...... .. ...»... . . . . by . 4%.. J. 2 .... I? 0.....00 0 .3"? ...”.v; 0.. . . v0.0 . . Lt. ... . w, I . . o . . .... ...: 2.. -.. ... 1.57”? $.85... . 11; ... .3 . r8...» 9.3.3....143. . ‘ 1 . .K .0. . I ‘ A \...A~...An. . .0 .. .... "”30. 0 ..090. .. .... Q’.0’..a . 9‘. l g. 00 . 0.1.3033 0.0 J .bl.».”fi ... {‘0‘ .fi. u 0 0 ...... .101... a... ...0 . ... . ..L . «.... . .... ..m." ...n ... ......ev 1. u. _. 0- . .12 0. 0. 0| . .n . _ . 0 . o n . t . . t . . . p . v . ...h... .0. .3. .58....t L ... . . ...? ....8. ....xhuofi: . .0.~ .0. 3300...... . . .0. .v .8. ._ ......1 :0. , v.0! ‘- 002.‘0.0.. P.0000 ‘60. . . . “‘0; “no.” .0 .0. 0...... 0 .. 0 0 ...—...; .10.... . .00 .A . ...-.... 3.......... 0.... u .11.. . . ,0... M .t (K... 2 . . n .. L... .0 .o . .‘0? ...- . . . .n. .... .. . ..0 2 , ... . .....- .wflkfl. . .. a. . ... . . . - . . . -410... ...-...m ...... . . ... o so...» .3. .. 0.. 3.1.3.. . 00...: .. n. A... .... 2... . 2...! .. .. ... ... .... : . . CO... . ‘ . 0| .33r3” . .00. v0, 0.. 3K. . 0. . .x. .Yu0 0. 00 o . .0 .0 . ... .00 $0000.... ..0‘ . .. .. .fi 7. _. .. 0P... 3...... . . . .0v . . a. 25*: . a. I J 0.0 . A .. ...... . . . . . ., . . . A ... ... . ... .. ..0. , . . l. . .6. 0. .. 3’. .0. . 000a a...“ W . . . .. H... 0. . ......l. .0.- . ..00. . .. r .. . . . . .. ... 0 . .0. . u . 0 ... 0... , 0. ....0 0 0. ...: n.“ ‘0! ”7.? . ”ft. .. 0 . 0. .0 . 0Q .0. ... . . . _. . . A . ... .. .0. 0. .. . £§flm #0 VII. . . 0.3a" ..«F: 0. . . A n. .0 0 d0 .0 00¢. . 0 I, A 00. . 00”— .. .... ., . . , . . . 0.0: 0 u.....h~n. m.‘o.u«? . 2 .. . . . «0.441.304.0100. .2 ..0. 3e . . .0: 0 ....0. 2.01.... . . 9 ..0 0 04 . .... 0700. . A0. . .. . . _ . _ 31.. .00....‘.00 .30.... C. .%.u'. 0.00 .00. . ... . . .. .3... .0. 0 . .. . . . .... 0. ... .. ..0.» .. ... 1 ....f! .1. 0.. 91.. , . . . . . .... .. . . , . . _ . , .1 ......03000130... 0.... . «.0. 0.08 :10 0., 7 u- . . .00.... _. . ... .- . 0.... L0... ...0... Y. .... c . ~ _ A... 0.001.... | . ... ... .Ih . . _ Y... . . 1.0.0 3.0.0.. .Ifiw. .. 0.. . . , J 1.1314. ..l... I 4 . .0. 5203.30. . 1 .... . 3.... a... .81.. .. . .. 0 ... o. 0 ... . . 0 . . . .r... ...-.0. .. ...‘3 0.00. . . . . . . . . 10.. .. .. ... f. r. . M. . 0 A... w , .. ... 73.»... .... .l . 00C... ... . 00.0 0-. . .. .... .5 3... “w. . . T 2...... .. . 0.....10 w. ..0 .... , .0‘ ...i... 0 ...A 0... 0...] . . . . A . . was: .02 ....A. 00000 ...... . 0 .1. . , A q. 4...... u , . ... . . n.\. .... s 1...... .2. _ ...... .... ...... A.. .. ... .... . .. .... . .. .. -..... _ . . .... . L}... ......é. 0.. it... .... _ .. . .. . Chi... 1.. 5 1.... . t , . .... .0 0 .. I . . 4 .21... l. g. .r..... 0... .. ...! . .0. .. . . . . 1.0.A3.\0.1J0.... I . a... .J.1.10...\.6 ’ ..er . u .. . 4‘. . 0 .01.»)... 50.1511 .0 ...; . . .53. A . . ...0 A. .139 up ...v... ...... .0. 1.. .0: o . .. . 0..“ . ....u o. , .. . . . .. . , :0... u .. . .0 .‘fJor‘ .3.” £0. . 0‘. . . bu v ”H9! .. ,Y 0 ,0. ‘..0. . 55.90 . ...-u .. .. . 4. .. .0. .r‘... . 0 . L 00” ..0. . V0. A .. . . . , .... .avA. .0 u u 0... o. .. 40 .9. r .04. I‘ . . .u . A , 1.5 ... .0. ..J ..Lu db- .03. r63.....‘ ‘0...“ . ... 2.0.. z ... 0. _. 32.11.31. .. '. 0.2:...00. . .4 . .. .w .. . 5.11.... . ...- ...“. :20. ., . ...J. . . A. (MIN... ink)... I. “. .05 5. .3 ...0. . ...0. ......I .00 ..t...‘ . .0 .. ..00. . 00*» 0 .0. ...... . . 0.. J . . .. A u}. ...... . .0 1.301. .. 0A. 0 If ”....t. . .v .01. 0‘9 ...-3A1-.. . . . , .038 ... . . . ...... .9. . .. . 27;»: . . .. . ...t .. . .... rug 3... A (.... .JWN.U . .....fl. ...LU. .xA ; .... t. .....u. -. .. . . . . .... . . ... . . ._ .53... 2.... ......t... _. . . _ ...) A3... 3...} 00;..00A ,0 .0“: 00.. . _. .... awn. . 00.... .... .... K. .... a . .. .0. .. .I . .0. A. .0 .0: . . 1 . .12 .... .1. .f......rn.l...¢m+.4 . .. .... . 0. ... s . 000.07- . .. . . . ...“. ... .....n. «50.0.: . . . ... . A. ..0. 70 .0 .o..... 4.0. ...! . 0r) A0. . «1t... .... . L L \\A .. . .... .A ...u 3...... . on... ...... . . .. .. e .... é . .. 3:32.: ., ...... ...}; ... ... .. . , If. v .. J ....u .5... r0. r3 .0. .. V . . it...» . ... 3 .v . 0 2...: .. : . . ..... ... . . 0 . . . .. 1 ft... .... .27 ‘0 ....svf... . «P1974. 2U. ....r 90 .5 .3818}... . 0... . . . n .0 .... 00.. 3...... ... o .. .. ...01. 7 fl». . ...A. ...k. .. 0... 0 . 0...... ...l. I . . . ....l. .. .90....- .0 Q0. 0.; ... ~hr... w... k V 308...... 0. 5.210% 0.2.0.000. .. .0 .... 1 0.0.. . ......0... 0 . .09... . .. . .... ...»...S..._10,..1 . .. 3fi...3.?....0.. r... I. . t . ..uhu. ......A <\:A . .ui .. .00... . 00. .. ... ... . . ...... q .. ... 1.0 .. 0 . 0. . .. . .. .. . . . ... .v 2 o... .. . .r. ...... .J‘. 0.4. r .I .40: 11/.-. .A\,, .... . ..u. . \E. . 8 unfit. . W. s. cit. nnfi. 0b.... In" ..I.. ...«..0..3.0“..m....... .w. .. .. ... 0... . .1803.“ . 1...... “uni“..VAfiJq 30.. . .....av. . .. . it: . . ... ...0..U.... Q ((5%... . Q. .n 0. .l0.. 0 o . . A . . 0. n. . o . .. . . .. . . .. . .. .0. c . . . 0.0 . . . A . ,4 0. . (a. . ,AA 0 3... . V‘i h3.‘§.;..3....... I.“ :0ng 0.3-... ...... . $.38 . 3.4.3010... . .... .. . _ Q. ... . .... . . In. .... .... Six-20.. . . .....Avofi . fir! J... (H.031. .. 0.3.5750}... : .. d“r....... 0. .0...000.000.0...‘.1..... 00L. . .... . . .. . o .. 0. ... .. 0 . ... 0.. ... . . . 3... ......h. .0.0 . . ......‘9 .(war., , ..A .1” (0.“- .0. . y . .3300 .. A0. a .... 0.01.1: ...? . v I. . . .1 . ...: .. .. . ... ... ... - . . .... . . A... . . Q ...“..If ... , v. v: 3.. 4 .... 000.330... 00.. a...“ 4. . . .fi. €§...2.0: . 0 . In”‘.0 L. o .0. 0 0 .m.». 0 . . . 0 .450“. .. 1&1. . ““00 ...”.U 00. . 3000“”) 00-;- A ‘1.“an 059.490. 3 15...?! 0.. . . .30.... ... t. .0. . ...... .. 0...... .. .... I . . . . . ...o. . .l ..«v .... ....1 . . ...-.0. , £3.51. .01.». .. .3021}. 1 ".... €013: . . ...... . .. 0.. 0.... ... («Hz-00.00.... ...... .A . ..0. .... .. ......1 ... .. . ....1... . 1.3.3000. ...—s... .8 .3. . . , £53.. 3......f40. .. ... ... 3 .\ Z. c . Q. ....o 0-. . . .... ...: ... I. A: .. a l . . .. 0.907. . .6er . . .0 .. . . . . .....0..._. .... . . 1 40.. 4.... . . . .0... . 0.12. O 1 J A 0.x}4. PA ...-.00 . T ‘u‘. .7 0;, Jun-scraw. ... 0 00.0 . 0:... ’.= 063....?. .h d0 .. . a. . .0.- 0. 0 .00 :0 .....0 .o . ... .0 .. .001. 00.0.0 . 0 .. 00....2. A ’2 . I "0 “$00“. ...... I} ...... .0. 0...: .0. ... .2100 .2! 3.1.... ..-: .... .. ... . v .. .. .. t... .. . . A . . 2v... . .-.-1:0. ...; .. .... .. . 30.9.3: .....10 ....3... . i... ...... x1050.) .... x 95.0.3... . $50? ... I I .00 . 0.. . 0. 0?. .2400...‘ 00. 6... 60.0... .1 .. . .. ,0. 5.... ... . . .0. G . ... ..0 19.0 ... . 3.0.0230... ...1 1‘30! .....10~.....v.0 ... Q. .ffilftl . .f... r. . .... .1; .L’ “.0 ...-00 0 .0. . a .00 09.0. .. 006 ’ . ..v... 0.-. . . ... , .. . .0. ...?“ ...... . . £01 0.... A! 0. 0n.- ., . .m- 0 .... .00.... .4 out 0. 00 . , ... ”.1 .2.v.0 .H .M’ I... a ...... ...... . . . ..............n “...... . (...... . ...... :3... .....e... .... .... .a._p............§..., ...............u. s .. 4.4...." ... ... . . n , . .1... H15... . . 0. ... 0 ... .. .. .. . . . . .A .. _ . .....1? . . .. ... .. o. v . A. A . 2 VA. . , 4, . a” ”(Janus-v. mgQuI . ...! .. .. .0. .. . .1. . .0 . ...r . . . ..loun .I. ‘ .0342... .0 ..- 0.) 000 3.3.”.mvm. 0..“0'0; 5% . 0M0“; “A: lfiflfilq ,HW. . . . .0 . . ... .. .0 f. .. . ...00 . . . . . . .. ‘ . . . Ian. 0.30“: .u... 00.1. I . . 00. . A . .0)... ... ...0 fl. o‘. . 00.0.... 3.1a...- »1.». .0. . *0. H40...“ A . . 3. .01.... :0 e. .... . . $3.20.... . .. . . . ... l...»0;...1.l.fl . nun-15.3: . . . , . z. ..._ .3 .... 3......» _ A . 33...... .... .. .2. .1. 3...... .. ..., . . i. . ..0 a . i... . , .c h..\0..\‘ I 0.. . .I “I. .o . 3300.. ‘Jmov . 009000.00“: n! 0. 00200“ ‘03..“ 30...“. 09...; . a .r.‘.. .9.” .4. 00’3““ 0. ‘wal . . IA . 0 . 0 .... v a . . 0 .0! A. . .0 ‘> . .. . .. . . .. . .. . . :0 . A . .. 0.000.“.4: 0... ”John“ I'm ..n nu. .5 00’" .... n 1.. 0 "0“ “roubfi 0.51:)...000 thi: 4. A . . .0410: - u 3. .3“ d .30”... “05......Inwv“ .00‘nh- . .... _ . , . .. 0 ...u. .10.". ...: .25.... . .... ...... ....-.nwi .m:...... . . . .. . . A 030.32.... 0...... a. .... ...... .50.. (.01....) 0! 4 3 . ‘3 . 1 .3013! . . .. 10.00....”0.‘ . v .00... Stu-.9 a 0 .’. ,. .. . . . . . . 000.... 0 .00 . A. I. .030 u 4. ..a»... ’0 305‘... . .30“ 0.1.... “0.03.9.4, .Fv»: . 00 1. .... . . . 30...... .. ... . . . . n X 300 v.1... .. .....0...’ . . . . .3032...” 0 1.45.0. .... 0.. . ...I’. ....0 .0 ‘ 0?“. 0.000.007.“ .00 . . . . n . 0. .. .16 0.... 0 7A... . .. ...-out. 0 I r 00..., .0000: o. .0... .... . Afia.‘ .. ....0 0 . ... $2.0 .23. a . .‘00-..€0 29.0.3.0. ... . . ...-0 . .. ... . ... .. ... v0.9.3 .0.‘3&..’ 0 50% 2 0......00.o,u . . . .0 0.3....) do.“ .. 20.00% ..c. .. 00 . . V . . 0 u .. . :1... . . . .07... . .A’ ... o 1 . ......AU .00. . 1?“. . .500... . . ...-5.1.: 0.130.. 0 ..., .0 l)‘..0._. «a... 10.. o. . .....Jd. ... . . . . 0. . .0.. . 0. . -0. ., .. , ....0... I00 . 0 (1.3.0.. .0 In . . .. coffiiuuit. ......32 A0. 0..., ..0. ... ”-0.086. 3.0.0.... .00...“ .. .0. .. o .0 3.1;, .0. . . » .... . .. .6. T”. $30.33.»... 0 . . . . ”....VS....0.S . x“ . . A 0. . .1 00. _. . 0 0.0 .0 .10., 0 .011 0 0 . L ...004 0.. .. .. 0. . . 1.004.. . . . . 0 .0 0. o: . «D. o v 0.. 00. . v. 0. . _ . . v '0'. .0 00 .0, .h.. 9300...“.3“ 0 0010 ...NA". 0....” ..p\0..xu..0.‘. $70.91. .0. LL... .5 ... 1. . 300:0. 0" r .. ......o0. ‘ . . . 0 .36.... . 9A . ...: . I... . 1.00 . . . . . . .. A . y....?&011§0—1}u—r. . ’23.... {007 A 2 0 ... 0... ...-5.3%.. o. ..l 0. A... ...V. \..§ ..5, .... . . . . . o .....s. .. .....‘u. . 0. 0 ......3...‘ .0..l}.§.0 . . . . ....0E..fl“'l..’u00od~.i... .4 is". .0. r. 3.0:... o r... 0. c 0 ...Jvm0.0.l..v‘ .0 0.‘ Artf‘ .0130}: ’ a 0 3.3.00.0“. .0 . 0 .0 .0 ....0.3‘. l 0 000. . , .000 ...." 000 a 90.0.... ...! J! .00, .......0\ ”I t .00. é‘#0fld0 0... A. l n . . .40“. g I?! 0.3!? 3-... F. 3...... u. | .1. 2.00.3. ._ ... A. . . . .0 .00. .... . . . . . . . ... .. . .....i. .. ... . ... . ..A. .. . .. . ... . ... . u . . .... (.0010 (...... i .0 1 . .............A. '7 .0. . _. .0...- .. dorm. 0. t. n. .0 .... .000 0. 4 . .0000010'. 0 .000... ..v 0.0...“0. 0 . ...100‘0. o , . . .00.. 90.00 ...0.....3.Q... . A. ... . . . . 33.0.00 .. . .0. . .0... . .. .303 . A A. .3...0\?0 . .0. .0... . .. .....(x. .. 2...“ 0' 0.0.1 0‘15".- «0 Vlg‘s... ‘0 . . I. . q. 01.1.0.0. h: in“... ... 0 0 I .0. 0 . ..00..0..'L.. .3049? .0. v . 0.... . . c . 0030...... ...... . f . _. p. . .. .. . 20...“. .. 0:000. .80.. . 00...... 0.00.0. 1 a. . 1 . .0.- .r ..0 ...|0 0).. .... .0 $.50... .0.r... A .0 R. 0. «...... . . . .... . . .... .02....- . . .o . 0:51.33... .1000? .. . ... . . .. ... . .. .... . . t .. 0 .....1 .. . .. . . . ....(. 050. . ...........0 .. . $7.30.... a . , . . 1' 3". ...... p. 0 0.. .. . v0. . ....0' 0 0 .0. ..‘1‘C0000!0..00 ......0 ... .. .00. . . .... v00 ...0... A. A 0.0.. . y 0. . 0 0 . ,. .. . ...: 0 . 3.3.0.9 . 00.0 . ...».W: 10.0.0. 00“ 000‘: A. 011‘ 7»... .0.!.L0.0 6-..:- . gri .0. 0 t ..00........0.x .. .0. ... .. .300. .l. .. ._ .. . C ...... . 0...... ... .I. . . . ... .... 1.... .. .. A . . . . . . A. .. .. .. .. . V . . ... ....x. ... . ...... ... 05.0.0 .5; 00 .A . ...A. .- 8.0.8... 00:0 3 . . 07;...- 0.0 (I. . .... “...!” u 0.03.0 ...... ...? . . . ... Q ...). .. 2... .. ... .. . . . .. .... . . . . .. ... ......02 .. A . . S O . V n10. :3”! . ..0. ..000 .0.....,Qn.. .00.. ..‘1 0.. .....Z... .0: .3530... ..k. 30.0. .0. . A... .00... . . . ...: _ ....0 . .2. . .0.- 0 ... $56..» . A .. 0 .. ... : . :10. .0. . 3.. .... ......0 . .. ... 0.0 o .0 0...... .. 0.. .... 0 _ ... .A . . . . .. ... 3.3.1.1....3 . ... . .0..lv ... ( an. A N07... . . . .. . owiA 0.1.00. 0. : ”3031. .0 _ .0 . 0.0. 1.01.“.- .. ......xcoo .0. ... ..g ‘2 p 1.0... .10. . . . ... A. -. r I; . ... ..0 .0. (.30.... {1.09 . aunt; 0i 7.. A. . I . A.§ ... .o 0.. a... . Z... ...0 ...; . 5.0 0...... .... . . ... . .0: 14.01.0232 . . .. . 012...... .. 3.83.: h u . to... ..l ...0 ... . ..l"..00‘u .000; .00.. 00 . . . .0 0 ...... a do .0 ...I0. 0.0.0.1.}... ... , 0 c 40. r .' ... 0.. . .. 0,. AmA’ I0 .. . . g? . A . I1 L .- ... ...: .01... .A .. .....d.’ . ... ... .. ‘39.... . . .. .. ... . 9. . . . A0... . ... .... . ..Ao. . .. .0. . _ ..U... 4 GM... ..n»\ ....‘..I. u. .11.) ‘. . 0.... .... . . 0. ' 0. ... 0.0. ..0. 0.. ..0 . . .1....d..0l3.. ...: \ ... . .. . . .., 0000. .. . . no.“ 0'... S ..\ 0... c '00 .‘br: 00.00 ‘01000 .0 0...... A . . . . ....0.. u . . .... V00..00-00&.0§.l¢ 09'. .0 . i 0 . .13. 0 .... 01.00.00 .‘3 ..0' t. .. ... 00 .0 .v . . 0 0... v a . . . 0. t .0 0 9 . 0.0.00... 0....0, 1.. ., . 0.0:): . ... 0.. 03.10 ...»? ~ . . . ‘.«\.....} .... .. . ...-00.... 1.0.. 0:! ...... .0003 ... . . _ . . , 0.0.! .. ... ... .. . ... .0 . . . .4.. .. 5 . .. ...... L ‘ 0 .0 . 049.000 04".? . 00’ A0, 10 9.00.“. ...-0 0.303.... . 0‘ . 00.21,. .0... .00 . . . _ . 0.7 .000 ..0.‘ 0 . -. g\.-0...0.I0. .0. .10.. . . , . «AL ... . ... 0... ... 0.. (:0 .0. 01... . . .0... . . 0.0 .. .. ......20307 0. 0.. . .00.- ,» .0200. . . . 10.0 . 0.. 0.. .. 0. . ... .. ...... 0 .. ...... ... . ..O . ., 3...»: .’ . . 00er “00+ .w .0000... .0. GI: . A. 100*...0‘0 0 ..00... 0.0 ’00 . .070 0.0....0’00000 D 0: 0.. ' .0 . 00.0 '. Q 0 0.... 00. .000.‘00..01 0.. O. .Q. ...0. . .. 0. .....2 . o .4: .00? ‘00.. 900000. .A0..:0..0“.€0 . 0 0 .0..‘. 0.. 0 .1 0000.0 .0033.‘0§. . . “L [10 y 0.. .....1A.... . ..I ...0... ...?004 .0.... 0.. 0v .0. . ‘. \u 00 . ...A. . ' 1‘, a 0 :0 00. I ..3...0 0 . V A. . .0 0 .10. . C 3.0. 0... . ... . . 0 ......0 70.. <0.-. I ...0. 0.00. .. o. .3 00 t ... I 0 a . .0 105.000 r0 ... 0 A. , A an)“. X X .. ...-Pr.- .SI... ... ...: 1 .02 10.5031... . 3 £010.03... .....r. ..90..«........ . ... a . . . .10. . ... $83.01, . ...... :8... .. ... 0. . . .. . ... ... ...-0.....30 .... . . 0.00... .....0... .9... . . . .w o «01% I. ...... ~10 ... .... I! . ... ... 01.. ....0. .. ...... .... .0. v ... .0 ...... 0...... 3.330. . ...: 0.1.0.. .. v .. . ... .... . ..u . .. ..0 .. .. . a. .. .... t .. ...001\. . ..n...:... J J. ADM Jaw. .w .0 a”- 0 t .‘0’. ......r. .... 0. 00. .L. .r . \‘lw. .3. .0 . rt. ...000 .... 0.0.0.0. .00 . o 000 0... ... . . . 3 ..A. . . _ ....0 .140. 0 0 0 I. .. 00 I ...? .0. .0 n . . .00 .3... ...0 00 ..0. .J. . Q. . ...... O. ’7. ..-..0" .01. ‘1 «.0 .u’ 2‘...0.. . 0.. .010. . {0.00: 000.0,... 00. .0 .3.- ..‘0‘2 . .A .0 . ll 00. n I . .v ..v...‘....ld 34.00.3310... _ 0.0... 000 00.0. .11 . . : ..0... .8... 3r<.8......3..fi. 3 ......v... . 0:10... 3... . ......0. ... 0. a. 0.. .... A0. . .. ...... ... : 0. .. 1a.... .... 2.0.... .. _ .A. ... 3...... ... 2...: 2.03.103... .... I." o. .. V... 4.... ... ...... 31.3070... . .... 0.... .. 0. 0 .\ ... . . ...... .. .000. . . . .o ...I. . . ... . .0 ...: .0... ... 00.0.. . :0. . 