have a consistent effect on read range across different antenna designs of general -purpose dipole antenna transponders was determined to be false. For all three general -purpose dipole antenna tags tested in this research, the read range of one tag was significantly differen t from that of the others , except the read ranges !89 tested wit h M4. The tagÕs read range with M1, M2 and M3 tested using the LP antenna was significantly larger than that of using the CP antenna. However, M1, M2 and M3 did not have significant effect on the tagÕs read range, while M4 always resulted in null read range for all three tags. The effect of packaging materials was consistent on read range across different antenna designs of general -purpose dipole antenna transponders. 5.3 Analysis of the Orientatio n Read Rate Test The O -RDRate was calculated as the count of Read Orientations over all 60 orientations (Table 4). Figure 46 to Figure 51 indicated that material M1, M2 and M3 had very little effect on the O -RDRate of a tag, whil e M4 resulted in 0% O -RDRate of all three tags. The data set of the orientation read rate test had two responses: TotCnt and RSSImx, while RSSImx did not have any value for the No -read Orientations . In order to further examine the effects of packaging materials on the tagÕs performance in the orientation read rate test , a completely randomized design analysis model was conducted and the response TotCnt was used to investigate how packaging materials influence the tagÕs O -RDRate . First, all the TotCnt data collected using the two antennas , three tags with four packaging materials was imported in the SAS program. The assumptions of normality and equal variance of the response TotCnt were checked. Figure 52 shows the results of SAS program and indicates that the TotCnt data was not quite normal. !90 Figure 52 Residuals for TotCnt in the Orientation Read Rate Test Then two models: one model with equal variance and the other one with unequal variance were run. The Akaike In formation Criterion (AIC) values of the two models were compared to decide the better model for analyzing the TotCnt data. Table 21 and Table 22 show the AIC value of the equal variance model and unequal variance model, respectively. The model with unequal variance had slightly smaller AIC, hence this model was used for further analysis. >þ?˜?!?˘?˜>Ì?˙?˚>Ì?˙? >ï?ˆ? >î>õ>ï>á>Þ>å>Ü>å>í>õ>ï>ï>á>Þ>å>Ü>á>í>õ>ï>á>Þ>å>Ü>á>û? ?">á>Þ>å>Ü>Ý>ò? >Ì>ÿ? ? ?˜? ?˜>ÿ? >Ì>ð?">Ý>â>á>Ú>å>á>ù?$?ˇ?!?ˇ>Þ>á>Þ>Ú>ä>Þ>ù?ˆ>ß>ß>ñ>Ù>Ý>ß>ù?ˆ?ˇ?!?ˇ>Ù>ß>ß>ã>Ú>ä>û?˜?˚?"? ?˙?ˆ?˜>à>ß>Þ>Ü>þ?˜?!?˘>Ì>ÿ? ? ?˜? ?˜>Ù>à>Ù>Þ>Ü>Þ>à>ý?!?ˆ? ?˘>Ù>á>Ü>Ü>Ù>Þ>á>Ü>Ü>Þ>á>Ü>á>Ü>Ü>þ?˜?!?˘>Ù>à>ä>ã>Ú>á>Ù>ß>Ý>Þ>Ú>á>Ù>Ý>ß>ã>Ú>á>ß>ã>Ú>á>Þ>Ý>Þ>Ú>á>ß>ä>ã>Ú>á>þ?˜?!?˘>Ü>á>Ý>Ü>Ý>á>Þ>Ü>Þ>á>ü?˚?ˆ? >Ü>Ý>Ü>Ü>Þ>Ü>Ü>ß>Ü>Ü>ü?˚? >Ì>ù?ˆ>Ù>Þ>Ü>Ü>Ü>Þ>Ü>Ü>þ?˜?!?˘!91 Table 21 Fit Statistics of the Model with Equal Variance Fit Statistics -2 Res Log Likelihood 52901.2 AIC (Smaller is Better) 52905.2 AICC (Smaller is Better) 52905.2 BIC (Smaller is Better) 52909.4 Table 22 Fit Statistics of the Model with Unequal Variance Fit Statistics -2 Res Log Likelihood 52730.1 AIC (Smaller is Better) 52740.1 AICC (Smaller is Better) 52740.1 BIC (Smaller is Better) 52750.6 With the unequal variance model, results of the analysis are presented below. P value s smaller than 0.01 are considered as indicating the variable had significant effect on the response of TotCnt. Table 23 shows the effect of different variables on the value of TotCnt in the orientation read rate test. The P values in Table 23 indicate that antenna, material and tag all had significant effects on the TotCnt value. Antenna by material, antenna by tag, and tag by material had significant interaction effects on the TotCnt. But antenna, material and tag together did not have a significant effect on the TotCnt in the orientation read rate test. Table 23 Effect of Antenna, Tag and Material on the TotCnt in the Orientati on Read Rate Test Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F Antenna 1 3330 55.55 <.0001 Material 3 2123 1009.66 <.0001 Antenna*Material 3 2123 4.13 0.0062 Tag 2 3330 199.00 <.0001 Antenna*Tag 2 3330 7.06 0.0009 Material*Tag 6 2603 14.87 <.0001 Antenna*Material*Tag 6 2603 0.73 0.6260 !92 Table 24 compares the least square means of the TotCnt values collected using the LP and CP antenna. This table again indicates that antenna had a statistically significant effect on the tagÕs TotCnt in the orientation read rate test. More specifically, the measurements with the LP antenna had an average of 25.14 more TotCnt per trial than those with the CP antenna, and this difference was significant. Table 24 Least Squares Mea ns of TotCnt Tested with the LP and CP Antennas Effect of Antenna on TotCnt Antenna Estimate Standard Error DF t Value Pr > |t| Least Squares Means CP 184.83 19.2436 59.6 9.60 <.0001 LP 209.97 19.2436 59.6 10.91 <.0001 Differences of Least Squares Means -25.1444 3.3737 3330 -7.45 <.0001 Table 25 shows the least square means of the TotCnt values with different packaging materials. The first four cells of the Estimate column show the average TotCnt of a trial tested with the correspondent packaging material. In the Differences of Least Squares Means category, M1, M2 and M3 had small values of Estimate when compared with each other, and these differences were not statistically significant as the correspondent P values were all larger than 0.01. M4 had a significant difference from the other three mate rials with P values in row 7, row 9 and row 10 are all smaller. !93 Table 25 Least Squares Means of TotCnt with Different Packaging Materials Effect of Material on TotCnt Material Estimate Standard Error DF t Value Pr > |t| Least Squares Means M1 261.72 19.3163 60.6 13.55 <.0001 M2 261.81 19.3070 60.4 13.56 <.0001 M3 266.08 19.3262 60.7 13.77 <.0001 M4 1.324E -9 19.6125 64.2 0.00 1.0000 Differences of Least Squares Means M2-M1 0.09352 4.0775 1917 0.02 0.9817 M2-M3 -4.2667 4.1243 1950 -1.03 0.3010 M2-M4 261.81 5.3064 1482 49.34 <.0001 M3-M1 4.3602 4.1677 1960 1.05 0.2956 M4-M1 -261.72 5.3402 1503 -49.01 <.0001 M4-M3 -266.08 5.3760 1511 -49.49 <.0001 Table 26 shows the least square means of TotCnt with three tags. Tag 1 averaged 42.73 and 82.41 more TotCnt per trial than Tag 2 and Tag 3, respectively. Tag 2 averaged 39.68 more TotCnt per trial than Tag 3. Differences between tags were statistically significant. Table 26 Least Squares Means of TotCnt with Different Tags Effect of Tag on TotCnt Tag Estimate Standard Error DF t Value Pr > |t| Least Squares