0 000.. d I . . e.0:.¢00~ ,. ... 10.. o. . 4 0 3...» S ...... ..00. .... .. .33 .. 3.7. ... .0... . . '4. 0. ...... 5.0.00.0. . I .0. .. .0... . I 3.30:... . . 1......0; ... F33: 9.03.0.3... . ..- . .. .3. “.... .0. ... -v...’ .... . 3010.0.“ AR0 021.. . .0.0.0.00?..0..~ .0. .4. 9.... . . .' . ,. o ... o. .0 ...... ... . . .. 0. . ... .w00. 3.7.7.0. ..0 . . .83 ...t 1:00.10. 00:0 . :00 .. . . 1|. . h J . O i 1.3.3.30 Snuff-...... 3....{I‘ .I... 1... . . 7.23.01.30.31... 930.11.! . v0.3.1.1; . . . ... ... . .0: . ...... ...-... 0. ...).......... . ....nt.... ... . . .... ...! ,A 0 . . ..1 0 .0. 0. .. ...... ......0: ......0... . . .. ..0 .0 ,0 f .I.... x ‘0‘; . n30... .. . 00 31....) .00 3.093.000.1120. . .. 0... .0 .0 o. . 0 . . .0 .. .\ .09....1 . 0 .... .. .r . ... 0 .... A. 1...... . . .- ... 0..0.~ . ., . .03. .... . F 0 .00.. 00.. . .4- .\ .1.n...-. ... .0... ... £00.01. 30-0.7.1... ...-0... . A. 00 00 0:00. .4 ... . .000... ., .IO‘.....‘. . . o. .l .... ..00. . 9.. . 7.430.»... .... (I “Wit-0. .. 200:... 8.5.0. .p . . . 3N . .. A.,. * ..,.......:v.u? . ............P. .. 3 . ..\ O....‘\ A . .. . . ... ... .... .. . 400.0 ... . t 6.0... .0.- J: . . ... . .... .. ... .... . . .0: . .....arn . . .. w a 0 . 040. ‘O 0.0r0-I ! ".I‘. .. ...; .... ‘00 0‘ .01 u. .00 0.0.0.00 . .00..V1§.00‘ - 00.0 0.0..“ '10 C0 .0... 0.... 0’00..0 . . . . . 0 . . ‘.A~.N“H) o 0.0“... _ u. .5. u . 0.3.7.. .. .0..‘.00.. 11...". .0. ...}...g ... ‘0 0v 00010.40 0 f. 0-0.. . . .R. 00. 0o .00.... .0. . , .. .‘0. . .I. . O 0 . . . v.0 .. 0.0 0.0 . .0. z... .0 ....J........v..v. ..l ....{0 ... .00....0. . I ... .. i...u..§.0.0.0~ £40.. 023:30... 0...... . A 0..3..‘...TA083\~...3 . . . . 0.0 ...... . .‘A. .. .. ..0 ..g 0 l‘ 0. 0‘ 0 Q1 . . 0 0...... Oat-000.... 00 .... v ...000 . . .0 .0 0 .. 300...... O... ...v. ‘ .0....~‘0 0' 0- ‘00-. 0 . ~ I . . . v. «9000‘s .. . 0 .0 . . u- |.. ...fllw: . , . . .n‘3.0.4. . . .. a . ....0' ....111... .05 w. .....0! 0.. :00... ... 0. .03.... 0... .. . ...... . . . . .. . .. .-.... ... . . . L .0.. ...r .0.0-1A.. .7 ...00 .0.- . [.3001 . .. I. . . A. .0 .00 000 0 .00.. 0 30.00.3090 00 . 0 .0, 0.. I ... 0. 0.... . . ...0A . o 0... o . . . . . t..00~........o....a..0f.flu . L . 0. . 0A. . 2.6.0:... .0. . ..00' L 0. . ...!3.=.I‘ ‘ ..J C... ..O 0.00. i: .90. c. 00.... ... 0000. ...-‘00 . . . . .. . . . . . . A . 1.000070. 0‘ .cv . 0 ... 0.0.1.... gl}..00.. . .00 100.04 .Lv‘.0.:0.. .9..00....‘00. '4 .00... . 0.0 .. . 01.... 0 p... ..00 0 :00 .. :0.‘ 0 . .0 00 0000 ...0 0 0.0.... 0 0.0.00 I .0. .0." :.‘0 n.‘ . . . . .....J: a . . 0.4000 0 A ..0..“fi.0§ .. 0.0.00. 0 p 0.2.... t... .. . ...1. . A .. . _. 2...; I . r. . .. $1.. 0021......3 00 . . 00. 9. . 0 ....0 .0 .0 .00 ... . ...: ... ... 0.70 ... \ o . . . .. 00... .0. .00 ......J....;.. . ....)T!...... ....I.V1..0. . .. .. ,‘J... ..n ..‘Ev’x0sav . . fl .0 .. ... 0. .0. .9: .5 .. 000; 0 l ......7‘ . .0... . 0... 000. .0. 00. . .036. y .0. 0 .. . :00334 ...0 ‘ .. ... 0.0.0 00.00..|00 ‘00.... .0 1.... 0.... .... 5r0'u0 . ...,- ..1 .... .; .9 .. . .1. ...I. .1, . I. ...x. o .. .....0. . . .... ~.A.!..-7—0 . ... .. . . .. . .. . ,. . 0 . .... i 0 0 p . s .... ... .. . .0 .I. 0.... 6.0... .. 00 00. ... ... .300. 0. ..00000 . 00 . 0. v0 . . Q 00.0 .. . 0.0 .x. 1.....00 0 . .. ...r... ... 0 .. .321... ...... .0. 313:... . .. .. 0 0... . .... .u 0 .... .. 0. 0.20.)..1 0 00 . ......40 ... . . . . . . . v... ....v; . .3.......0..0.~_ . ..n 0.... .0:.00....00.. .0... ....30 p .. .03... 000 .4. 0 0.. A0 .0 .. .1. . . ... Q0 .. . .l. . . . . . . . . _ . . _ , . . 0‘. 03L :0 .... .‘y ...!‘1...:. . . . 00.....V0 .91... .0. 1... 1!. .0 0...... ... .. . .. 0.0.30 ...0...0 ... . . f . 0 00.... . . . .. . . . . .. o o. a. 0 .... ... u .0. . r. . .0 0 ..0‘ . n 0 .O. 01,700... .0... . .. ...00 700 .. ...-0‘ 0.»... . . . I‘ 00 ...... . .. . .0. . , u _ y .. . . 0‘0”- 0....1-..’ .0. 0.4 ......3 $041.... .... .. .. ...-00,. 3......5 . . . .00 ..0.0..‘.. ......f... ... . 3.1.0)! 0... o . 0. 0. .00 I .. .....1 .. . .00}... ... A 0‘0. .... 2.0.90.1 .0. .. n. 0 . . . .4 .v\ . .. . .0 a . . ... 000. .O.‘.. .u 0. 1 u . 0....0 . . {0.0.300 200...... . . ... . . . 0...... .00 000. 00.0. 0 ., .30....0.‘ 1.. ...0.. HO. 0.... J. ... 322.1 I... ... 20.. ... . . . ...: $0.0... 010. . ... . 1?. . 0... .I. t3...~0 ...-0.0;... .A...-.. ... ...l .....0...'0. . ...? o .0... . 7... . . . . ... 030 . ....9. 3...: I033... (...!3.0.0.....<.:.13.05.1.02? .. . . .00. I020 . . , X 0. . .... . .. . D. I . . . . A '0. 000000 0. 0.0.10.1: 0...... . 009 .. 90 O u. ‘ 3.0.1.... ...-0 N . 0.. . . . c . 0.. . . .009!- .0,\ .0"... . 0. AT 0. 0‘00...“ . . . .... ..‘.00........0..0 ... ..0 .- 0.. A ‘10.... .0... 0 ....l .00... .0. ‘0... .0. ‘0: . 0 .. ..0. 0.1.. 4.7 0 . . . .. . . . ‘ .00 0.. 0........ . 0 , ...,0..0..1‘. 0 0. 00.10.20.00- .. .. c .. . .0. 3...... .310... ... . . .vo .... . . . . . 0.. .v . .... c .. 0. 0 v.0- 0.. ..A. . . . _ 0." .!,.0.... . ...-:51... ...v . .. .. . . ... ..0 ..K. 0.30}; . .. “$0.... .0 , . .... C ..f..-...0'. 240...... 90.. O. .... .. .00.. o .........O. 00.....0. .. 1 .. ... . ......10...‘ V . . o 7‘ . . . 3 . 0 .... . .. . . .....J .00. A... . ...-..0. 0-0.2... .. . ... v. . ...00..\00. 0 0'. 0... .‘l. «0 . . 0 ... . 100 ... 0 .0... . (:00. ......0 0...? 0 . . ...v .0... H 0.8;.11. . . . .0 .... l 3. . ...! 0. .. .\ .l‘. 0 . .... ... ...00030...0..00 1 ~ .I..Q .00.. . . .t. L {In .. . ..0 .. .00 ... 1 0.0.0. 0 0. . . .0. ..0’L’..Q§ a... . 0.0‘ . .0 000.300 .0. ..0. . ..0 ‘s 000 .002"! —. 0 . .0 00‘! .0. . . . . . 7 ... . 0 , .00..» .... l.- 00... .. I. . r00 . .... A . ....» .... ...-.0 . .. .. ... .. 9 00. . 0. ; ....) .. 0. ...... .0 .... .0. ... 0 .. Au... . . . ... 0 .0 .A .r... . {.00 .010: 0.0.. 0.00 v3... .0 0.0.. 0. .20 0... . . 0.. 0. ....000 0.. .0 ..0 . .|. . .0...‘0.0.000 .. 0...» .0: .0. . ...... . 7.0.0. - .. . .. ...0. .. 09.0.! d 01 . 0. ......3. 0 0. .00 l o ..00 c0.0..l\ 0.0.! . 0.- .. 00 . 03003.... ...00. 0 0 00... .. 00.? . 0 _ 00 ‘ . ... it‘da 0 .... 00 .. . 00.0 L... 0‘ O. a. . . . . 0. b. t...,.‘.. n ... 37:13:03... . 0v . ... . . .. A ... ... .0... ...... .. 0.. ... .. ... .. ‘M. .... . . ’32:): 0.0. ..J 1 Jan”... . . .... 0.. .....0 . .. . 00.. .. .5 A . v.0 . . 0... 0.. .0 . 0 .. . . . ‘ 1 O. . .0 . C... .I. . . . . .. ..vl . 0&0: L0. .... 0....0 .13.. 00. 09.0I. .. . . .4. 0. E. . 0.010.203}... .0.. .. 0. r. r. .0... .. ...... . 0. .. . .. :0 . ...... 0....00. . l .... 0. 30.8.1.3. {30...}... . 00200.». v... .... . .... .1 3:2,... 2... ... .0v. . 0. . .. J. . ..0 .o c. 0' 0 .... . .. -.....- ..I 0. ...: . . . . 00.0....\.z....‘t0.2...0 . .0300. 0 . 0. ...; A .0. ... O 00.. Iiiv 90.. 0.. 6.0 . . 000. ... 00 . . .. ... 0. 0 .. o. 0 0 . .. 10000003..... 4 0 .0:0l:..onh .li.. . lo. . .... .. . .... ... 0 v . .... . .. . A. .I: n a». 4.00.0310. .3 . . .. S .0 .0 .. .10. .l . .. . ..0..00. ..0 . 0... ..0‘. ..0 9.. 0 as... ...-.0 .... 5.00 .0 0...~ (000.300.. - ....‘.. .. .. ‘. ..0... ....00 . 03.0.... ... ... ... .. ... 0 ..3 .... ..0 00. . . . _. ...}...2. . 1h. .. . . . . . . ...... . 0 . M. I. 03001.0 .. ..‘00-7‘...‘ . .. .... . .. .0 0.091.000 0.0.. .0 . .. . . A.. . . ...-v.0: ....- ... ..r . . ..." L ..A 0000:. 00;. _ ‘3‘ 0. 0 .. 1,139.... 0 .. . . . .....1... .0. ... ._ ...L! ...0 ....» . .0..0... 3 . . 2.5.... .0. .... . ... . . .5. .0. .. .. . 1 . I... A 0...... 5.0.00.2. q .AJCQJ 3 0001‘ ' . 0.....70... .000 .0 . ...0-..’.... ' ......:. 0.10.70... .0 . . . . .. . . _. .A. ‘0 .. 00. . .2.....0... .. . . _ . .. , .00. .. . . . . .01 , . 0.0": 0. . A. . , . ‘- . . . . 30.0..“ . .00 . 0.00. 30....“ £300! .0‘ ‘93‘..‘. .MP. 500.030.03.00... .x.‘...‘b0 9 ....-.0.1‘ ... A d. . 0.. .- y. 3 .00 .0 20.0.3.0 . .0 u A .. ... 0... 3.107.... ..., 0.. 0 . . ... .. 09100.... 00 0 3.2.... . .0 p I. 0 0.00 . . . u a o . . A . . A . It... ..0 0.. . . .. , . . . : . . . .. 0 . . u .. ..A... 0. “who... .. . , . .. . . . .. .. . . 000. ....u . . Sham ..“.00.§§ . . . .0... .0 .00 . 1.0. .... “...... .... . ‘ 0 . 313...: i :, .1 0.0, 0 00....0! I. .o 0....- ...O. ._ I. .0. ... 22.2.0.0. ,9. .. . . .....z ...0. .. . .. 0 .00.. .0.1..Q.: .0309... . .04.!- ..... a ...? 0 01...... ...0000020. ... .3 . 11.0.00. . . .. ....| .. . v .0 . Q00... .. .,0A . “'0. ..0ui0. .. 0 u. .520...) O.0.0-. .. ...‘l .‘ . .I. . .. v. o. 0 .. . . . . , . . 0 I. . .00. ‘ . 2.0.. -. 0&0 .00.... ... A. | .50 .... . ..IQ _ . . .I .o ..0 0‘ r;0........, ... .... ... .... .00....130 .. £00 .. .0! .0. ... .- ... ... . 0. . ...-.0. 0 .I. 0.0. .0 . . . . . 0. . . . . . . , . ... O...0 ...-u..- Isl-L ..0 . 3.1.... 0...! . . ..0 .‘O 0. , ... 3530050.... I o. . 0...... 0.. .. . . 0 ... . .......o . . . .‘ . .... ....030. 00 a 0‘. ~- 15.000. .1307 0 0 0.. .3. .. ... .0- . .. ...... 00 0 00;. . 0. ... .J A . . . 0-. . , . A0. . a . .o 0%. a. ..g...000. . .. . u . .x ....0... ..00 .. .0 0. 0., . A .U .. . 0 ... 0 ... . . . . . . 0 - ...: .. . . . _. . L .0.. .0 ... . .. .v..( ... . 0 0“ .. . . .. 0.00.0. .... .. ’0. .93.... ., ... . .00. L... ... 40.. :0 . . ... . , . 0000. 0 3‘...) . . (0.21.05.001 ..I . . . .. ...v . . . 0.. 0 .. . . , 0..|. . . 04.. . o. .0. .. . , r . . .0. . . . . . .... . . r . . . . . .... . 10.0.... .3 ... ... .00: 0:... 3...!“ .I ...... ....QRI. ... ... .0 .a .... 0 .f. .... . . . 30.0 0.... ...0. ... . . . .. .. .4..00.... . . 5.0 . .‘ ‘0 . .7 0 ... .. 0.0.. .M o 9- . . . . .0... ..l w.... . " ...... 0 . .. .0. . 0 . 0. o. . o. u 0 ‘00. . A . . . .. .0.0..0 ... . . ......0._ .I....h0 ..00 .0 . ...-0. . . .... 0.00. 0 0 0... o. I. . v.30. ...... 0 .. 0 . ..0 00.. . 0.3.1... . . 0000 00-.» ... .....00 0.00. .... ... 0. 300... . ....s . ..0. . 0 . . .. l 3 .. ... .... 0 ... . c _. . 6 . .0. . .3. on .-I .L . 3- .WJ .. 0.0030013. .0. ......0 a. .10": . 0 3"; .... .v .. 0 . ; A u. . :90. . . . 00 0.. . . ..0 00. . A .. .. .0 . .. . A. ‘IoH 00...? 00... ..k....l?.~w¢.0.0. A 1.0. 0.0320 .... . . (A. . . :0. v ‘. 0... ... ..00. 0. ...0 fl 0......... . . ..A :0: . 9. s .. t. ... . .. 0 I o a .. . 0 ..0 l. ‘0‘ ..l. . _ .10.... .00... .... .. o, ,0 . . ....pr. .00.. ... . .0 y JL..000. v 0:" 0‘ 0. .0 .. . .0 0. .... .0 .o .0 I . Q. . .. ..- .. :00. .. 0. . r . ...-.... . a. n .0! Z ..0‘ A .10....0. 03.:0: .0 L70. 0. ..0I..0...O.000v. 0.. ‘00. . .0. .. 0. . . ~.‘..&..t.r 0.0 ... . . . . s .... | .. ‘ ... . . 0 .. a . 0.. .0 . . A 0 . .Q . . . ., . . .... v 3 ... . . ....‘0 J.» ... 1. . . . O .0. ..0 . 0 . Qfi _. .00. o. I f... .00. .. .l 0. 0 0. . .. 00 ... 00 .v .0: ._ ...0. o 0 D A 0 . .0... ... . 0.. ...\ 0 b ..0, 0 .. ’ .0. ,. . r .. ... 0 .. . ...: :0. . . . A. . . O A . ..O .10. 0.... ... I. 3.0.: .. .. .. ......0 ...: 0. .0. A. u .. .. A ...... _ . . ... . . .. . ... .. v. .0.» . . .V .1... | O . 0 ..0 n. . A 0 .o. . .V A 0. 0.7!. . .00“ ... .. 0 . . 00 :0 0 I . . .0 .‘.0 A . .. ...? 0 .6... ‘0. 0. .03... . . ...I ..-. -«.00.00. .0 ..I. .. 07.00 0.0 .0. ..0 0 ... . . ¢. ..0. w. 0.0. .... .... .. 0....- .. I... . 0 . s . 02. .0 .0 0 .9... 0. . .— ..n r so. .2.‘ .....0 3 .. ...0 . . . o ...y. 0. 0.. ... n 0......0 . A... . . 1...... . . 0.....‘0 0.. . . . ...... .00. ...... . .. .. . . . I .. ._ .1... i .. . . . I 0 . . 0 .. . . . ... ... .. . . .... . .02... .0. ...... 0-90 ....Y.03 .... ...... .. .....00.0.. ...0. .A ...: 0.. . . . 0. .000 .. .. c. , . ... . ... 0...... 0 .....Qv! .. . L 0 a. .. ... .0. .A ... .900...~ .A . . . 3.9.... . . . . . .0.- . .i . . c .. . . .00.: ...»?04‘.‘ . .‘ka ...-0. (... 0 ... ., .. . .6 A . . . ., .... .0..CJ?. .. . .. . .... . ... .. ... v . ... . . . ... .... _ ......l; .9... \ 0.. .. .0... 31003.0 1.0.01.5}! . 0. n ..L I ......4 r K. 3.} .0 !.. .0... ...... .. .0... . .3... ... .... A . .. .. - . ... . ... .... v . . . . 222...}... 1.0. . . . .. . 7!: 11.00. ... 00.11431...0..0..0.! ‘.103‘H30§'..1,.10- .0 .... . 0,0. . 0. _ 0 . ...... . 0.. .0 . . . , . . .... . . . o .. . .. . . .. . ... ... .. . _ .. . . . ..0 . I . . 3.5!... .9 0.0.0.. ... ..0000..l.0. 0.0:‘000..’.4ll. . .. . .v ..-. 2 0 . '0 . . .... .. ... 0...... .....0...0 A . . .. 1 —... ... ..n. . . .o n . t -... 3.0. .- . . . . .. ......TN .10. . (...i‘: 0 .. 0 .. . A .. ...... 0A. ......o... .0 ... 0... u. . . . . . A I. . .. 0‘... X 0 _ .... I. . 0 a . ......0. . 0. . . 0 4'90. L...0...o 0. .... .. . _ . .. . .. . 0A ... .. ....l0'0; .0- . .. t . ... 00 . o. . . ... ..00 . o . . . .. V 0. .. 0 ..0 . V..Y, .... . . . . ,. .. 0.. 0., .. 0.... 00:0 0.. ...... 0.0.0. .0. . . o 80.. .0 . . 0. .) ... 5.00.. . A .v. 0... 0!. . .. 0 . .... 0.0:... 0 . . . . . . . . . ...l . . . .... .. o . 0A . ...... 00. .... ..00I.u. .. O o . n 0 0.00. r . . ..0 A... . . . .0. . s0. . .Q . .v000.00.v0l..l... . .050 L... 0. .4! .30... . 0 0 0 . ., . '0 0 . , 0' 00 . .0. 0. 0.0 .0. 000... . Q. . . . . .0. .0l . .. . 0-. ‘ .. 0 .... . . . .20.. ... .02 .. 00 . . Ir. . ... 0.... ..H0.V.O .~ .. .i .. I. .. I! . .. 0.. . . . ... ..0. ...... Z . '0. . . .. . .. ...00 0‘. . . ... 1.309 ... .. ..v . .. 0. . . q .. . {iv-...... .00! \ . 4 Q. o. i. _ .44. .0 .. 0 0 .fi. 1 0. ...... . . .0 0.... . .J 0 . . ... 00.0 I. $3.0... ... I: 000. 0. 0... 0.:0000 A . 00 . A3 0 . , .0 . 0.. . - I... 0 as...“ 0.. 0. .. .4...) ,. 0 .. :0." I0. . 0.v . .. \ .... ... A: ~ ... . . .... . . , .... 0 . . .. . . . ,. . 0 .0 s. . . . .4. 10.1.0. 0‘09... v. .. V . .00 ...-.- . . .. 00 .V. 0 . ....l .. ...: 1.. . 0 . .. u A . . . I. 1 . .... ... 0. . . . . .‘0 0:... ... $50,... _ .. . .. .v , A .... .. ...... ... . . .0: {...-:1 . .. A ...... .. o ......0 . 10... . : . ... ..0 . ..0. .J...A..v \ .'..I ... .0 0 . ... . .... Q . .... 0 0......00.. . .. ... 0...... 0 .12! .It . ....vf .. .... ... .0... A .... 3. . v. . .. 5.1.... ..0 0 .-. ......D . ... 0 . Q. J 00 . :0 . . .. . . .3‘ . . . 0 .0 :00; .0 .. . . . . . «(ugh-"fiz‘ .l“. _ 0... . .12.. .... . .. . 0.... .0. .. 0 .0 000 . . .. P.» . O. o . .0. .. . 3 u . . . .. A ... b0”v“’l‘"£} A I 0. . . .... 7.0. 2 .... A ..I0. 0 0. . a” . 0. I I. v . . ... 0. . .. ., . . . . ‘3..‘10..’ 310‘ ”98‘- . . . A . . 63:... .... . x . . I. .. . . ... Z .3. . .. ...... .. . us: . .. . . . . ... 000.03.18.‘..$0u10o .. . 1 .0. .. .. . .... ‘ 0 ., . o ... -. . .. . .v I ... . . . . .. . . ......00. 0‘. 00v. . . . ..01 .u...... I... 40, 0. 0 .. ...0. x 0 0 ...:0 ..100 ..—. ... . .. . . . . . , . 000:0. ..‘0.‘.0‘50.0 NA . ..0'0. . .0: . ... . Q ‘11. . _. 0 . .0 .. O... . v .. Ill .0 . . 0.. . .. A3 . o. 0001.“)... 1000.23... .0 1‘09 0 0' . .0 ...} .. . ... I. . 0.... . .0. ..(o A . ‘ .. 0. . . . 103.7 . ... . . A . . 1 0- ...b .20"... no... .0000. £010 ... 0 .. I .01.. ... :u .. ... 0 .0. V > 0 .1: ' s. A... I . ... . . ..0 A .. m: ’. . . . . .... ’00.... 0 . .0...00...0m.v.l‘ .‘1. . . . .0 ..A .. .. v... .. A0 . . 0.. . . . 0 00 0.0 A. 0... . . . . .. . . . ... ‘ii‘rol.’ A ...I. .0 0.. u'.)..'.‘. 0...... .... 