Means Tag 1 239.12 19.3174 60.5 12.38 <.0001 Tag 2 196.38 19.3174 60.5 10.17 <.0001 Tag 3 156.70 19.3174 60.5 8.11 <.0001 Differences of Least Squares Means Tag 1 ÐTag 2 42.7313 4.1319 3330 10.34 <.0001 Tag 1 Ð Tag 3 82.4139 4.1319 3330 19.95 <.0001 Tag 2 Ð Tag3 39.6826 4.1319 3330 9.60 <.0001 Table 27 presents the analysis output of the effect of the interaction between the antenna and tag on the TotCnt. The Difference of Least Squares Means category in Table 27 shows that even though using the LP antenna to test Tag 1 resulted in a greater average TotCnt than using the CP !94 antenna, the effect of antenna on Tag 1Õs TotCnt was not significant as the P value was 0.0173. However, since 0.0173 is close t o 0.1, it needs future reseach to further evaluate the effect of antenna on Tag 1Õs TotCnt. For Tag 2 and Tag 3, the effect of antenna was significant. Table 27 Least Squares Means of TotCnt with Antenna by Tag Effect of Antenna*Tag on TotCnt Antenna*Tag Estimate Standard Error DF t Value Pr > |t| Least Squares Means CP*Tag 1 232.16 19.5371 63.3 11.88 <.0001 LP*Tag 1 246.07 19.5371 63.3 12.60 <.0001 CP*Tag 2 187.06 19.5371 63.3 9.57 <.0001 LP*Tag 2 205.71 19.5371 63.3 10.53 <.0001 CP*Tag 3 135.27 19.5371 63.3 6.92 <.0001 LP*Tag 3 178.13 19.5371 63.3 9.12 <.0001 Differences of Least Squares Means CP*Tag 1 Ð LP*Tag 1 -13.9153 5.8434 3330 -2.38 0.0173 CP*Tag 2 Ð LP*Tag 2 -18.6583 5.8434 3330 -3.19 0.0014 CP*Tag 3 Ð LP*Tag 3 -42.8597 5.8434 3330 -7.33 <.0001 To summarize, the analysis of the orientation read rate test showed that Hypothesis 2 : Packaging materials will NOT have a consistent effect on orientation read rate across different antenna designs of general -purpose dipole antenna transponders was false. For all three tags tested in this research, material M1, M2 and M3 did not have significant effect on the tagÕs O -RDRate, and M4 always led to a zero percent O -RDrate. Packaging materials had a consistent effect on O-RDRate across different antenna designs of general -purpose dipole antenna transponders. Three tags had significantly different O -RDRates. For the O -RDRate of Tag 2 or Tag 3, the effect of antenna polarization was significant, and the tag read by the LP antenna had the higher O -RDRate and the more TotCnt per trial. However , antenna polarization might not have a significant effect on the O -RDRate of Tag 1 (P=0.0173) . T he effect of antenna polarization on O-!95 RDRate might not consistent across different antenna designs of general -purp ose dipole antenna transponders, and needs further study to investigate. To apply the results to the real -world application of RFID system s, on e can expect the consiste nt effect of different pa ckaging material on general -purpose passive dipole -antenna transponder performance. In other words, generally speaking, a transponder that performs better than another while attached on a given materi al will still perform better on another material, and general -purpose passive diopole antenna transponder will not yield any read while using with materials that contains madel. As the tagÕs antenna designs had a significant effect on tranponder read range and orientation read rate, it is suggested to perform t he read range test and orientation read rate test described in this research with all the tags on a packaging material to determine the tagÕs performance, then choose the desired tag or tags for the proposed application. According to the results of this re search, one can expect that a tag with larger read range generally will have larger orientation read rate. Furthermore, reader antennaÕs polarization is a significant effect on the tagÕs rea d range and orientation read rate (except for Tag1Õs orientation r ead rate), and generally the linear antenna will result in larger read range and larger orientation read rate. However, linear antenna may have blind spots around 10 feet due to the Gen 2 protocol [53] ( e.g. Figure 12), and circular antenna may have blind spots around 7 feet (e.g. Figure 15). Thus, i t is suggested to perform tests with both antenna at planned location to avoid blind spots before complete installation of an RFID system. !96 Chapter 6 Conclusion s and Future Research In this thesis, three general -purpose dipole antenna passive RFID tags were evaluated when they were attached to four packaging material samples. Tag performance of read range, which is the maximum distance the tag can be detected by the reader, and orient ation read rate, which is the percentage of orientations in which the transponder can be read, out of all the possible orientations evaluated within a three -dimensional space were quantified and analyzed. Four commonly used packaging material s were chosen, including uncoated corrugated paperboard, PE film, PE corrugated board, and PET/Al/PE laminated film to investigate the effects of packaging materials on tag performance. A transponder support that could rotate, tilt, and incline the tag with the packagin g material sample was made for this research. Both the LP and CP antenna were used in the test. A survival analysis model that was a novel means to analyze read range test data was conducted. Results showed that all tags tested with the PET/Al/PE film coul d not be read by the reader, while the other three packaging materials yield no evidence of significant effect s on the tagÕs read range. Thus, it was concluded that packaging materials have a consistent effect on read range across different antenna designs of general -purpose dipole antenna transponders. Additionally , reader antenna polarization and tag antenna design both had significant effects on the tagÕs read range. Except for the measurements with the PET/Al/PE film, the tag had a larger !97 read range whe n tested using the LP antenna, and the tag with the larger dimension (see tag dimensions in Table 3 and Figure 7) had the larger read range . The orientation read rate test was analyzed u sing a completely randomized design analysis model. Results showed that all tags tested with the PET/Al/PE film had 0% O -RDRate, while the other three packaging material yield no evidence of a significant effect on the tagÕs O -RDRate, leading to the conclu sion that most packaging materials have consistent effect on O -RDRate across different antenna designs of general -purpose dipole antenna tags. In addition, it was found that except for the measurements with the PET/Al/PE film, the tagÕs antenna design had a significant effect on the O -RDRate, with the l arger antenna dimension