1.....0‘. I- K. .0.¢0V .. . .. .0. I .I .... ...} . .. .. .. ... .00... I... ‘T; 0 ‘ I.... 00.. .0. .. 0 . .. .. . . ... . .. .. . ... . .. . , . 0.....81 ... 5 0 .0 ... C... .... .....9...Y o... C. . .. .... . .... . 0 .... ,. . o . . ..~A0 1... ...00. .v.‘ 0. ..0 ’IO..' .A 0 '2 I. ,. r... 0. . 0 _ 00....3§:0..00‘ , 00.8. .O/o 0 7040.. o .0 0 .II 0 .O .80.. 0| .... 00.. . . 0 A- 00:00‘ .. :07: 2... ...: ...... v0 1.. . .... .A . .. .0: .A .. ,. .9}. 0.0... ..‘r . .0. ..k .. V .. 0.. ‘ 4... . v .... [.37. ... _ . . . . 0.3.0.. . .o 0.... a . ... .0 .. . . . $0....0..0000.10.Jl’.’of 01,001.! ... ... o ....A .- 0‘30 l.0~ 0. .31. .3 who . .- .... .. u 0. .. .. ... _ . 9:01.! .... . . .. . .0 0. ...... .. . . . ... .. . . . ..!.i.‘.0.0~...2001 0‘ 00.0. 04):..n00...’ ....0. 0 2.1.00. I ,. ... .. ... ...... V .. 0. . ..0. .. .. . .o ‘ . . ... .. . . .. . . . .. . . . . . .. . . . .. . .. 0.3.10.0. ..‘0 0... 30..“ A0. 333...... .30“ ..X...~A...00. .... ... .... . .Vw _ . 0' | .30. .0 _ . . . ..WI . A .v .. . I. ...... .. .. . . . . . . 70.7.22: , . . . . . .. . .010..0.A.0....f01.9!‘.0.0 20.00 .....‘000. .0. n 0.. . ...... L.:.v\ Tl . , ... 0 J ‘ . . I v . . . . . . . . . . .03....000 . 0.. .. .0 0. . . . 0 ... .. 00a .0000 0. ... . . . . . . . . . .‘.‘01fi0,0.00’0§‘| v00 5"‘ .’. . c 0 2.0:! ._ .... A. L. . ... 9.00 ..I. 0 .. c .. 0100‘ 0. 0. . .0 . .00. ..0. 0. o. . 00 . . . .Q- .0... . . . .. .I . 0 .0. I: 0. 0.. . . . . . . . . . . . . . ...-.30 I120. ‘.’00..l".g'l‘ .0 ; 6......9- \. —. on“ . u . .x. . . 0.... . . . ... . . .. .. 0. .0... ..0 . . . . 0.... 00 I . . .... n o .u 00 0 . . . . x. . . A..0..‘0 ..."r§‘..\0'.00§0¢ A. .7 0“. : ... .. . 0.. . . .. A. . . A..,.. . . . 0.... . . . ._ . A .... . . . ... \ .2. 0. .0. . ... ...... 0 . ..0 . . J. .. . . .I..0.0..00..vru... 000.000. ....)‘I - 000‘0 ..3. ul.. .0. .. ...CA...0.. Vn. ... .. .0. 0 0. Z... ... ... 1. ....... . 0. 00. v.9)... 00-. .A. .~.‘. , .. n . . ....v .. , . . . . '00un.lul... 00.....0‘,'.20’0\‘0’0.‘0'. ... ..0 ... v1. . ...: . . ... . .. A . . .. ... -u . .25.... .. .003. .. . ...... ......3 . :0. 2...: .. .. . . .. . . ...: . .....0. . .QQP“... $4.. (05.40. ”.‘h... 0h 9:10 0....»‘0’0... I O a. A .. . l....‘..9r. . . . .. . .... . _ 0. 1 . 0 a. . . I ... .. .0. .. 0.. . . v .00 00.0. . 0 .. .0. 02...... .10.. )0? 0.00".) 0‘ ... ... v . 9 ....0 ....3. ...-o. I O . Q. . a. 0 . . . .' ..0.........0. . 0 .. C... I .. 0 . ...... . 11. 0. o. .1... . . 0 0. . . .. .. I0 l\.l. . ...0. ’0. .. Ion-...! . .— . .- . 0- .. a ...0 .. . . .. . s . . ., . 9 . .0 0 0. . u. . n .0 L. .0 0 . 1 ... 0r . C“ v .0 0.. .0 I‘”. b. .9 _ v . . ... ‘.v.. 2 0 .00 v 000‘. 00.». 0 7.. A. . c0... .... . 0.000. . . v. . . .00.. 2.... . I 0.0 ..‘........0 0 . . . . .0 .. 0 .w ... 2.7.... ..rl . _ ....L: . . .. ..... .‘. .. 0 0:0 .... ... . ...0 0» ..0 o . . ..A . . .. . .r. 0-. . . .P. .A . v .03... . ... . . n.. 0. c. 0 .. K 0 . .... ..‘0. . . . 0.0 . I... 0.3!: 0 .. ... U .0 «.0 .0lfi '0 ... .... . o. 0 ..Q. .. ....Q ...00 . .. .I .. .. ..0 0. 00 ... I. . . ..., .. ...P... . ... 0 ... a. . . o . .10.... .. . ... . . . . A Z... . . I . . . ...... . . . . . . 00.. .. . .A. . . 0 A... . . .. .0 A 0 . 0 . ... f . .. 0.. . . .... I . 1.. . . . 0 ... . . . .. .. I . .. c .' 0.0. .— . '0 .. .... ... .1.... . . . . ..0.. .. .. ... .. - ... 0 ... .0. . .. 0\ . c . . 0.0.. v . ...... .v. .... .. ... . . .. .. . .... .0 .. . . . . . A I u a. I . 0 0 0.. . . .... .0 n .V. . ... .. 0f... 0‘ . ... ... ..‘g100.,. . . . . ... 1 . .....0.‘ . - .. , . 9.10.0... 9' 7... . 0. , 0 n . 00 . . .. . 0 .0 0 0 . L . .o .. ... 0. ...?‘0 . . . .. a... 9. 0 n 0.. .0 ..N‘. . ...‘ .. . 0 . 0.... . 0. . 0 . .. 4 . . ... 00“) :00 all-.000 . . . ink-0'. 01000.0... “‘3‘. .u .. ..l. .0.....00..0! ‘4‘. 5.. . ,. n J t . ...}...IPJNN52 ...i) .0001 ”A. .39. \f\: ..c.. L; I 0. .. . ‘ .00 z. 101." ’. av. .. 9%.: K. .0 O V 5d} ._ v. 0.0 . 0 '0... .0"... 0. I . ‘I .w y d ..w a. v 0. FO+ (J. 5‘ J 0.. q— A / .... A 3 .... . ... ,.. . . . .3 .....NH. 1 . .A . 0 ... 0 0 . #2....11 I...¢.0M..} . . ”3;; on. . . ... A 0 «‘19... v I .. . A .00 .. it... .. .- 9.... ... ...." L. . .... .1... :. ... - . . 00 1 .. .. ... -.. . 0. . 10;. ... . . .2 ... . 0“}: a 3 $43.“..z. ‘ «.3.... A . A An Al . .. . . ZI . . ......1... . u .. . 1.. 4 .P" ... t. . r - .1 _ . .. . A . .I. .ll'. . 0‘ ..‘0 0 .u 1.. 0. .0 .34. ..0‘ .5. 9.0.0‘.r 0+ 0J0...” .. t . %. h . 00 0 .9 . 0:... .. lull. II I. ~.. . 0 .0. 4 ..0n .....5 0004 0. 60. . J . ’I .¢ . I V v. . . . .. .... .. . . .... . ... 1...... .. ... ... . .... .. _ ' ll . .... 0.‘ -..... . «... .. .... . . I! . A a ‘8 I'll . 0 .. 0 .. L .. .121... 0. ...-0:0 I J , . . . . .0'§.~. 0: 19 0‘ a. . . . I . A'.‘0 . .190‘. . I !'| .. .10... I A This is to certify that the thesis entitled NETWORK PLANNING AND DIMENSIONING PROCESS IN WCDMA presented by OSAMA MUSTAFA ABUSAID has been accepted towards fulfillment of the requirements for the Master’s degree in Electrical Engineering . “x Major Professomiafire :DfiéaxL/(r )éi 2,0 ) 0 Date MSU is an Aflirmative Action/Equal Opportunity Employer eflLlBRAR—Yme Michigan State University o~-.-.-.—.-—-—.-.-.—._._ - V PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE WWI 43 $13?“ Fa 5/08 K:lProi/Acc&Pres/ClRC/DaleDue.indd NETWORK PLANNING AND DIMENSIONING PROCESS IN WCDMA By Osama Mustafa Abusaid A THESIS Submitted to Michigan State University In partial fulfillment of the requirements For the degree of ‘ Master of Science Electrical Engineering 2010 ABSTRACT NETWORK PLANNING AND DIMENSIONING PROCESS IN WCDMA By Osama Mustafa Abusaid An overview of the mobile system/network architecture with Wideband Code Division Multiplexing Access (WCDMA) air interface is highlighted. The main network parameters are first introduced. The WCDMA network planning and dimensioning process is then described, with emphasis on the link budget, coverage and capacity calculations for a specific quality of service (Q08). A cooperative graphical and algorithmic technique is developed to efficiently compute the optimal network load factor and cell radius. The optimum solution of coverage and capacity calculations result in the optimal cell size, and thus the minimum number of Base stations (BTs) required for a given geographical region. Then, applying a modified (energy-based) Self Organization Map (SOM) algorithm, an optimization for BTs positions is achieved as a computational solution to the dimensioning process. As a case study of scale, we apply the developed WCDMA network planning and dimensioning process for a specific quality of service to a test regional area, namely, the Janzour region with an area of 17.38 km2 and 100,000 in population. The introduced process is implemented for three different services using software environment developed in MatLab. The promising results provide an example of the applicability of the developed planning and dimensioning process. ACKNOWLEDGMENTS I would like to express my sincere gratitude to Dr. F athi Salem, my advisor for his guidance and encouragement throughout the years of my graduate studies. I also want to thank Dr. Ren Jain, and Dr. Tongtong Li for their kindness to be my committee and their valuable comments and suggestions on this work. Finally, I wish to dedicate this dissertation to my parents for getting me started right and continued support throughout my studies, my sisters and my friends. iii TABLE OF CONTENTS LIST OF TABLES ......................................................................... v LIST OF FIGURES ........................................................................ vi 1. Introduction ............................................................................ 1 1.1 Back ground ................................................................. 1 1.1.1 History ................................................................ 3 1.1.2 UMTS Definition .................................................... 5 1.2 The Important of Network Planning ..................................... 7 1.3 Objectives of this Project .................................................. 7 1.4 Organization of the project ................................................ 8 II. Overview ................................................................................ 10 2.1 Growth of WCDMA Systems ............................................. 10 2.1.1 The Architecture of UMTS network with WCDMA air-Interface. 11 2.1.2 Core Network ......................................................... 12 2.1.2.1 Mobile Services Switching Center (MSC) ............. 12 2.1.2.2 Home Location Register .................................. 13 2.1.2.3 Visitor Location Requestor .............................. 14 2.1.2.4 Serving GPRS support Node .............................. 14 2.1.2.5 Gate way GPRS support Node ........................... 14 2.1.2.6 GPRS Register (GR) ....................................... 15 2.1.3 UMTS terrestrial Radio Access Network .......................... 15 2.2 Network Parameters ........................................................ 18 2.2.1 Traffic capacity in cellular CDMA Network ..................... 18 2.2.2 Types of services ............... .. ..................................... 23 2.2.3 Handover in UMTS ................................................. 24 2.2.4 Power Control ........................................................ 27 111. Network Dimensioning Process .................................................. 32 3.1 Introduction ................................................................. 32 3.1.1 Network Planning Terminology .................................... 34 3.1.2 Dimensioning ......................................................... 35 3.2 WCDMA Parameters in Link Budgets .................................. 36 3.2.1 Coverage Calculation steps ......................................... 39 3.2.2 Load Factor ........................................................... 42 3.2.1.1 Uplink Load Factor ......................................... 43 iv 3.2.1.2 Down Link Load Factor .................................... 48 3.2.1.2.1 Down Link common channel .................. 50 3.3 Cell capacity calculation ................................................... 52 3.4 Simulation Procedure ........................................................ 57 3.4.1 Simulation Procedure by Controlling the number of users on the cell 57 3.4.2 Simulation Procedure by controlling both the Radios and Load Factor .................................................................. 58 IV. WCDMA Radio Network Flaming Process ..................................... 72 4.1 Introduction ................................................................... 72 4.2 WCDMA Radio Network Planning Problems ............................ 74 4.3 Self Organizing Map Process ................................................. 76 4.3.1 How SOM is a competitive process .................................. 76 4.3.2 The cooperative in SOM Process ................................... 77 4.3.3 Adaptive Process ...................................................... 81 4.4 The Self Organizing Map (SOM) algorithm ............................. 83 V. Case Study .............................................................................. 87 5.1 Introduction ................................................................... 87 5.1.1 Area under Study ...................................................... 90 5.1.2 Area Classification ...................................................... 91 5.2 Design and Calculation ....................................................... 93 5.2.1 Design and Calculation of soft Blocking ........................... 93 5.2.2 Design and Calculation for Dimensioning Process ............... 98 5.3 Base Station Position Planning and optimization ........................ 111 5.4 Design and Calculation for RNC and Microwave Links ............... l 16 VI. Results and Comparisons ........................................................... 117 6.1 Results of Planning Process of WCDMA Network ..................... 118 6.1.1 General results ......................................................... 118 6.1.2 Results related to Janzour area ...................................... 119 6.2 Comparison between different scenarios by definitions of the Radio Frequency Link budget Parameters ....................................... 121 6.2.1 Propagation Related Parameters .................................... 121 6.3 Comparison between different different scenarios by definitions of Quality of service parameters ............................................. 125 VII. Conclusion ........................................................................... 120 7.1 Conclusions .................................................................. 127 7.2 Future Trends ................................................................ 129 LIST OF TABLES 3.] Typical powers for the downlink common channels ............... 5.] Soft capacity parameters for.12.2kbps ................................. 5.2 Obtained capacity results for 12.2 kbps ............................... 5.3 Soft capacity parameters for 144 kbps ................................. 5.4 Obtained Capacity results for 144kbps ................................. 5.5 Soft capacity parameters for 384 kbps ................................. 5.6 Obtained capacity results for 384 kbps ................................. 5.7 Link budget parameter for 12.2kbps voice service with 120km/h speed 5.8 Hata propagation model parameters for 12.2 kbps service ........... 5.9 Link budget parameter for 144kbps real time data service with 3km/h, indoor user covered by outdoor base station ................................. 5.10 Hata propagation model parameters for 144kbps service ............ 5.11 Link budget parameter for 384kbps non real data time ............... 5.12 Hata propagation model parameters for 384kbps service 5.13 Values of speed of service & number of BTs ........................... 5.14 Relation between number of subscriber & number of BTs for different 6.1 Different values for Eb/NO for different speed of services ............ vi 5 1 93 94 95 96 97 98 1 00 101 103 104 105 106 107 108 124 2.1 2.2 2.3 LIST OF FIGURES Architecture ofthe access plan Interference in radio network planning Simplified calculation of traffic capacity without inter-cell interference 2.4 Generic range non-voice applications 2.5 2.6 3.1 3.2 3.3 3.4 3.5 Generic range non-voice applications Hand Over process Cell range calculation Main steps Coverage-based dimensioning .................................... Relation of coverage, capacity and load factor nu; ......................... Relation of coverage, capacity and load factor nu] for different of services Simulation diagram Procedure by controlling both radios & load factor of the cell ........................................................................ 4.1 Gaussian Neighborhood function 5.] Satellite image ofJanzourtown 5.2 5.3 5.4 5.5 5.6 Relation of speed of service and number of BTs ............................ Shows suggested population distribution for Janzour area ................. Show the random distribution for initial position of BTs in the area The optimized position for the BTs and the effected users by distribution The final optimized position for the BTs and the effected users by Distribution ...................................................................... vii 12 19 20 22 23 25 41 58 64 66 7O 79 91 108 109 113 114 115 Chapter I Introduction 1.1. Back ground: The digitization of our private and working lives is one of the most radical changes that have taken place in recent times. Mobile telephony and the intemet are lead technologies symbolizing a massive, epoch-making transformation. The mobile communication wave has engulfed us, gaining a level of acceptance and spreading at a pace that has exceeded even the most optimistic predictions. And things are only just beginning. While in some parts of the world mobile phone ownership is already close to saturation point. It will be possible for mobile phones, handsets, and mobile terminals of the UMTS generation to be kept permanently ready for operation - “always on”. It will be easier to form virtual communities (“communities of interest”). Personalized service designing will be possible. Each user’s familiar service environment will also become mobile; this means providing a mix of services tailored to individual interests and needs (Virtual Home Environment, Virtual Office Environment). The services will be especially capable of offering hints and tips on a proactive basis in relation to location, time, occasion, event, or activity. Messages can then reach users on a context-related (or action-oriented) basis. The need for high speed data and the quality service demanding completely clear, as your wireless mobile set controls everything in your life and become reliable tool in any application like: -Your mobile set now can do lock and unlock for the car, check the battery level of your car and the general condition of the car. - Now you can control a lot of things in your home like alarm alert, electricity, TV and a lot of other things. Mainly WCDMA network systems are designed for multimedia communication. But, with them personto-person communication can be enhanced with high quality images and video, and access to information and services on public and private networks will be enhanced by the higher data rates and new flexible communication capabilities of third generation systems. This, together with the continuing evolution of the second generation systems, will create new business opportunities not only for manufacturers and operators, but also for the providers of content and applications using these networks. In the standardization forums, WCDMA technology has emerged as the most widely adopted third generation air interface. lts specification has been created in 3GPP (the 3rd Generation Partnership Project), which is thejoint standardization project of the standardization bodies from Europe, Japan, Korea, the USA and China. Within 3GPP, WCDMA is called UTRA (Universal Terrestrial Radio Access) F DD (Frequency Division Duplex) and TDD (Time Division Duplex), the name WCDMA being used to cover both F DD and TDD operation. Forth generation systems are currently under development and will be implemented somewhere in 2008-2015. They will not only connect people to people, but machines or houses to people. 1.1.1 History: The rapid growth in traffic volume and increase in new services has begun to change the configuration and structure of wireless networks. Thus, future mobile communications systems will be distinguished by high integration of services, flexibility and higher throughput. To support such features, the efficient use of spectrum and optimum management of radio resources will be essential. To meet these challenges standardization bodies like ETSl (now expanded to 3GPP), have selected the Wideband Code Division Multiple Access (WCDMA) and the hybrid Time Division — CDMA as the radio techniques for the Universal Mobile Telecommunication Systems (U MTS). Hence, UMTS conceived at the eve of this new millennium will without doubt have a large impact on future Wideband mobile networks and serve as the leading platform for wireless multimedia communications. [1 ll Exactly the UMTS project was initiated by ETSI (European Telecommunications Standards Institute) under its RACE (Research & Development in Advanced Communications technologies in Europe) workgroup as a 3G wireless system. The lntemational Telecommunications Union (lTU) proposed UMTS in 1985. This study, called lntemational Mobile Telecommunications-2000 (lMT- 2000), planned lTU's vision beyond the year 2000. In 1992, the World Radio Congress (WARC'92) allocated the fi'equency bands of 1885-2050 MHz and 2110-2200 MHz for lMT- 2000. [MT-2000 was basically the design architecture and authentication of various competing coding and modulation technologies 3G networks would use. 1.1.2 UMTS Definition UMTS is the term used in Europe for 3rd generation networks, and it is intended to make the transition from 2nd generation networks smoother, but eventually replace them. This means UMTS will, in the long-tenn, support all applications currently served by 2nd generation cellular systems such as GSM and PDC, cordless systems like DECT, and satellite systems like lRIDlUM. It will converge contents from the telecom industry like video telephony, IT industry like lntemet applications, as well as from the broadcasting industry like video on demand. UMTS will have to support this wide array of services with data rates ranging from 8 kbit/s to 2,048 Mbit/s (later even higher) regardless of location, network, or terminal (adaptive terminals). The costs of such terminals could be kept low by being compensated from a mass market generating huge volumes of data traffic. The SMG decided to make UMTS backward compatible to GSM in the beginning, but also to upgrade GSM beyond its initial capabilities. UMTS development would therefore, not be compromised by too many compatibility issues. In order to achieve that goal, UMTS is being developed in a modular Way. This also reduces the risk for operators and allows the creation of a consumer basis for mobile data, which will eventually drive the full deployment of the system as its demand for improved services increases.[l 8] UMTS/UTRA is providing much greater data-transmission rates than the current 2G networks using GSM. The rates are as follows: - 2.048 Mbps for low-mobility outdoor applications, particularly in the pico and micro cellular environments. - 384 kbps for limited-mobility outdoor applications in the micro and small macro cells (in urban/ suburban areas). - 144 kbps and 64 kbps for full outdoor mobility applications (large macro cell applications). - 14.4 kbps for continuous low speed data applications in very large cells. - 12.2 kbps for speech (4.75 kbps - 12.2 kbps). - 9.6 kbps globally (satellite). 1.2 The importance of Network planning Network planning is a major task for operators. It is time consuming, labor-intensive, and expensive. Moreover, it is a never-ending process. which forces a new round of work with each step in the network's evolution and growth. Sometimes extra capacity is needed temporarily in a certain place. especially during telecommunications conferences. and network planning is needed to boost the local capacity. Changes in the network are also needed with changes in the environment: A large new building can change the multipath environment, and a new shopping center can demand new cell sites. and a new highway can create new hotspots. The quality of the network-planning process has a direct influence on the operator‘s profits. Poor planning results in a configuration in which some places are awash in unused or underused capacity and some areas may suffer from blocked calls because of the lack of adequate capacity. The income flow will be smaller than it could be, some customers will be unhappy, and expensive equipment will possibly be bought unnecessarily. For the WCDMA radio access network the problems of dimensioning/planning and algorithm- evaluation/-optimization is a challenging task, which requires methods and tools, that are essentially different from the ones applied for 2G systems. In order to generate reliable results for the highly flexible and dynamic WCDMA network, sophisticated simulation tools have to be applied. WCDMA cellular networks are being rapidly deployed and expanding with the changes happen on the structure. But, the change of WCDMA network happen without proper planning, a WCDMA radio network can be neither successfully deployed, nor be successfully expanded. Changing the structure of the region and/or type of users on it, demands new planning for the WCDMA network. So, while WCDMA radio network is under operation it will go through optimization and development operation to contain the demands and new business models, this is similar to the new planning except we are saying there are some Base Stations BTs fixed. As well planned and optimized WCDMA radio network can provide some 30% extra capacities under the same infrastructure cost. The most noticeable thing is network planning and optimization plays vital role for the deployment and maintenance of WCDMA radio networks. From the other sided the planning process will stay challenging issue for long time, because the WCDMA system have self-interference, “ breathing effect” and many other effects. The coverage capacity of the BTs restrains each other. So the planning process will be needed frequently as the structure of the region change with time to make sure the trade relation between the coverage, capacity and distribution of the users. Technically in WCDMA network planning the coverage, capacity and quality of signal are inter-related, multi rate and mixed signal feed to the same carrier at the same time, that shows the complexity of the technology and how the planning process could handle it easily. For capacity calculations of WCDMA radio interface, due to the possibility of resource allocation for different classes of traffic and sharing the same frequency band, is much more complex. That means all users are going to be served by the same frequency band and the only thing to distinguish them is the orthogonal codes. Things are going to be more complicated when you apply the multipath in to the calculations. 1.3 Objectives of This Project The main objectives of this project are: 0 Develop an understanding of common propagation model used in WCDMA systems. 0 To Develop suitable ways capacity and coverage of UMTS are interrelated and their influencing parameters. 0 To have an idea about the information required to start and implement a system design (customer requirements: 008, coverage; traffic for casts, etc.). 0 To understand how to calculate the coverage and capacity of a WCDMA network. 0 To study UMTS network planning, different parameters and to investigate their effect on the performance of the system. 0 To understand the main operations on the network like power control, sofi handover o Toward implementing an optimization algorithm for planning position of the base station by using Self Organizing Map technique with applying some modification on it, a deep understanding for SOM from side of view related to Artificial Neural Network. 0 To develop complete program doing the dimensioning and planning process to any type of region as one package. 0 To see the possibility of using some signal process technique for planning, like use different vision of convex sum technique for positioning the base station BTs. 0 To implement WCDMA dimensioning and planning methodology in a specific case study. 1.4 Organization of the project: This project is organized in 7 chapters, with chapter 1 as introduction. Chapter 2 Gives an overview for hall mobile systems, it describes Their network architecture especially for UMTS network, Network parameters for WCDMA like " types of handover, power control, traffic capacity " and type of service. Chapter 3 introduces the principles of WCDMA. Like factors that Influence WCDMA network planning & dimensioning process which explain each parameter in link budgets for WCDMA network & its capacity calculation and the idea of dealing with both of them at the same time. Chapter 4 Talking about Self Organization technique, and what’s make it very Powerful technique for optimizing the planning process for optimization. Gives brief talk about SOM modified algorithm with applying the convex sum to the SOM and apply it to the result of dimensioning process. 8 Chapter 5 presents real case study as an implementation for planning process & dimensioning for WCDMA network where we take Janzour town with area 17.38 kmz, with 100 000 population as a suburban area as a case study. Chapter 6 Gives the results obtained and a comparison between parameters obtained from different cases of the implementation. Finally chapter 7 gives conclusions and recommendations of the project. Chapter 11 Overview 2.1. Growth of WCDMA Systems: In the 2002 as GSM was commercially launched, ETSI had already started the standardization work for the next-generation mobile telecommunications network. This new system was called the Universal Mobile Telecommunications System (UMTS). The work was done in ETSI‘s technical committee Special Mobile Group (SMG). SMG was further divided into subgroups SMGl—SMG l 2 with each subgroup specializing in certain aspects of the system. [3] Practically and based on the map of frequency bands of different wireless networks. But There have been (and still are) several competing proposals for a global standard which may show more advantages than WC DMA like advanced TDMA, hybrid CDMA/TDMA, and orthogonal frequency division multiplexing (OF DM). But we will speak with some explanations for WCDMA: The bandwidth ofa WCDMA system is 5 MHz or more, and this 5 MHz is also the nominal bandwidth of all WCDMA proposals. This band width is enough to provide data rates of 144 and 384 Kbps (these were 30 targets), and even 2 Mbps in good conditions. 