resulting in the larger O -RDRate value. It was expected before the orientation read rate test that the tag measured by the CP antenna would have a higher O -RDRate than that of by the L P antenna. However, the results showed that the O -RDRate of a tag measured by the CP antenna was equal or smaller than the O -RDRate measured by the LP antenna. For Tag 1, antenna polarization did not have a significant effect on its O -RDRate. But for Tag 2 or Tag 3, the O -RDRate measured by the LP antenna was significantly higher than that measured by the CP antenna. So reader antenna polarizations did not have a consistent effect on O -RDRate across tags evaluated . This study evaluated three dipole antenna passive tags with four packaging materials in a simulated manufacturing environment. Several ideas were spun off during this study, but were not conducted due to the limitations of time and resource. For future work beyond this study, it is suggested that more transponder designs and packaging materials be evaluated. This research was !98 unable to clearly determine how a specific 3D orientation influences the tagÕs readability. Understanding the internal relationship between the tagÕs 3D orientation and tag pe rformance would be the next step in continuing this research. In addition, the orientation read rate test in this research was conducted at a fixed distance of 2 feet from the reader antenna. For future research, the interaction effect of distance and 3D o rientation on tag performance can be tested. Last but not least, this research was conducte d in a room simulating the real -world environment and could not control the relative humidity. Thus, another testi ng idea for future study is to conduct the read range test and orientation read rate test for tags with packaging materials under different environment al relative humidities . !99 APPENDICES !100 Appendix 1 Pilot Test 1 The initial purpose of pilot test 1 was to develop a testing methodology to evaluate a UHF passive transponderÕs three -dimensional (3D) read range in a simulated manufacturing environment, and to find a method with the least possible measurements to accurately map a passive tag Õs 3D read range. The tag used in pilot test 1 was AD -381m5, which was Tag 3 in the thesis. In pilot test 1, the tag was attached to the center of 7* 7 inch square high -density fiberboard (HDF) ( Figu re 53), and then was mounted on the tran sponder support described in Chapter 3. The RFID reader and antennas, the laboratory environment and setups were the same asin the tests in Chapter 3. Figu re 53 Tag AD -338m5 and the High -Density Fiberboard !101 In pilot test 1, the tagÕs critical distance and acquisition d istance were both measured in 3D space. Thus the tag had two 3D read rang es. O ne was based on the critical distance and the other one was based on the acquisition distance. The procedures for mapping the tagÕs critical 3D read range are described as below: 1)!Assemble the LP antenna to the reader, and configure the reader to a transmitting power of 20 dBm, and a receiving sensitivity of -70 dBm; 2)!Attach the tag AD -381m5 to the center of the HDF with # =0¡ using small pieces of masking tape, the angle # is kept constant throughout the entire pilot test 1; 3)!On the transponder support, rotate the HDF with the tag to ! =0¡ and " =0¡ , then place the transponder support at the distance r=0.5 foot; 4)!Turn on the reader to automatically read the tag 3 times, 5 seconds each time. Delay between two runs is 5 seconds; 5)!Save the data file with the information of reader antenna polarization, ang le ! and " , and distance r; 6)!Move the transponder support to r=1 foot and repeat step s 4 and 5; 7)!Move the transponder support outwards from the antenna with a n increment of 0.5 fo ot each time, until the tag can not be detected by the reader for 5 concutive in creases in distance ; 8)!Record the last distance r the tag can be read as R 0-0-cr. The R 0-0-cr is the tagÕs critical distance at ! =0¡ and " =0¡; !102 9)!Move the transponder support back to r=0.5 foot, turn the angle " to 15¡ and repeat step s 4 to step 7, record the R 0-15-cr; 10)!Increase the angle " by 15¡ each time and repeat step 4 to step 7 until " =345¡; Record correspond ing R!-"-cr at each " ; 11)!Rotate the angle ! to 15¡ and " to 0¡, repeat step 4 to 10; 12)!Increase the angle ! by 15¡ each time until ! =165¡. Repeat step 4 t o 10 for each ! . The test was done when all the critical distance R !-"-cr, where ! is every 15¡ from 0¡ to 165¡, and " is every 15¡ from 0¡ to 345¡, as well as the TotCnt and RSSImx of all the distance r of every 0.5 foot within each R !-"-cr were recorded. Then the CP antenna was used to measure the R !-"-cr of the tag by following the same procedures above. The tagÕs 3D critical read range could be outlined by all the R !-"-cr. The procedures for mapping the tagÕs acquisition 3D read range were conducted reversely. For each ! and " , where ! is every 15¡ from 0¡ to 165¡, and " is every 15¡ from 0¡ to 345¡, the measurement of the tag was started at the distance r that was 2 fee t gr eater than the corresponding R!-"-cr. The tag was moved towards to the antenna with an increment of 0.5 foot, and the first distance r at which the reader could read the tag was recorded as R !-"-ac. The test was done when all the acquisition distance R !-"-ac, as well as the TotCnt and RSSImx at distance r of every 0.5 foot within each R !-"-ac were measured. While both the 3D read ranges were determined, data managing processes for pilot test 1 were the same as the tests discussed in Chapter 3 . There were 288 sets of !-" measured for mapping !103 each of the 3D read range. Figure 54 and Figure 55 below shows the comparisons between the R !-"-cr and R !-"-ac of the tag measured using the LP antenna and CP antenna, respectively. Figure 54 shows that 93.40% of the R !-"-cr and R !-"-ac were equal to each other while using the LP antenna, and Figure 55 indicates that 87.20% of the R !-"-cr and R !-"-ac were equal while using the CP antenna. Observation of Figure 54 and Figure 55 shows that moving the tag outwards or towards to the readerÕs interrogation zone had little effect on the tag Õs read range. In order to examine whether the tagÕs 3D critical read range is statistically identical to the 3D acquisition read range, the same survival analysis model described in Chapter 5 for analyzing the read range test was used. Table 28 shows the test of equality over the different procedures for determining the tagÕs critical distance and acquisition distance using the survival analysis model. Figure 54 Comparison of the R !