10 The WCDMA radio interface proposals can be divided into two groups: network synchronous and network asynchronous. In synchronous network all base stations are time synchronized to each other. These results toward more efficient radio interface but require more expensive hardware in base stations. For example, it could be possible to achieve synchronization with the use of Global Positioning System (GPS) receivers in all base stations, although this is not as simple as it sounds. GPS receivers are not very useful in high-block city centers (many blind spots) or indoors. 2.1.1 The Architecture of UMTS network with WCDMA air- Interface: UMTS networks support both types of switching, in each case with special nodes in the Core Network CN. Here the nodes necessary for circuit-switched transmission are shown in the top-right comer of Figure 2.1. This part of the network is heavily based on the existing GSM networks. The nodes used for packet- switched transmission appear underneath. These nodes have already been introduced with GPRS into the GSM architecture, even if some of the protocols are different. Both parts of the core network CN use the same radio access network. The interface between core network CN and Radio Area Network RAN (lu-interface) is divided into the interface for the circuit switched part of the CN (lu-CS-interface) and the interface for the packet switched part (lu-PS-interface). ll ...... Figure2.l Architecture of the access plan The RAN contains two types of nodes: the Radio Network Controller (RNC), which controls resource management in one or more base stations (Node B). Node B in turn supplies one or more radio cells. The interface between RNC and Node B is called the lut-interface. A new feature, and not available in this form with GSM, is the direct connection of RNCs over the Iur-interface. The UE is connected with Node B over the Uu-interface. [1 1] Compared with the previous generation the major changes are in the radio access network (RAN) with the introduction of code division multiple access (CDMA) technology for the air interface, referred to as Wideband CDMA (WCDMA), and asynchronous transfer mode (ATM) as a transport in the transmission part. These 12 changes have been introduced principally to support the transport of voice, video and data services on the same network. 2.1.2 Core Network: 2.1.2.1 Mobile Services Switching Centre (MSC) The Mobile Services Switching Centre (MSC) is a switching node that supports circuit-switched connections. In addition to its switching tasks, an MSC must also support user mobility. If a user moves area while maintaining a connection, the MSC forwards the connection over the appropriate RNCs and Node 33 to the location area of the user (Handover). In addition, the MSC stores (in attached databases) the current location area of the user so that a connection can be set up in the right cell in the event of an incoming call (location management). The MSC also participates in the mechanisms for user authentication as well as in the encryption of user data. The Gateway Mobile Services Switching Centre (GMSC), which also offers interfaces to various external networks, e.g., the Integrated Services Digital Network (ISDN), is a special variant of an MSC. The MSC is the central element of the circuit-switched part of the CN. [1 1] 2.1.2.2 Home Location Register (HLR): The HLR contains the permanent subscriber data register, which sounds like the personal file of the subscriber. The HLR can be implemented in the same equipment as the MSCN LR, but the usual arrangement is to have the MSC/VLR as one unit, and the HLR/'AuC/EIR combination as another unit. [I l] 13 2.1.2.3 Visitor Location Register (VLR): The Visitor Location Register (VLR) is a database similar to the HLR and stores a local copy of the data from the HLR. However, you can name the data at VLR as temporary data, the data in a VLR is dynamic. As soon as a user changes location area, the information in the VLR is updated. [11] The VLR contains such data that the normal call setup procedures can be handled without consulting the H LR. This is important especially if the user is roaming abroad, and the signaling connection to the home network is expensive. 2.1.2.4 Serving GPRS Support Node (SGSN): The Serving GPRS Support Node (SGSN) carries out tasks for packet-switched transmission similar to those of the MSC and VLR nodes in the circuit switched part. The current position of a user is stored in the SGSN so that an incoming data packet can be routed to the user. In addition to routing functions, the SGSN also handles authentication and stores a local copy of the user information. 2.1.2.5 Gateway GPRS Support Node (GGSN) The gateways to other packet data networks, such as the Internet, are connected to the Gateway GPRS Support Node (GGSN). Consequently, The GGSN usually incorporates a firewall. Incoming data packets are packed in a special container by the GGSN and forwarded over the GRPS Tunnel Protocol (GTP) protocol to the SGSN. l4 2.1.2.6 GPRS Register (GR): The information required for the operation of a packet-switched transmission is stored in the GR, a database that is part of the HLR. It includes, for example, a user's authorizations for access to the lntemet. 2.1.3 UMTS Terrestrial Radio Access Network: The UTRAN is the new radio access network designed especially for U MTS. Its boundaries are the lu interface to the core network and the Uu interface (radio interface) to user equipment (UE). The UTRAN consists of radio network controllers (RNCs) and Node 85 (base stations). Together. these entities form a radio network subsystem (RNS). 2.1.3.1 Radio Network Controller (RNC) The Radio Network Controller (RNC) is the central node in a radio access network . It takes the place of the Base Station Controller (BSC) familiar from GSM and assumes the management of the resources in all attached cells (channel allocation, handover, and power control). The RNC essentially is responsible for the following: I. Call admission control 2. Radio resource management 3. Radio bearer set-up and release 4. Code allocation 5. Power control 15 6. Packet scheduling 7. Handover 8. SRNS relocation 9. Encryption 10. Protocol conversion 1 1. ATM switching 12. O&M operation and maintenance. [11] 2.1.3.2 Node B: The name Node B is an unfortunate choice: During standardization this name was planned as a temporary solution until the introduction of a more appropriate term. However, the name stuck nevertheless during the course of the standardization activities and therefore the base station in UMTS is called Node B. This node corresponds to the Base Transceiver Station (BTS) familiar from GSM. The tasks directly connected to the radio interface are handled in the BTS. The inputs comes from the RNC. A Node B can manage one or several cells and is connected with the RNC over the Iu—interface. Node B is the counterpart of BTS in GSM. It supplies one or several cells. Along with the antenna system, Node B includes a CDMA receiver that converts the signals of the radio interface into a data stream and then forwards it to the RNC over the lub-interface. In the opposite direction the CDMA transmitter prepares incoming data for transport over the radio interface and routes it to the power amplifier. There are three types of Node B corresponding to the two UTRA modes: UTRA-FDD Node B, UTRA-TDD Node B and Dual-mode Node B. which can use both UTRA modes simultaneously. l6 Currently, the Node B is linked over an ATM link to the RNC. Due to the possible large distance between Node B and RNC and the length of the processing times, certain particularly time-critical tasks cannot be stored in the RNC: this includes Inner Loop Power Control that in a C DMA network ensures that all users receive at the same signal strength. The RNC has to have as exact a picture as possible of the current situation in a cell so that it can make sensible decisions on handover, power control and call admission control. Consequently, mobile stations and Node B periodically carry out measurements of the connection quality and interference levels and transmit the results to the RNC. In the special case of softer handover, the splitting and combining of data streams of the various sectors are also already handled in Node B.[l 1] 2.1.3.3 User Equipment (UE): The last important network node is the user terminal (U E). This equipment can support one or more radio standards and contains the USIM. It is simultaneously the counterpart to Node B, RNC and the CN Like Node B, the UE is responsible for processing the radio signal. This compute-intensive task comprises error correction, spreading and signal modulation as well as radio processing up to the power amplifier. On command of the RAN the mobile station must adapt to the transmitter power. [12] As a counterpart to the RNC, the mobile station participates in the signaling for connection set-up and release as well as in the execution of handovers. For this purpose it measures the received field strength of neighboring cells and transmits the 17 measured values to the RNC. The encryption and decryption of communication also take place with the RNC in the UE. This long list of tasks handled in a U5 is accompanied by the users' request for larger displays that also support the decoding of video data. It is also planned that a camera with an accompanying MPEG codec will be integrated into the equipment. Since a terminal should also enable the playing of games, an efficient processor with substantial memory is to be used. The most important thing, however, is that the equipment should remain small and easy to handle. The complexity of UE is very high and it will take a big attention on the part of the developers involved to accommodate all the requirements and still develop attractive tenninals.[1 l] 2.2 Network parameters: 2.2.1 Traffic capacity in cellular CDMA networks: Since the size of a cell depends on the traffic in the network, the location planning for a WCDMA network is always only optimal for a certain volume of traffic. When traffic volumes are low, large areas of overlapping occur; with very low traffic volumes the cells can shrink to such an extent that gaps occur in the radio coverage. Furthermore, the cells breathe differently for services that require different C/I ratios. The planning of a CDMA network must therefore not only take into account a certain traffic volume but also a certain service or service mix and the respective quality of service (see Figure 2.2). So the gross transmission rate is reverse proportional to the spreading factor. The required carrier-to-interference ratio at the receiver therefore increases the higher the transmission rate of the respective service. A higher required C/l ratio also means a 18 smaller maximum range of radio coverage. Radio coverage for services with a high spreading factor is therefore higher than for services with a low spreading factor. Depending on the network planning, such high—rate services can only be used in the proximity of the base station, whereas services with a lower transmission rate, such as voice services, are available in the entire cell. [ Radio Network Planning ] Due to the cell-breathing the cell size depends on the amount and on the type of the traffic signal Figure2.2 Interference in radio network planning The capacity of a cell in a CDMA network essentially depends on the orthogonality and number of spreading codes used. Perfectly orthogonal codes guarantee that the different physical channels do not cause mutual interference. In this case, the traffic capacity of a CDMA cell is only determined by the number of mutually orthogonal codes. 19 In normal cases where only a scrambling code is used on the downlink in UMTS FDD mode, the theoretical capacity is determined by the number of available Orthogonal Vectors spreading codes. If the spreading codes are only quasi-orthogonal, as is the case on the uplink of UMTS F DD mode, then cell capacity is determined by the interference this creates. The transmitters of the mobile stations are controlled by the power control in such a way that in ideal circumstances the signals received at the base station all have the same power, thus preventing near-far effects. Of n transmitters in a cell, one of them thus always supplies the wanted signal; the remaining n — 1 contributes to the interference power. As Figure 2.3 shows, the maximum user capacity per cell therefore now only depends on the minimum required C/I ratio. Cell Capacity FDD UL: quasi—orthogonal codes, Single-user detector, Perfect power control C/l=1*C/(n~1)*C o . . uh-u-n. :.;‘~';'r- -.a..‘.- no" - . » . . ' - --- . ."' - u. . .‘v' Hu- . «....r'nr-un -- - x..'---» - .' - . «~17. ' -.-.. .' u... ;' - . ---.-'.: ----....' ---.. _-...._ .-"- u. ' ---.. . . _ . ......,.,_._ u- ...-.... 5...... ,~_- .. ..; ,- ~ ' .... -~--- . ..:' ' - . .,;'---- - . .' ' - ...,_-~-.. .,:"~x- . .'-- - . :‘u -.. :'-- - - --- . .,;*-- -.: ."--~- .." ~.... -'~- ..' ;' “n.4," - - .'---- ..;'u.- . :----m..«.,t‘:?s'i':-;--XT.::.‘.‘.-.o - . - ;.‘ ....:,:,.:-‘-_._.... v...;r:';1- - :‘z-g'g ~ 3:5... . :';- ....,-_-‘....;'.',- ..::;:,-,- .':':--_v-'.'..f.‘:---..:':-_~- :,:::?.v--‘.:.-',--.-‘2:t:~;-I- ;’:-_-‘.':::-A :t:-;-...:::-;8..f Figure2.3 Simplified calculation of traffic capacity without inter-cell interference 20 In CDMA networks the interference power is a factor that determines the network capacity. CDMA networks are therefore also referred to as being interference- limited The connection between user numbers and C/I is illustrated in Figure 2.5 the CV] at the receiver is spread over the number n of active senders in a cell. The more senders that are active in a cell, the higher is the interference power at the receiver and the smaller the C/I ratio. The C/I ratio required depends on various factors. What is particularly important for determining the required C/I ratio are the spreading factor, the channel coding and modulation as well as the tolerable residual bit error ratio. It is obvious from Figure 2.4 that with a required 01 of -20dB around 100 users can be active in an individual cell or that with 100 users only a C/I ratio of a maximum of -20 dB can be achieved. The calculation of cell capacity in a multicell mobile radio network has to take into account the intracell interference as well as the intercell interference and thermal noise (see Figure 2.5). In UMTS the spreading codes of different cells are only quasi-orthogonal to one another. All senders in co-channel cells thus produce intercell interference. The interference shown on the uplink in Figure 2.5 originates in the 21 Multi-Cell Capacity C/'l=1“'C/(n-1)*C+linter(n)+noise . . :‘t' . .. . .... .... .-£i , , ........... , , . . , , . , _ . , , . . ......... . ..... ................. . ""'--cu¢.-.-.:.. . ....... -- . _,._ ........ . , . .-_~ . . . . . ; _‘ - . . .'.- - . . . . ., . _'- , _ _- ~--....,_, ' :--.ou-.. “h ‘ulco-v.'.a.. ..... 