-"-cr and R !-"-ac Measured by the LP Antenna (GPJS QRGJPS RGUKS (GRQS PGRNS LP Antenna, Tag AD -338m5 with HDF, Comparison of the Critical Distance and Aquisition Distance at each $-%R!-"-cr is 0.5 foot less than R !-"-acR!-"-cr is equal to R !-"-acR!-"-cr is 0.5 foot more than R !-"-acR!-"-cr is 1 foot more than R !-"-acR!-"-cr is 1.5 foot more than R !-"-ac!104 Figure 55 Comparison of the R !-"-cr and R !-"-ac Measured by the CP Antenna Table 28 Test of Equality over Procedures for Critical Distance and Acquisition Distance Test of Equality over Procedure Test Chi -Square DF Pr > Chi -Square Log -Rank 0.0000 1 1.0000 Wilcoxon 0.0000 1 1.0000 -2Log(LR) 0.0000 1 1.0000 All the thr ee tests in the Table 28 give p=1.0000, which indicates the possibilities of reading a tag past distance r by moving the tag outward or towards to the readerÕs interrogation zone were equal. In other words, the tagÕs critical distance and acquisition distance were statistically identical. To simplify the procedures for mapping a tagÕs 3D read range, measurements of only one type of the maximum read distance is enough. Further, it was observed in the results of pilot test 1 that the ang le ! did not have much effect PGRNS KGTTS UTGKPS NGUUS PGOQS PGRNS KGTTS CP Antenna, Tag AD -338m5 with HDF, Comparison of the Critical Distance and Aquisition Distance at each $-%R!-"-cr is 1 foot less than R !-"-acR!-"-cr is 0.5 foot less than R !-"-acR!-"-cr is equal to R !-"-acR!-"-cr is 0.5 foot more than R !-"-acR!-"-cr is 1 foot more thanR !-"-acR!-"-cr is 1.5 foot more than R !-"-acR!-"-cr is 2 feet more than R !-"-ac!105 on the read range of the tag. Although the complete set of the pilot test 1 results is not provided in consideration of the reasonable length of this document , the analysis of the effect of ! on the tagÕs read range is given below in Table 29. There were 12 different angles ! in pilot test 1. Table 29 shows that ! did not a have significant effect on the probability that the reader can read the tag past distance r, as all three p values in the table are greater than 0.05. Thus, in t he read range test descrived in Chapter 3 , angle ! was fixed to 90¡. Table 29 Test of Equality over Angle ! Test of Equality over Angle $ Test Chi -Square DF Pr > Chi -Square Log -Rank 12.7816 11 0.3078 Wilcoxon 5.5241 11 0.9032 -2Log(LR) 2.6626 11 0.9945 The initial purpose of this pilot test was to determine the least po ssible amount of data that was necessary to map the same 3D read range with the complete data set in a simulated manufacturing environment. Specifically to say, in pilot test 1, the tag was measured at every 15¡ of angle ! , every 15¡ of angle " and every 0.5 foot of distance r. The 3D read range of the tag was outlined using this set of data. However, the test of the critical read range using the LP antenna consisted of more than 2500 Read Points, and that of using the CP antenna had more than 2000 Read Points. Taking these many measure ments is time consuming and not efficient for practical use. So, it is necessary to investigate if measurements with greater increments of ! , " and r can draw a statistically same 3D read range compared to that determined by the measurements with smaller i ncrements. !106 In order to achieve this goal, a surface regression analysis model was developed using the SAS software. The procedures for analyzing the data of pilot test 1 was planned as 1) Find the best -fit function of ! , " and r to predict the TotCnt or R SSImx value with the complete data set. The value of root mean square error (RMSE) of the regression function will be given; 2) Find the regression function with a smaller set of data. For example, using data of ! at every 30¡, " at every 15¡ and r at ever y 1 foot to determine the regression function. Then the RMSE of the new regression function will be calculated; 3) Compare the values of the two RMSE to determine if the two regression functions are similar. If so, then the data set with the larger increment is sufficient to map a 3D read range that is similar to the 3D read range using the data set with the smaller increment. The surface regression analysis results with the complete data set of the critical 3D re ad range test are given in Table 30 to Table 37. Table 31 shows that the R -square of the regr ession function defined in Table 30 was 0.4526, that means the regression function could only represent 45.26% of all the TotCnt values collected in th e critical 3D read range test using the CP antenna. The RMSE in Table 31 was larger than 130 and also indicates that the regression function could not predict the TotCnt value well. Similarly, the values of R -square and RMSE in Table 33, Table 35 and Table 37 all indicate that the complete d ata set in pilot test 1 using either the LP antenna or the CP antenna could not predict the TotCnt and RSSImx well. In other words, even with measurements at every 15¡ of angle ! , every 15¡ of angle " and every 0.5 foot of distance r, the 3D read range was not !107 accurate enough to predict the TotCnt and RSSImx value at a given distance and orientation of the tag. Hence, the method was considered as failed. Pilot test 1 di d not successfully determine a measuring procedure with greater increments of ! , " and r that can draw a statistically same 3D read range compared to that determined by measurements at every 15¡ of angle ! , every 15¡ of angle " and every 0.5 foot of distan ce r . Thus, in consideration of pratical time consumption of testing , as well as by reading relevant research , it was determined that in the read range test described in Chapter 3, the increment for " is 30¡ and for r is 0.5 foot. Table 30 CP Antenna, TotCnt as a Function of ! , " and r Parameter DF Estimate Standard Error t Value Pr > |t| Parameter Estimate from Coded Data Intercept 1 429.395522 17.146470 25.04 <.0001 237.618156 $ 1 -0.377823 0.225056 -1.68 0.0933 -11.975450 % 1 -0.140543 0.106373 -1.32 0.1866 -12.854500 r 1 4.642091 6.360012 0.73 0.4655 -285.035018 $ *$ 1 0.001921 0.001124 1.71 0.0877 13.073532 % *$ 1 0.000263 0.000473 0.56 0.5780 3.748578 % *% 1 0.000262 0.000242 1.08 0.2803 7.790216 r* $ 1 -0.030519 0.026365 -1.16 0.2472 -9.441808 r* % 1 -0.010830 0.013937 -0.78 0.4372 -7.005798 r*r 1 -8.972400 0.727842 -12.33 <.0001 -126.174377 Table 31 CP Antenna, Response Surface for TotCnt Response Surface for Variable TotCnt: TotCnt Response Mean 302.966443 Root MSE 130.966667 R-Square 0.4526 Coefficient of Variation 43.2281 !108 Table 32 RSSImx as a Function of ! , " and r Parameter DF Estimate Standard Error t Value Pr > |t| Parameter Estimate from Coded Data Intercept 1 -34.832497 0.630599 -55.24 <.0001 -54.901310 $ 1 0.050054 0.008280 6.05 <.0001 0.217177 % 1 -0.018359 0.003877 -4.74 <.0001 -0.306015 r 1 -7.779622 0.257372 -30.23 <.0001 -8.656255 $ *$ 1 -0.000334 0.000041360 -8.07 <.0001 -2.272750 % *$ 1 0.000008611 0.000017424 0.49 0.6212 0.122542 % *beta 1 0.000028897 0.000008837 3.27 0.0011 0.859874 r* $ 1 0.001548 0.001113 1.39 0.1644 0.446850 r* % 1 0.001476 0.000580 2.55 0.0110 0.891259 r*r 1 0.615511 0.031838 19.33 <.0001 7.540012 Table 33 CP Antenna, Response Surface for RSSImx Response Surface for Variable RSSImx: RSSImx Response Mean -48.943860 Root MSE 4.409493 R-Square 0.6678 Coefficient of Variation -9.0093 Table 34 LP Antenna TotCnt as a Function of ! , " and r Parameter DF Estimate Standard Error t Value Pr > |t| Parameter Estimate from Coded Data Intercept 1 451.198221 11.836325 38.12 <.0001 248.466267 $ 1 -0.404511 0.158548 -2.55 0.0108 -26.676099 % 1 -0.122817 0.073313 -1.68 0.0939 -12.547912 r 1 -9.579790 2.404469 -3.98 <.0001 -227.150132 $ *$ 1 0.003151 0.000795 3.96 <.0001 21.447806 % *$ 1 0.000004625 0.000335 0.01 0.9890 0.065824 % *beta 1 0.000327 0.000170 1.93 0.0542 9.740338 r* $ 1 -0.062797 0.010077 -6.23 <.0001 -33.674998 r* % 1 -0.009034 0.004897 -1.84 0.0651 -10.129247 r*r 1 -1.330519 0.158135 -8.41 <.0001 -56.214437 !109 Table 35 LP Antenna, Response Surface for TotCnt Response Surface for Variable TotCnt: TotCnt Response Mean 291.598593 Root MSE 131.517969 R-Square 0.4518 Coefficient of Variation 45.1024 Table 36 LP Antenna, RSSImx as a Function of ! , " and r Paramete r DF Estimate Standard Error t Value Pr > |t| Parameter Estimate from Coded Data Intercept 1 -31.544414 0.437705 -72.07 <.0001 -56.125717 $ 1 0.029538 0.005944 4.97 <.0001 -0.316744 % 1 -0.019646 0.002747 -7.15 <.0001 -0.277280 r 1 -5.620905 0.092702 -60.63 <.0001 -8.958679 $ *$ 1 -0.000266 0.000029868 -8.91 <.0001 -1.811815 % *$ 1 0.000012724 0.000012619 1.01 0.3134 0.181081 % *beta 1 0.000039680 0.000006369 6.23 <.0001 1.180722 r* $ 1 0.001237 0.000398 3.11 0.0019 0.637891 r* % 1 0.000489 0.000195 2.51 0.0122 0.527019 r*r 1 0.296380 0.006236 47.53 <.0001 11.577345 Table 37 LP Antenna, Response Surface for RSSImx Response Surface for Variable RSSImx: RSSImx Response Mean -49.588350 Root MSE 4.542763 R-Square 0.7184 Coefficient of Variation -9.1609 !110 Appendix 2 Pilot Test 2 Pilot test 2 consist ed of two sub -tests. The first test was to determine if angle # had an effect on a tagÕs performance. The tag AD -381m5 which was Tag 3, and the HDF that was the same as in pilot test 1 were used in the first test of pilot test 2. The RFID rea der and antennas, the laboratory environment and setups were the same as for the tests in Chapter 3. The procedures of this test were 10)!Assemble the LP antenna to the reader, and configure the reader to a transmitting power of 20 dBm, and a receiving sensiti vity of -70 dBm; 11)!Locate the transponder support at the distance r=2 feet for the entire test; 12)!Assemble the HDF to the transponder support and turn the angle ! to 90¡ and angle " to 0¡; 13)!Attach the tag to the HDF at # =0¡ using small pieces of masking tape, the center of the tag is aligned with the center of the HDF; 14)!Read the tag 3 times, 5 seconds each time, and 5 seconds delay between each run; 15)!Save the data file with information of reader antenna polarization, angle ! , " , and # in the filename; 16)!Keep the same ! and " , and rotate # to 30¡. Repeat step 5 and step 6; 17)!Rotate # every 30¡ from 0¡ to 330¡ and repeat step 5 and 6; 18)!Turn the HDF to ! =120¡ and " =315¡, and repeat step 4 to step 8; 19)!Turn the HDF to ! =150¡ and " =345¡, and repeat steps from 4 to 8; !111 20)!Assemble the CP antenna to the reader and configure the reader the same, repeat steps from 2 to step 10. The TotCnt results of the first test are shown below in Figure 56 to Figure 61. Each value of the TotCnt used in the figures was the average TotCnt of the measurement of three trials at angle #. Figure 56 to Figure 61 show that at three sets of !-" , angle # had different effects on the TotCnt of the tag. The tag at certain angles of # could not be read by the reader, no matter whi ch reader antenna was used. In a ddition, the angle # resulted in 0 TotCnt was not consistent across different !-" combinations and different polarization of antennas. For example, at ! =90¡ and " =0¡, # =90¡ and 270¡ resulted in 0 TotCnt while testing with the LP antenna ( Figure 56). At the same !-" but testing with CP antenna, # =90¡, 270¡ and 300¡ resulted in 0 TotCnt ( Figure 59). At ! =120¡ and " =315¡ and usin g LP antenna, # =60¡, 240¡ and 270¡resulted in 0 TotCnt ( Figure 57). !112 !Figure 56 LP Antenna, Tag AD -381m5 with HDF, ! =90¡ and " =0¡, TotCnt versus # Figure 57 LP Antenna, Tag AD -381m5 with HDF, ! =120¡ and " =315¡, TotCnt versus # !"!!#!!$!!%!!&!!!'$!'(!')!'"#!'"&!'"*!'#"!'#%!'#+!'$!!'$$!'LP Antenna, Tag AD -381m5 with HDF, $ =90and %=0 , TotCnt vs &TotCnt !"!!#!!$!!%!!&!!!'$!'(!')!'"#!'"&!'"*!'#"!'#%!'#+!'$!!'$$!'LP Antenna, Tag AD -381m5 with HDF, $ =120and %=315 , TotCnt vs &TotCnt !113 !Figure 58 LP Antenna, Tag AD -381m5 with HDF, ! =150¡ and " =345¡, TotCnt versus # !Figure 59 CP Antenna, Tag AD -381m5 with HDF, ! =90¡ and " =0¡, TotCnt versus # !"!!#!!$!!%!!&!!!'$!'(!')!'"#!'"&!'"*!'#"!'#%!'#+!'$!!'$$!'LP Antenna, Tag AD -381m5 with HDF, $ =150and %=345 , TotCnt vs &TotCnt P(PPKPPRPPJPPNPPPVRPVOPVQPV(KPV(NPV(UPVK(PVKJPVKTPVRPPVRRPVCP Antenna, Tag AD -381m5 with HDF, $=90 and %=0 , TotCnt vs &TotCnt !114 !Figure 60 CP Antenna, Tag AD -381m5 with HDF, ! =120¡ and " =315¡, TotCnt versus # !Figure 61 CP Antenna, Tag AD -381m5 with HDF, ! =150¡ an d " =345¡, TotCnt versus # !"!!#!!$!!%!!&!!!'$!'(!')!'"#!'"&!'"*!'#"!'#%!'#+!'$!!'$$!'CP Antenna, Tag AD -381m5 with HDF, $ =120and %=315 , TotCnt vs & TotCnt !"!!#!!$!!%!!&!!!'$!'(!')!'"#!'"&!'"*!'#"!'#%!'#+!'$!!'