31‘- Con-...... _ . - - 0 ' - - - ................. ~ ~ .. - - . . - . . . - . 9 .-.';-.-ext-J.'.:;7:::-:"mamas:'r-r-m::.?.7.'..::§:::-r::---‘=-=» Figure2.4 Calculation of traffic capacity in a multicell CDMA network transmitters of the mobile stations. In a multicell situation the equation given can no longer simply be resolved based on the number n of active participants, because the inter-cell interference also depends on the number of participants in the co-channel cells. However, the cell capacity can also be calculated for a multicell situation. It is easy to see that the cell capacity for a single-cell network is always larger than the cell capacity for a multicell one. Due to thermal noise and intercell interference, the cell capacity in multicell networks is only about half that of single-cell networks that having no intercell interference or thermal noise. [I l] 22 2.2.2 Types of services: The strength of UMTS services will not reside in one or two applications, but in the conjunction and complementation of a series of application and technologies, which will generate different sets of services. Figure 2.3 illustrates a generic set of application targets. Primarily for PS networks including multimedia features. In this illustration, we can see the characteristics of connectionless and connection oriented services, i.e. variable and constant bit rate. Man agement Directory Picture leading Messaging Figure 2.5 Generic range non-voice applications. 23 We not only need to know to what groups we can address these services (e.g. enterprises, communication firms, telematic centers, content and location based providers, commerce organizations, and typical wireless operators aiming to minimize operational costs and charge). We also need to know where the end user is and how does he/she applies technology. To meet the needs implies making available the correct tools and environment. Now, if we assume that the infrastructure arrangements will take care of the environment, it remains a big task to find a tool or user equipment device to satisfy users. A terminal not only needs to be a smart device capable of accessing a PS network, support bandwidth on demand, audio streaming, multimedia, it will also need versatility and have multiple capabilities. A multi-functional device will make the difference in future usage and acceptance of higher transmission rates offered through UMTS. Market penetration and widespread usage of these of multimedia services will depend on the available and affordable terminals, as well as the pragmatic applications. Wireless device interconnections, intelligent voice recognition, wireless e-mail, simultaneous voice and data, user defined closed user group, location services personal profile portal, location based delivery and marketing will only occur with efficient integration and inter-working of multiple technologies. 2.2.3. Handover in UMTS: The term handover always relates to services that are operated on a circuit-switched basis. Packet—switched services use a different technique, which will be explained later. In UMTS there are mainly three different types of handover. The simple understanding of hand over procedure is shown at Figure Active set = Node 81 & Node 82 Node 82 SS > add threshold Active set = Node 82 Node 81 SS < drop threshold Figure2.6 Hand Over process The main types of handover are: 1. With hard handover already familiar from GS M, a connection is switched hard at a particular time. This method is also used in UTRATDD- mode, because sufficient time is available between the individual transmitters and receivers phases in a mobile station to switch to a new cell. The changeover to the new cell thus occurs from one frame to the next one. 2. A sofi handover is when a mobile station communicates simultaneously with up to three sectors from different Node-Bs. The data is split up in the RNC (splitting), broadcast over the Node Bs and combined again in the mobile station. Data from all 25 participating Node B5 is received on the uplink and forwarded to the RNC. The RNC combines the two data streams again and transfers the data to the CN. This technique is also called macro-diversity and offers several advantages: a) The connection becomes more resistant to shadowing due to the reduced probability that, considering all supplying base stations, the mobile station will end up in shadowing. If an interference object cuts off a connection to a base station in a soft handover, there is the possibility the connection will function over the second station and the communication will not be cut off. b) When the minimum received power is calculated, a small reserve against fast fading through multipath propagation can be incorporated. Since the drop in received power through multipath propagation is almost static in the case of static transmitters and can amount to up to 30 dB, it is possible that a static mobile station will not be supplied adequately. Soft handover offers the option of transmitting data over the second Node B and thus maintaining the communication. c) Furthermore, a soft handover offers the possibility of reducing the near-far effect. 3. Sofler handover is a special version of soft handover in that transmission can also run in parallel over different sectors of the same Node B. The advantages mentioned in conjunction with soft handover also apply to softer handover, although the Node B can already be entrusted with the task of combining the two data streams and only transferring one data stream to the RNC. 26 2.2.3.1 Handover types in UMTS: A distinction is made between various handover types depending on which network elements participate in the handover: 1. Within a Node B a connection can be switched hard or soft from one sector to the next (Intra-Node B, Intra-RNC Handover). 2. Between different Node B3 of the same RNC a connection can be hard switched (Inter-Node B, Intra—RNC-Handover). 3. A connection can be soft switched between the different Node 35 of the same RNC (Inter-Node B, Intra-RNC-Soft-Handover). In this case, the RNC is responsible for the combining/splitting. 4. A connection can also be forwarded between RNCs within the RAN over the lur- interface. This is the case in which SRNS and DRNS are involved (internal Inter- RNC handover). The term internal handover relates to the fact that the handover is carried out totally within the RAN. 5. Switching with relocation of the lu-reference point is called SRNS relocation. This is an external Inter-RNC-Handover. 2.2.4 Power control Power control in a WCDMA system is crucial to its successful operation. This is because each handset transmits on the same frequency and at the same time as other handsets. Each of the handsets therefore generates interference, raising the overall noise level in the cell, and the base station has to be able to distinguish a particular user out of this interference. If a single mobile device is transmitting with too much power, or is physically closer to the BTS, this may drown out the other UEs. Conversely, if a UE is transmitting with too little power, or is physically further 27 away, the base station will never hear it. This is commonly referred to as the near— far problem. There are two main concerns regarding power control: Distance from the base station and fast fading. Within the WCDMA system three types of power regulation are used, open loop, inner loop and outer loop power control. Each of these mechanisms will be dealt with in turn in the following subsections. The key goals of power control are to: 0 provide each UE with sufficient quality regardless of the link condition or distance from the BTS; - compensate for channel degradation such as fast fading and attenuation; - optimize power consumption and hence battery life in the UE. 2.2.4.1 Open loop: This is used when a handset first enters the cell. The handset will monitor the CPICH of the cell and will take received power level measurements of this channel. This information will be used when setting its own power level. The power radiated by the base station will reduce as the distance from the tower increases. Simply having a received power level is not enough information for the mobile device to set its own transmit power. This is because it does not know at this stage at what power level the CPICH was transmitted. The base station sends out power control information on the broadcast channel, which includes an indication of the power level at which the CPICH is being transmitted. The UE can now determine how much power has been lost over the air interface and thus has an indication of its own distance from the base station. The handset can now calculate what power to transmit at; if it has received a weak signal 28 it will transmit a strong signal since it assumes that it must be a long distance from the base station. Conversely, if it receives a strong signal it assumes that it is near to the base station and can thus send a weaker signal. The actual mechanism is that the UE will listen to the broadcast channel. From this, it will find out the following parameters: ° CPICH downlink transmit power 0 Uplink interference The UE will then measure the received power of the pilot (CPICH RSCP). The initial power used is then: Initial power = CPICH downlink transmit power - CPICH, RSCP + UL interference + Constant value The constant value really provides a correction figure for better approximation of an appropriate start power. It should be noted that this is a rather approximate mechanism since the base station will be transmitting in a different frequency band to what the handset will use and the power loss over the air interface may be significantly different for each band. We can summaries the work of power control as the two different mobile devices, depending on their distances from the base station, will send an access request on the RACH at different power levels. This will reduce interference to other mobile devices. The diagram shows separate BCCH and CPICH for the two different mobile devices; in reality only one of each of these is broadcast and all the mobile devices receive it. 29 2.2.4.2 Inner loop Inner loop power control feedback is a form of closed loop power control. Information is sent in every slot, i.e. 1500 times per second. This can be compared with 18-95, which has feedback 800 times a second, and GSM, where it is only carried out approximately twice per second. This type of power control is required because the open loop system can only give a rough estimate and is not accurate enough to deal with problems such as fast fading. To control the power level on the uplink, the base station performs frequent signal to interference ratio (SIR) measurements on the received signals from each of the mobiles. This value is compared to a target SIR value and if the power from the mobile station is deemed to be too high or low it will be told to decrease or increase the power accordingly. Since this task is executed 1500 times per second, it is much faster than power control problems, such as fast fading, that may occur, and hence can compensate for these. This fast power control is very effective for slow to moderate movement speeds of the mobile device. However, benefits decrease as the speed of the mobile increases. This also deals with the near—far problem, where signals from mobile devices which are far from the base station will suffer greater attenuation. The object of the fast power control is that the signal from each mobile should arrive at the BTS at its target SIR value. The same type of power control is used on the downlink. When communicating with mobile devices that are on the edge of the cell the base station may marginally increase the power it sends. This is required since these particular mobiles may suffer from increased other-cell interference. 30 2. 2. 4.2 Outer loop : As noted, inner loop power control measures the power from the mobile device and compares it to a set SIR target. This target value is set and adjusted by the outer loop power control within the RNC. This value will change overtime but does not need to be adjusted at the same high frequency. The target value is actually derived from a target BER or BLER that the service is expected to meet. Some errors with the data received from the mobile device are expected. If there are no errors, then the UE is assumed to be transmitting at too high a power, with the consequences of causing interference and reducing the battery life of the device. To implement this method of power control the mobile device will compute a checksum before sending any data. Once received, a new checksum is computed on the data and this is compared to the one sent by the mobile device. The BTS will also measure the quality of the received data in terms of BER. If too many frames are being received with errors, or frames have too high a BER, then the power can be increased. The target set-point is not static: it does change overtime. This is required so that the cell can be more efficiently utilized. [18] 31 Chapter III Network Dimensioning Process 3.1. Introduction: The Planning and dimensioning process is the initial step on the radio network planning for WCDMA or any other 3G Network. In general we can see three main phases on the network design process: starting with (a) the preparation phase which responsible on setting the principles and collecting the all needed data for the network planning process, followed by (b) the high level network planning phase “ Network Dimensioning”, and (c) the detailed radio —network planning phase. This chapter talking mainly on the Dimensioning Process for WCDMA network which the initial step on the whole radio network planning process and we use its outcome as an input for the final stage of the planning process. The dimensioning process uses the all available parameters of the network to get all amounts of instruments we need on installing the network. The dimensioning must satisfy all the operator’s requirements on capacity, coverage and quality of service, where in this chapter we introduce the knowledge of: both coverage and capacity which are related to each other .So, we have to optimize defining them simultaneously; this will be one of our challenging targets during this chapter and its case study. As the network become under testing mode, we observe its perfonnance by measurements. So, it will be easy to use these measurements to optimize the network performance. But, with intelligent tools, network elements and observation 32 of the changes of all network parameters we can get the optimum values for the network. In a WC DMA network cell, the available data rate depends on the interference level—the closer the UE is to the base station, the higher the data rates that can be provided. Thus, an operator that is aiming to provide 384-Kbps coverage must use more base stations than an operator that is aiming for l44-Kbps coverage and so on. All this is related to the cell shrinking phenomena but not the main explanation of this phenomena that will be explained in details later on next chapters. There are other planning parameters must be mentioned like: the allowed blocking probability, migration aspects (if the operator already has an existing cellular network). the quality of service (008), and so on. If call blocking is allowed with a non—negligible probability, then less capacity needs to be allocated as result, and the network will be cheaper and easier to implement. Now it is clear the WCDMA network is interference limited, where the use of one carrier frequency within all users on all the cells of the network gives the network soft capacity limit (Soft Limit Capacity). The interference limit is mainly effected by the load factor which the main parameter on the dimensioning process and the planning optimization process as well. This load factor parameter is clearly affected the coverage calculation and the capacity calculation as we will see in this chapter. all these make this parameter playing the main role and make all the process of planning depends on it. It is clear now after getting through the introductory chapters there are some preliminary steps that an operator must apply like: acquiring the population and vehicle traffic information from the planned coverage area (the project position). 