$$!'CP Antenna, Tag AD -381m5 with HDF, $ =150and %=345 , TotCnt vs & TotCnt !115 The purpose of the second test of pilot test 2 was to determine if different materials have a consistent effect on tagsÕ read range across different tags. The tag AD -227m5, AD -233m5 and AD -381m5 were used, and they were the same t ypes of tag tested in the read range test and the orientation read rate test described in Chapter 3 . The materials used in this test were an uncoated corrugated paperboard (material M1 in the Chapter 3 tests), a medium -density fiberboard (MDF, see Figure 62), a low -density fiberboard (LDF, see Figure 63), and the HDF that was the same as in pilot t est 1 and the first test of pilot test 2. Except for the uncoated corrugated paperboard, the other three materials in this test are not typically used for packaging. The RFID reader and antennas, the laboratory environme nt and setups were the same as for the tests in Chapter 3. !Figure 62 the Medium Density Fiberboard !116 Figure 63 the Low Density Fiberboard The procedures of this test were: 14)!Assemble the LP antenna to the reader, and configure the reader to a transmitting power of 20 dBm, and a receiving sensitivity of -70 dBm; 15)!Assemble the material sample to the transponder support and set the angle ! to 90¡ and " to 0¡. ! and " are constant for all the following procedures; 16)!Attach the tag to the m aterial sample at # =0¡ using small pieces of masking tape. For the uncoated corrugated paperboard and the HDF, the center of the tag was aligned with the center of the material samples. The MDF and LDF samples are not 7 * 7 square inches for the reason of li mited resource. So the tag was attached to the MDF and LDF at the same height above the ground as it was attached to the HDF and the uncoated corrugated paperboard. Angle # keeps to 0¡ for all the procedures; !117 17)!Move the transponder support to the distance r= 0.5 foot; 18)!Turn on the reader to automatically read the tag 3 times, 5 seconds each time. Delay between two runs is 5 seconds; 19)!Move the transponder support to 1 foot and repeat step 5; 20)!Move the transponder support outwards from the antenna with a increment of 0.5 foot each time, until the tag can not be detected by the reader for 5 consecutive increases in distance ; 21)!Record the last distance r the tag can be read as R max . Thus the maximum read distance of the tag with the material at !=90¡, " =0¡ and # =0¡ is measured; 22)!Repeat above steps for all the three tags, four materials and two antennas. The results of the second test are shown in "#$%&'! OJ and "#$%&'! ON. "#$%&'! OJ indicates that the tag AD -381m5 had the smallest R max with the uncoated corrugated paperboard, while it had the greatest R max with the MDF, and the R max tested with the MDF was much larger than that with LDF and HDF. The four materials did not have same effect on the tag AD -233m5 and AD -227m5. For the tag AD -233m5 in "#$%&'! OJ, the R max tested with MDF, LDF and HDF were the same, while the R max with the uncoated corrugated paperboard was 5.5 feet less than those with the other three materials. The R max of the tag A D-227m5 with the four materials tested by the LP antenna were identical or very similar shown in "#$%&'! OJ. "#$%&'! ON indicates that the tag AD -381m5 had the smallest R max with the uncoated corrugated paperboard, while the R max of AD -381m5 with the MDF and the HDF were the same. For AD -233m5 and AD -227m5 tested by the CP antenna, the R max of either tag with different materials were identical. So, a conclusion was !118 drawn from the second test of the pilot test 2 that materials may not have a consistent effect on tag performance across different tags. !Figure 64 LP Antenna, Rmax of Tags with Different Materials while ! =90¡, " =0¡ and # =0¡ !Figure 65 CP Antenna, Rmax of Tags with Different Mate rials while ! =90¡, " =0¡ and # =0 ¡! (JGP! (JGP! QGN! (JGP! (JGP! UGN! (JGP! (JGP! (JGP! (JGN! UGN! NGP! PGP NGP (PGP (NGP KPGP H:WKKT;N H:WKRR;N H:WRU(;N Rmax (ft) Tag LP Antenna, R max of Tags with Different Mateirals While $=90 , % =0 and &=0Uncoated Corrugated Paperboard MDF LDF HDFUGN! OGP! OGP! UGN! OGP! JGP! UGN! OGP! OGP! UGN! OGP! RGN! PGP KGP JGP OGP UGP (PGP (KGP (JGP (OGP H:WKKT;N H:WKRR;N H:WRU(;N Rmax (ft) Tag CP Antenna, R max of Tags with Different Materials While $ =90 , % =0 and &=0Uncoated Corugated Paperboard MDF LDF HDF!119 Appendix 3 Figures of TotCnt and RSSImx of Each Read Point in the Read Range Test !Figure 66 LP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Point Read Point !120 !Figure 67 LP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Point !Figure 68 LP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Point Read Point !121 !Figure 69 CP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Point !Figure 70 CP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Point Read Point !122 !Figure 71 CP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Point !Figure 72 LP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 1 with M 3, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Point Read Point !123 !Figure 73 LP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Point !Figure 74 LP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Point Read Point !124 !Figure 75 CP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Point !Figure 76 CP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Point Read Point !125 !Figure 77 CP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Point !Figure 78 LP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Point Read Point !126 !Figure 79 LP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Point !Figure 80 LP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Point Read Point !127 !Figure 81 CP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Point !Figure 82 CP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Point Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Point Read Point !128 !Figure 83 CP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Point -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Point Read Point !129 Appendix 4 Figures of TotCnt and RSSImx or Each Read Orientation in the O -RDRate Test Figure 84 LP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Orientation Read Orientation !130 Figure 85 LP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Orientation Figure 86 LP Antenna, Tag 1 with M3, TotCnt and RSS Imx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Orientation Read Orientation !131 Figure 87 CP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Orientation Figure 88 CP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 1 with M1, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 1 with M2, TotCnt and RSSImx of Each Read Orientation Read Orientation !132 Figure 89 CP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Orientation Figure 90 LP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 1 with M3, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Orientation