33 How many people live in an area? How many people work there? What is the vehicle traffic density on main roads during rush hours? Are there any special places that may require lots of capacity at certain times? These could include sports arenas, conference centers. and sites ofpublic festivals. Then the operator must estimate the mobile-phone penetration and the amount of traffic generated by each user. Note that an average business user probably generates more traffic than an average residential user. The business calls will probably be longer. and many of those calls may include data traffic. 3.1.1. Network Planning Terminology So here are some concepts and terms used in network planning. It mainly gives general idea about Erlangs or blocking probability where its purpose is to clarify the required 008 : ° T raffle intensity: is measured in Erlangs. One Erlang is equivalent to one call continued one hour. Thus, the traffic intensity can be calculated from ° [Number of calls (per hour) >< average call duration (in seconds)]/3,600 0 If the result is smaller than 1 Erlang. then quite often the appropriate unit is the mErlang (= 0.001 Erlang). 9 ° Traffic density measures the number ofcalls per square kilometer (Erlang/km ). Traffic density value is only usable for circuit-switched voice calls. For data . . . . 2 servrces, the traffic dcnsrty usually measured usmg Mbps/km . 0 Spectral eficiency is defined as the traffic that can be handled within a certain bandwidth and area: ’7 Traffic intensity (Erlang)/(Bandwidth >< Area) = bps/(MHZ >< km") 34 ° Outage is the probability ofa radio network not fulfilling a specified 008 target. - Cell loading indicates the relative occupancy of the cell. This is given as a percentage of the maximum theoretical capacity. 0 Loading factor defines the amount of interference loaded into the cell by surrounding cells (outer cell interference). This is given as a ratio of the power received by a base station from other cells to the power it receives from mobiles in its own cell. Notice that all power received from outside the home cell is interference. [9] 3.1.2.Dimensioning: WCDMA radio network dimensioning is a process through which possible configurations and the amount of network equipment needed to install the network are estimated, based on the operator’s requirements and through the following calculation. [13] Coverage: _ coverage area; _ area type information ; _ propagation conditions. Capacity: _ spectrum available; _ subscriber grth forecast; _ traffic density information. Quality of Service: __ area location probability (coverage probability); _ blocking probability; 35 _ end user throughput. Dimensioning calculation including radio link budget and coverage analysis, capacity estimation and, finally, estimations on the amount of sites and base station hardware, radio network controllers (RNC), equipment at different interfaces, and core network elements (i.e. Circuit Switched Domain and Packet Switched Domain Core Networks).[l3] 3.2 WCDMA Parameters in link Budgets: Transmissions in different cells of a CDMA cellular system are not divided into groups of different frequency channels as in an FDMA cellular system, but are Wideband and share a common spectral band for all the users. Rather than having the repeating frequency reuse plan of sets I to 7 of narrowband frequency channels and take care about the frequency reuse distance, a CDMA cellular system is designed to use the same set of frequencies in each cell or one carrier frequency at time. The effective size of a CDMA cell, expressed as the radius of a circle or hexagonal, is the maximal distance at which the forward or reverse link can operate reliably, so in the calculation we are working on the less radios which mean the worst case. This distance for a given radio link is the pointer or the indicator for maximal propagation loss that can be tolerated while receiving the signal with sufficient power strength to overcome noise and interference so it can be gotten, as determined by the link's "power budget." It is clear now the effective CDMA cell radius can be completely different on the forward link than it is on the reverse link, where all the parameters of link budget calculation will change completely depending on if the calculation is for up-link or down link. In general, there is more effecting of interference on the upper link (reverse link), that’s make the power of 36 signal sent by mobile user is less than the strength of power can be sent by the base- station and that difference of power level are noticeable as we will see later. Therefore, as will be explained below, an effort is made to "balance" the forward and reverse links by adjusting power and other parameters to make their maximum tolerable propagation losses approximately equal, all this is done under Power control, where in Planning Process we are working with worst case scenario.[9] First there are some WCDMA-specific parameters in the link budget calculation are not included on 2nd G planning process where here we need to main keys for our calculation because in WCDMA the network are based on interference limit not on frequency reuse factor. The most important new parameters effected the link budget calculation on WCDMA networks are as following: * Interference margin: The interference margin is needed in the link budget because the loading of the cell, the main indicator for interference margin is the load factor which affects the coverage. The more loading is allowed in the system, the larger is the interference margin so larger load factor which lead to smaller coverage area. So we have two output cases of Link budget calculation: coverage limited or capacity limited. For coverage-limited cases a smaller interference margin is suggested to get as large as we could, while in capacity-limited cases a larger interference margin should be used to get the maximum capacity. In the coverage-limited cases, the cell size is limited by the maximum allowed path loss in the link budget, and the maximum air interface capacity of the base station site is not used. 37 * Fast fading margin (= power control headroom): It is headroom is needed in the mobile station transmission power for maintaining adequate closed loop fast power control, as we need to apply fast power control to compensate the fast fading. This applies especially to slow-moving pedestrian mobiles where fast power control is able to effectively compensate the fast fading. Typical values for the fast fading margin are 2.0—5.0 dB for slow-moving mobiles. [13] * Soft handover gain: The soft and the hard handover are gives gain works against the slow fading, where the slow fading is meaning no correlation between the base station and the coming signal. So we apply the gain of hand over as reduction on the required log-normal fading margin and by making a handover the mobile can select a better base station. The clear constructed way on WCDMA is the soft handover gives extra gain which we can benefit of it because we can decrease required Eb/No relative to a single radio link, due to the effect of macro diversity combining. Typically the total soft handover gain is assumed to be between 2.0 and 3.0 dB in the examples below, including the gain against slow and fast fading. Other parameters to be specified at this stage include the data rates, mobile speeds. coverage requirements, terrain types, and asymmetry factors. These values can be based on empirical tests or assumptions. [9] Practically. using of fully dynamic simulator that implements the all effecting radio resources in the link budget calculation like Radio Resource Management simulators (RRM) use input data indicate the Power Control, Soft Handover. Packet Scheduling, etc. 38 3.2.1 Coverage Calculation steps: The main calculation step in any wireless network is the link budget calculation or as they named coverage calculation, it is based on the all parameters effecting the signal during its trip from the source to the destination where any obstacles will effect on the negative direction and the gains will effect on positive direction. So, in link budget calculation all the physical distribution of the area should be mentioned in the estimation. Now we know the load factor and the other WCDMA parameters is effecting the link budget calculation. / So from the first phase we have to calculate the maximum path loss from this equation: Lmax = EIRP — receiver sensitivity +GBS R — Lcabe] BS — fast fading gain (3.1) Where: Fast fading gain: it’s gain depending on the power control which depends on speed of the user. - Lcabel BS = Loss per Km * the height of the tower of Base Station (3.2) - G33 R : the gain of receiver antenna at the Node B (Base Station). - Receiver sensitivity = Eb/N0 —- 10 log (chip rate / data rate) + Receiver power noise + interference margin. (3.3) Where 39 03° chip rate = 3.84 Mbps is stander for WCDMA 0: receiver power noise = KTBTF (3.4) Where 23 K: poltizman constant = 1.381 X10“ .l/K T : Temperature in Kelvin BT: transmitted BW F: Noise figure ‘3' interference margin: 10 log (I- load factor for type of link) (3.5) EIRP : Effective Isotropic Radiated Power = maximum radiated power from UE transmitter + UE antenna gain + Body loss (3.6) V For the second phase and fiom the link budgets above, the cell range R can be readily calculated for a known propagation model, for example the Okumura— Hata model or the Walfish-Ikegami model. The propagation model describes the average signal propagation in that environment, and it converts the maximum allowed propagation loss in dB to the maximum cell range in kilometers. So, instead of having the limited power we will have limited distance or maximum distance between the mobile user and the base station. There are different propagation models already established and done for different types of area based and now we have its own constants and ready to use. As an example we can take the Okumura—Hata propagation model for an urban macro cell with base station antenna height of 30 m, mobile antenna height of 1.5 m and carrier frequency of 1950 MHz 40 L = 137.4+ 35.2 loglo (R) (3.7) Where L is the path loss in dB and R is the range in km. For suburban areas we assume an additional area correction factor of 8 dB and obtain the path loss as: L = 129.4 + 35.210g10(R) (3.8) [131 \/ Now after we know R max, we have a scenario that is doable: (l) we can see the area of cell and from all area we got the number of Base Stations (node B) and with the distribution of people we can estimate the number of users in the cell. After that, we use the load factor equation; by trial and error solution we can get the actual R for the cell by changing the estimate value of load factor. Repeat all the steps above to get the suitable load factor as shown in figure (3.1). System Level parameters [Max path loss calculation [dB] j G The probability dwwm Margins to guarantee service ] requ'md “Winge including the shadowigqidai Type of the ...... rPropagatlon model to convert ] region 1 mwer to distance Figure 3.1 Cell range calculation [13] 41 Using all the previous steps we can decide that, the perfect way of design is to design where a transmitting bit rate is the highest, one but this will create large cost, so we can customize it. (2) Clearly, the first way depends on trial and error which may not converge and bring us a suitable value. The other clear way for us now, is to change the meaning of our limit from maximum radios to maximum number of subscribers, With this way, we will be clearer with the meaning of the limit. To get the maximum number of subscribers allowable for certain radios we fellow these steps: I- Get the density of subscribers of the whole area we want to install the network on it by: subtracting the population of the region over the area of the region. 2- Using the maximum distance between the subscriber and the base station, which we got for the link budget calculation as radios of hexagonal cell and find the hexagonal cell area. 3- Take the output of the (density of subscriber x hexagonal cell area) as the maximum number of subscribers the base station can cover based on the link budget calculation. Now we have the maximum limit of coverage on terms of number of the subscribers which is going to help us on the optimization stage after we get the output of the capacity calculation. 3.2.2. Load factors: As clear the importance of the load factor is playing on the WCDMA calculation, where it has the biggest effect is on the link budget calculation, as well as in noise rise and fading margin. So it is very important to see what it means exactly, how can 42 we calculate it or get the optimum value of it, what are the most important parameters effecting it, and how is it effecting the planning process. It is estimating the amount of supported traffic per base station site. To see the meaning of its values: if the frequency reuse of a WCDMA system is 1, the system is typically interference-limited and the amount of interference and delivered cell capacity must thus be estimated. It is clear now that the load factor is the indicator for the interference in front of the receiver and it is different in uplink to the downlink. 