Read Orientation !133 Figure 91 LP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Orientation Figure 92 LP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Orientation Read Orientation !134 Figure 93 CP Antenna, Tag 2 with M1, TotCnt and RSS Imx of Each Read Orientation Figure 94 CP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 2 with M1, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 2 with M2, TotCnt and RSSImx of Each Read Orientation Read Orientation !135 Figure 95 CP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Orientation Figure 96 LP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 2 with M3, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Orientation Read Orientation !136 Figure 97 LP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Orientation Figure 98 LP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt LP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Orientation Read Orientation !137 Figure 99 CP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Orientation Figure 100 CP Antenna, Tag 3 with M2, TotCnt and RS SImx of Each Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 3 with M1, TotCnt and RSSImx of Each Read Orientation Read Orientation -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 3 with M2, TotCnt and RSSImx of Each Read Orientation Read Orientation !138 Figure 101 CP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Orientation ! -70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 050100150200250300350400450RSSImx (dBm) TotCnt CP Antenna, Tag 3 with M3, TotCnt and RSSImx of Each Read Orientation Read Orientation !139 REFERENCES ! !140 REFERENCES [1] X. Zhang, ÒThe influence of water content in hydrated superabsorbent polymer on dielectric properties, antenna radiated power, and RFID tag readability,Ó Michigan State University, 2006. [2] J. Tazelaar, ÒThe effect of tag orientation and package content on the readability of radio freque ncy identification (RFID) transponders,Ó Michigan State University, 2004. [3] G. Robinson, ÒEffects of radio frequency noise on communication capabilities of ultra -high frequency passive transponder,Ó Michigan State University, 2010. [4] M. R. Ryan, ÒA mod el for the implementation of a radio frequency identification system into a warehouse environment,Ó Michigan State University, 2002. [5] R. C. Eppelheimer, ÒAn evaluation of the UHF radio frequency identification technology for application in archaeologica l repository management,Ó Michigan State University, 2010. [6] J. Onderko, ÒRadio Frequency Identification transponder performance on refrigerated and frozen beef loin muscle packages,Ó Michigan State University, 2004. [7] T. J. S. Crawforth, ÒThe effect o f antennae configuration, product and tag type on readability of passive UHF RFID transponders,Ó Michigan State University, 2005. [8] J. R. T. Falls, ÒThe effect of conveyor speed, packaging materials, and product on the readability of radio frequency iden tification transponders,Ó Michigan State University, 2006. [9] P. Sanghera, RFID+: CompTIA RFID+ Study Guide and Practice Exam (RF0 -001), 1st ed. Syngress, 2007. [10] Y.-P. Luh and Y. -C. Liu, ÒMeasurement of Effective Reading Distance of UHF RFID Passive T ags,Ó Mod. Mech. Eng. , vol. 3, no. August, pp. 115 Ð120, 2013. !141 [11] H.-G. Choi, S. -S. Kim, M. -T. Cho, H. -J. Joo, and E. -S. Lee, ÒA Study on the Tag Performance Test for International Standards Using RFID Emulator,Ó in Computer Science and Convergence CSA 20 11 & WCC 2011 Proceedings , J. J. Park, H. -C. Chao, M. S. Obaidat, and J. Kim, Eds. 2011, pp. 621 Ð632. [12] S. Ayyer, ÒEvaluation of passive RFID system in a dynamic work environment,Ó Rutgers, The State University of New Jersey, 2012. [13] S. D. Mello, E. Mathews, L. Mccauley, and J. Markham, ÒImpact of Position and Orientation of RFID Tags on Real Time Asset Tracking in a Supply Chain,Ó vol. 3, no. 1, pp. 1 Ð12, 2008. [14] L. Ukkonen and L. Sydanheimo, ÒThreshold Power -based Radiation Pattern Measurement of Passive UHF RFID Tags,Ó Prog. Electromagn. Res. Symp. Proceddings , pp. 87 Ð89, 2010. [15] R. Clarke, ÒThe Influence of Product and Packaging Characteristics on Passive RFID Readability,Ó in Smart Packaging Technologies for Fast Moving Consumer Goods , J. Ke rry and P. Butler, Eds. 2008, pp. 167 Ð195. [16] R. Bridelall and A. Hande, ÒPerformance Metrics and Operational Parameters of RFID Systems,Ó in RFID Systems: Research Trends and Challenges , M. Bolic, D. Simplot -Ryl, and I. Stojmenovic, Eds. John Wiley & So ns, Ltd., 2010, pp. 23 Ð55. [17] K. Finkenzeller, RFID Handbook , Third. John Wiley & Sons, Ltd., 2010. [18] D. M. Dobkin, The RF in RFID , Second. Newnes, 2013. [19] M. Bolic, A. Athalye, and T. H. Li, ÒPerformance of Passive UHF RFID System in Practice,Ó in RFID system: Research Trends and Challenges , M. Bolic, D. Simplot -Ryl, and I. Stojmenovic, Eds. John Wiley & Sons, Ltd., 2010, pp. 3 Ð21. [20] GS1, ÒImplementation Guide for the use of GS1 EPCglobal Standards in the Consumer Electronics Supply Chain,Ó 2010. !142 [21] R. A. Kleist, T. A. Chapman, D. A. Sakai, and B. S. Jarvis, RFID Labeling: Smart Labeling Concepts & Applications for the Consumer Packaged Goods Supply Chai n, 2nd ed. Banta Book Group, 2005. [22] Impinj Inc, ÒRFID Standards.Ó [Online]. Available: http://www.impinj.com/resources/about -rfid/rfid -standards/. [Accessed: 21 -Mar -2014]. [23] J. I. Aguirre, ÒEPCglobal: A Universal Standard,Ó Cambridge, MA, 2007. [24] I. Poole, ÒRFID Standards.Ó [Online]. Available: http://www.radio -electronics.com/info/wireless/radio -frequency -identification -rfid/iso -epcglobal -iec -standards.php. [Accessed: 21 -Mar -2014]. [25] I. Poole, ÒRFID Frequencies and Frequency Bands.Ó [Online]. Available: http://www.radio -electronics.com/info/wireless/radio -frequency -identification -rfid/low -high -frequency -bands -frequencies.php. [Accessed: 21 -Apr -2018]. [26] GS1, ÒRegulatory status for using RFID in the EPC Gen 2 band ( 860 to 960 MHz ) of the UHF spectrum.Ó 2016. [27] EPCglobal Inc, ÒSpecification for RFID Air Interface EPC TM Radio -Frequency Identity Protocols Class -1 Generation -2 UHF RFID Protocol for Communications at 860 MHz Ð 960 MHz Version 1.2.0,Ó Intellectual Property . 2006. [28] EPCglobal Inc, ÒSpecification for RFID Air Interface EPC TM Radio -Frequency Identity Protocols Class -1 Generation -2 UHF RFID Conformance Requirements Version 1.0.6.Ó 2011. [29] GS1, ÒInteroperability Test System for EPC Compliant Class -1 Generation -2 UHF RFID Devi ces Version 1.2.4.Ó 2006. [30] GS1, ÒEPC/RFID.Ó [Online]. Available: www.gs1.org/epc -rfid. [Accessed: 20 -Jul -2014]. [31] GS1, ÒEPC Tag Data Standard Version 1.9.Ó 2014. !143 [32] N. Tai, M. Frey, and G. Ow -yang, ÒGS1 EPCglobal Transport and Logistics Industry U ser Group Implementation Guide Ð TLS Pilots,Ó no. 1.4. 2010. [33] GS1, ÒRTI ( Pallet Tagging ) Guideline,Ó no. 2. pp. 1 Ð21, 2010. [34] Wireless News, ÒABI Research Report: ÔCargo Container Security and Tracking,ÕÓ 2011. [Online]. Available: http://go.galeg roup.com/ps/i.do?id=GALE%7CA271749622&v=2.1&u=msu_main&it=r &p=ITOF&sw=w&asid=707c147ea88d8133806eebbd59fc0fba. [Accessed: 02 -Jul -2013]. [35] GS1, ÒMoving products across borders with greater visibility and efficiency.Ó 2011. [36] K. Daschkovska and B. Scho lz-Reiter, ÒElectronic seals for efficient container logistics,Ó in Dynamics in Logistics: First International Conference, LDIC 2007, Bremen, Germany, August 2007. Proceedings , 1st ed., H. -D. Haasis, H. -J. Kreowski, and B. Scholz -Reiter, Eds. Springer Berl in Heidelberg, 2008, pp. 305 Ð312. [37] J. Patton, ÒRFID as electronic article surveillance: feasibility assessment,Ó University of Arkansas, 2008. [38] GSI, ÒLevi Strauss & Co. Pockets of Opportunity with GS1 EPC -Enabled RFID.Ó 2014. [39] GS1, ÒGS1 EPCglob al RFID -based Electronic Articles Surveillance ( EAS ) Technical Implementation Guide,Ó no. 1.0. 2009. [40] H. Lehpamer, RFID Design Principles . Artech House, 2008. [41] EPCglobal Inc, ÒTag Performance Parameters and Test Methods Version 1.1.3.Ó 2008. [42] P. V. Nikitin and K. V. S. Rao, ÒTheory and measurement of backscattering from RFID tags,Ó IEEE Antennas Propag. Mag. , vol. 48, pp. 212 Ð218, 2006. !144 [43] D. Deavours, ÒUHF RFID Antennas,Ó in RFID Systems: Research Trends and Challenges , M. Bolic, D. Simplot -Ryl, and I. Stojmenovic, Eds. John Wiley & Sons, Ltd., 2010, pp. 57 Ð98. [44] L. Ukkonen and L. Sydanheimo, ÒPerformance Characterization of Passive UHF RFID Tags,Ó in The Internet of Things: 20th Tyrrhenian Workshop on Digital Communications , D. e. al. Giusto, Ed. 2010, pp. 229 Ð238. [45] P. V. Nikitin, K. V. S. Rao, S. F. Lam, V. Pillai, R. Martinez, and H. Heinrich, ÒPower reflection coefficient analysis for complex impedances in RFID tag design,Ó IEEE Trans. Microw. Theory Tech. , vol. 53, no. 9, pp. 27 21Ð2725, 2005. [46] ÒRadar Cross Section (RCS).Ó [Online]. Available: http://www.microwaves101.com/encyclopedia/Navy handbook/4.11 Radar Cross -Section (RCS).pdf. [Accessed: 13 -Nov -2014]. [47] V. Derbek, C. Steger OVE, R. Weiss OVE, J. Preishuber -Pfl, and M. Pistauer, ÒA UHF RFID measurement and evaluation test system,Ó Elektrotechnik & Informationstechnik , vol. 124/11, pp. 384 Ð390, 2007. [48] Y. Huang and K. Boyle, Antennas: From Theory to P ractice . Chichester, UK: John Wiley & Sons, Ltd., 2008. [49] National Physical Laboratory, ÒDielectric properties of materials.Ó [Online]. Available: http://www.kayelaby.npl.co.uk/general_physics/2_6/2_6_5.html. [Accessed: 07 -Nov -2014]. [50] ÒDielectric Co nstant, Strength, & Loss Tangent.Ó [Online]. Available: http://www.rfcafe.com/references/electrical/dielectric -constants -strengths.htm. [Accessed: 07-Nov -2014]. [51] R. Suwalak and C. Phongcharoenpanich, ÒImpedance Effect of Elliptical Curved Surface on RF ID Tag Performances,Ó in 2017 International Symposium on Antennas and Propagation (ISAP) , 2017, pp. 3 Ð4. !145 [52] Y. Ziyuan, H. Xiaoxiang, Y. Yang, and T. Liu, ÒUHF RFID Tag Antenna Performance on Various Dielectric -Background,Ó 2016 IEEE Int. Work. Electromag n. Appl. Student Innov. Compet. , vol. 5, pp. 8 Ð10, 2016. [53] J. Bolton, E. Jones, R. K. Punugu, A. Addy, and S. Okate, ÒPerformance and Benchmarking of Multisurface UHF RFID Tags for Readability and Reliability,Ó J. Sensors , vol. 2017, 2017. [54] EPCgloba l, ÒDynamic Test ,: Door Portal Test Methodology For Applied Tag Perormance Testing Rev 1.1.4.Ó 2006. [55] EPCglobal Inc, ÒStatic Test Method For Applied Tag Performance Testing Rev 1.9.4.Ó 2008. [56] GS1, ÒTagged -Item Performance Protocol ( TIPP ) Tagged I tem Grading ,: Testing Methodology Guideline.Ó pp. 1 Ð17, 2017. [57] ASTM International, ÒASTM D7435 -08 Standard Test Method for Determining the Performance of Passive Radio Frequency Identification ( RFID ) Transponders on Loaded Containers.Ó 2014. [58] AST M International, ÒASTM D7434 -08 Standard Test Method for Determining the Performance of Passive Radio Frequency Identification (RFID) Transponders on Palletized or Unitized Loads.Ó 2014. [59] ASTM International, ÒASTM D7580/D7580M -09 Standard Test Method f or Rotary Stretch Wrapper Method for Determining the Readability of Passive RFID Transponders on Homogenous Palletized or Unitized Loads,Ó Annual Book of ASTM Standards . 2014. [60] P. N. Roque, ÒPerformance analysis of effective range and orientation for U HF passive RFID,Ó 2008. [61] F. Lu, X. Chen, and T. T. Ye, ÒPerformance analysis of stacked RFID tags,Ó 2009 IEEE Int. Conf. RFID , pp. 330 Ð337, 2009. !146 [62] W. Bin Zou, Y. Wu, and Y. Zhao, ÒAutomatic testing system for UHF passive RFID tag performance,Ó in 2009 International Conference on Networking and Digital Society, ICNDS 2009, 2009, vol. 2, pp. 79 Ð82. [63] Z. Huo, Y. He, B. Li, K. She, L. Zuo, and Y. Zhu, ÒDynamic performance test on UHF RFID tags,Ó in The 5th International Conference on Computer Science and Education (ICCSE) , 2010, pp. 1336 Ð1339. [64] L. A. Kosuru and D. D. Deavours, ÒOptimum performance for RFID tag immersed in dielectric media,Ó 2011 IEEE Int. Conf. RFID , pp. 11 Ð18, 2011. [65] R. Colella, L. Catarinucci, P. Coppola, and L. Tarricone, Ò Measurement Platform for Electromagnetic Characterization and Performance Evaluation of UHF RFID Tags,Ó IEEE Trans. Instrum. Meas. , vol. 65, no. 4, pp. 905 Ð914, 2016. [66] P. Stasa, J. Svub, and F. Benes, ÒDevelopment and Evaluation of Automated System for RFID Tag Performance Measurements,Ó Adv. Electr. Electron. Eng. , vol. 15, no. special issue, pp. 613 Ð625, 2017. [67] Avery Dennison, ÒAvery Dennison RFID Solutions.Ó [Online]. Available: http://rfid.averydennison.com/content/dam/averydennison/rfid/Global/ Documents/RBIS -Web -RFID -Solutions -portfolio -v1.pdf. [Accessed: 02 -Jun -2014]. [68] Iminj Inc., ÒMonza¨ 5 Tag Chip Datasheet Rev 1.51.Ó 2014. [69] Institute for Digital Research and Education, ÒSAS Seminar Introduction to Survival Analysis in SAS.Ó [Online]. Available: http://www.ats.ucla.edu/stat/sas/seminars/sas_survival/. [Accessed: 12 -Jan -2015]. !