3.2.1.1. Uplink Load Factor: Moving toward a full understanding of load factor, we will start to understand the main parameters effecting the load factor, even theoretically. Additionally, spectral efficiency of a WCDMA cell can be calculated from the load equation, whose derivation is shown below. We first define the Eb/No, energy per bit to noise spectral density ratio: Eb Signal power of userj — = j . , (3.9) No 1 Total receiver power where Gj is the processing gain of user j This can be written: No 1 a.Rj ' Itotal-Pj (3.10) 43 Pj is the received signal power from user j, W is the chip rate (= 3840 kbps) for WCDMA network, 0:} is the activity factor of user j where services are not in use 100 % of the time, Rj is the bit rate of user j and [total is the total received Wideband power including thermal noise power in the Base Station. Solving for Pj gives: 1 P- = W .Itotal (3.11) 1 1+[ ] By defining Pj = Lj >< ItotaI, we obtain the UL load factor Lj of one connection: (3.12) L'- 1 1 _ 1+[W/(E_b)j.Rj.o+q,# - . ,Nifi+++¢¢}fl 4* *‘ ++++- t N++¢¢+¥++fii¢+++ ++¢++i 4r +“++++¢+¢¢fi#+f++ ++++++ +‘*+++¢+4+++*++‘+ 4+ ++++++ ++++++¢+3¢u++g+++ ++$*++. **+‘§#****++** +++ -++ r; +~++++~+++=¢++++++++ Figure 5.3 Shows suggested population distribution for Janzour area We applied several modifications to the SOM algorithm to do the optimization: - The first modified Self Organization Map (SOM) algorithm toward signing the suitable positions of the base stations. With having the distribution density of the subscribers in the area based in distribution of the building and knowledge of structure of the area, we could find the best distribution of the base station numbered by the dimensioning process. That can be done by applying modified Self Organization Map algorithm to the data we have related to the positions of the subscribers. [6] Our modified SOM algorithm work with the input data, which are a huge number of two dimension vectors, represents the target users in the area. We use this huge number of vectors as training data to another certain number of two dimension 112 vectors which are presenting the base stations. Initially we randomly initiate the position of BTs as you can see the small blue triangles in the next figure: ‘* * ¢ ***+##*++t+£¢tr+t¢f¥tfr+ist.., ’ f g ‘ +5 fins“ 5* =4é+4§§¢+9+~+w+++§++. 'T ¢ if; + *** ’1‘“? +31“ ' ’i “'4? ;**‘*‘***¢****¢:‘¢*‘_t+4;- *w * ;i*f*‘f*i’ ‘. f*.%w£*** * ##4443‘ . *****h9r. gfibi» '~+++'+t ,-. «Hire. ... ;+«r+ + w» t 4' as; ‘ firtaw ' ‘5‘ 4» +¥*+t+'l rfi++~+¢§ .Q# 1: +41% . ' *1; ' in * ;t+u+' is i . hits ~**+¢# aukooq'r‘ fibre .4; g 4"} # if; t . . +34.» 4» ~¢+¥3‘ ~g+' at fitrrr :tiat-M. #4» *ug;. ‘ths r++¢¢t ‘a¥¢++~;‘¢ ‘ gas -.#+4* _ f+¢+ft4y first»; :-++"M++ - ..4_r,#+ + +£+flhjr+ + + + ‘HH._**.+ f311+++¢¢vifi3++r giiim. w++t++s¢+iwszd++e . . + . rifidtff **:#+r+ie*fit*+f mat?“ a++s+o++¢zfl+ee ++++t *#fi*1*+**#++## Figure 5.4 Show the random distribution for initial position of BTs in the area We take these initial positions of BTs as weights in Modified SOM algorithm, so the process keeps running until the weights diverged. Those weights explain the distribution of the subscriber with less number of vectors as you can see in the Next figure. ”3 i fill - "-. 'U .....- '3 a *‘ Iflu . . . ‘ 1 ,V- 2. 7+ 2 ,‘V‘ . . l.’_ ..I ' 1.9%; q; 4.:- gb _:. " ll "2 N “" ‘ . r, .' ';..'1 Y ' ‘ ‘ ..;. -. -,. 422-. . ”" ' " " . gum.” . :..:.~ - ... . .3; g- a. 1. ‘- ‘ . .. 2|. , 'L . . ‘ 2' ' . ..:l><‘ - , ’ _; 2.... ...“_ . ’2 r .- " " .j‘ft; . '.:_ -:- . , ~ .. . - . .., .- .' ‘..' 2 . - ..‘t _ ; ‘. 5.5:}. ; ,. u - - 'k' ;, .- - 2 . ("i.n" ;. "u .. :. .‘ " ' l ‘ g. .+,. Q J. 1.. 'E '- a- + . ’ '2‘ , . .2 4.; ... o 2 ’2.- .- - r .2- W ~ 1; 3:1,. "3%? 3:22. 222+ 2 “* ‘2f. '_.ic.757‘tt : *Lgilfi .u- 22:2;‘3122t ; .2..2 2ts2.¢::.i;.:r:: ; r233: f&:2 2222.2 1222 4 :224: +++++++++++4+++++Af ++++++ ++++++++2222 ++++'#$++wfitg-a-.d N + ts¥i‘~ #7. + '0 Figure 5.5 The optimized position for the BTS and the effected users by distribution The modified SOM algorithm we used is based on regular SOM decision maker, which is looking for the minimum distance between each input point and certain ' weight. So, the closest weight to the input point will be activated and the neighborhood weights will activated with less amount based on the neighborhood function. But, in our case we are using very small width of neighborhood function up to updating only the activated neuron (the closest one). One of the other modifications we applied to the SOM to get what we want is: The amount of the updating has been done to be the average of the all input points that affected the specified weight. 114 With this we got weights (BTS position) make very clear clustering and it is positioned based on the distribution of the input data in each cluster as shown in the last figure, the plus shape in the figure represents the position of the BTs. - The second modification we applied to SOM algorithm is by changing the update rule of the weights on the original SOM algorithm. Wejust change the power of the difference measure and apply different powers which should be odd. For our case just make it to the power 3 is gives completely better result. So wejust need to work on equation (4.14) by applying odd power to the deference in the equation. With using this kind of modification of SOM to our case we notice it works easier and gives smother optimization for the positions of the base stations. As you can see in figure 5.7. M .‘.__ ‘. .tt . 11:3)“. ._ ; ,. a: + + 't .+ + 4- + + ‘9 - + + $”‘ . ~~ + +T + +' "+1 1“" Q- s . ‘f “f z+ _‘ ' : . ..p +' V + O ~ 4- ‘ ' - «'33:? t +5 -+ +"‘1"‘M§++ + . 34-0-0 7 ' t a. .. » ; '3» “' : . 3 ..+..+r.:‘..+..‘x.fi W ‘ .*. L.... ‘. Q. .},_,..9 a; .t. .13.; + + *4? +*r {f + f 4"“ “f; . ‘l 1' M" 3 s +3 ...... wegtww +++¢a1++¢+ +t-+4++? . .. »~‘~ + +- 4 + + +4+ 4» 1 i» é + 4— 4? , r ;, t : +‘*+4++-¢*i+++j+‘r+g+¢-+g+~ ' ~ ~ ..+..+..§..+..+..+.+ ......t...+.. ”+1.4-‘+.‘:.+.. . . . : + " : : + ' Figure 5.6 The final optimized position for the BTs and the effected users by distribution 115 5.4 Design and calculation for RNC & microwave link: As mentioned above the suitable scenario for planning is that one which designed for 144kbps speed of service, that gives the ability to use the multimedia service to grantee for a subscriber that can make a video call from any place of the cell. This is important but the network can deal with less speed of service for non real time 384 kbps data service so for back bone mobile network calculation will be based in 144kbps real data time case as well. l-) Microwave links are suggested to link BTS with the radio network controller RNC centered in Janzour which is connected by microwave link to the head—quarter operator center in Tripoli, where the Mobile Switch Center (MSC). The capacity of each link is calculated as follows:- By using data from table 4.4 & table 4.5 we got: For each of the BTS], BT82, BT83, BT84, BT89 1049 users x 2lmErlang = 22.029Erlang By using table B Erlang for 1% blocking probability 22.029 Erlang —> 32 channels which need 2E] for each link between each BTs and NRC this gives very safety spare channels for microwave links between RNC and each BTs. 2- ) for the microwave link for RNC and the operator building in Tripoli ' abuseta' is need: Where E4 has I920 channel and we have 9 x 32 = 288 channel ”6 Chapter VI Results and Comparisons The Universal Mobile Telecommunication System (UMTS) has introduced a completely new high bit rate technology known as Wideband Code Division Multiple Access (WCDMA). So you can say the best thing come out on UMTS is sharing the same frequency band among different users or base stations. In order to preserve quality of existing connections, a new call can be accepted only if the estimated load in the access cell and in the neighboring cells will not exceed particular thresholds set by the radio network planner either in uplink or downlink. [15] By applying this network system in Janzour area some results have been gotten for different service, where some of these results are general for UMTS network system and an other specifically for Janzour area case study. A comparison between the requirements and parameters have been applied for different service that applied to the network system From the other side we found the results as definitions for the main parameters of the design and the specific values from Janzour area case study. 117 6.1 Results of planning process of WCDMA network: 6.1.1 General results: 0 ln WCDMA, the cell sizes are not fixed, but depend on the required capacity. 80 coverage and capacity parameters are dependent on each other. This means that both parameters have to be planned together online. 0 In both Up Link and Down Link the air interface load affects the coverage and as the effect is not exactly the same in the UL and the DL. The calculation of planning will be based on up link because it has less maximum path loss. 0 In a WCDMA cell, the available data rate depends on the interference level the closer the M8 is to the B8, the higher the data rates that can be provided. Thus, an operator that is aiming to provide 384 kbps coverage must use more BTs than an operator that is aiming for I44 kbps coverage. 0 An interference margin is required to prevent pulsating (breathing) cells. In WCDMA, the loading of a cell affects its coverage. The higher the load (higher number of users or higher bit rate). the smaller the cell size. 0 The best way to optimize the positions of BTs is to make their positions as brief explanations of the distributions of the subscribers on the area. This comes after deep analysis for defining the distribution of the people in the region. 0 Modified SOM in terms of convex sum technique gives the best averaging position for the BTs in terms of effected users. Averaging the effect of data on IIS neurons based on the size of the window. So, each neuron (B'l‘s) effected by certain data will be centered at the best place. 0 Applying the modified SOM and then just repeating the clustering will optimize the results for the region. 80, we can find the best position for the neurons (BTs) and then we will look for which data (users) closer to it for clustering. 0 The regular SOM algorithm gives fair optimization and clustering where the weights will describe the probability density of subscribers and how frequently each weight has been updated. But, the modified 80M technique gives positions to the weights describejust the density of the users only, which more accurate. 0 Free space loss increases as distance increases; also, increasing the frequency will increase the free space path loss. 0 Propagation models vary widely in their approach, complexity, and accuracy. 6.1.2. Results related to Janzour area: * The dimensioning phase, the capacity parameters are dominating than the coverage parameters in Janzour case study. This dependency appear in high speed of service, because of the high density population the area has, even if it is classified as suburban area. * The uplink has less value for maximum path of losses so it's the limit for design as done in Janzour case study. * At Janzour area with its population, we found that to get 12.2 kbps speed of . 2 . . . servnce we have to make 5.7km as the maxnmum distance between the mobile set and the BTs. For I44kbps speed of service the maximum distance is 858 meter “9 between the subscriber and the Node B. For the maximum distance between BTs & MS is 683 meter. This difference shows the inverse relation between speed of service and distance ( figure 5.4) where when the speed of service increase the maximum distance will decrease where the number of BTs increase. * Janzour area needs 9 BTS to cover 60,000 subscribers by I44kbps as speed of service. * The best speed of service to grantee Video call is I44kbps where we have ability to make video call by 64kbps but with low quality so the best scenario for Janzour area is to apply 9 BTs for all types of user. * To obtain the best distribution for the BTs in the region. The distribution of the Base Stations BTs should be mainly depending on the distribution of the users in the area. In Janzour case and based on the all information we have about it, it’s not very well designed, there are normal distribution for the users around the main, there are high density populated area in some centers of the city as well and the other parts you can see uniform distribution. * To distribute the dimensioned number of BTs on the area need an optimization technique. Self Organizing Map SOM is the best technique can be used with some modification on it for the assumed distribution for the Janzour area. * Self Organizing Map algorithm has been applied with some modification on it. With using the modified version of SOM we reach to the best distribution for the BTs. The way is using the convex sum technique which we applied by averaging the effect that each BTs effected by. Convex sum gives good optimization as well. l20 6.2. Comparison between different scenarios by definitions of the Radio Frequency Link Budget parameters: There are two main purposes for establishing the RF link budget for CDMA designs. The first one is to establish system design assumptions for all of the gains and losses in the RF path (such as vehicle loss, building loss, ambient noise margin, maximum subscriber transmit power, etc.). The second purpose of a link budget is to establish an estimate for maximum allowable path loss. This maximum allowable path loss number is used in conjunction with the propagation model to estimate cell site coverage, which ultimately determines the number of cells required for adequate system RF signal coverage and hence the system cost. The parameters within the RF link budget can be divided into four major categories. 6.2.1 Propagation Related Parameters: Propagation related parameters are those gains or losses of a link budget that are constant, independent of the multiple access technology chosen or vendor. 6.2.1.1 Building Loss Building loss is associated with the degradation of the RF signal strength caused by a building structure. In our calculation done for indoor case which has constant losses for voice l2.2 kbps & video call by l0 dB where we apply it l5dB for 384 kbps because the building loss will have effect. 6.2.1.2 Vehicle Loss Vehicle loss is the degradation of the RF signal strength caused by a vehicular enclosure. 12] So in our case we put Vm=8dB as vehicle loss in our calculation, this was only for mobile calculation because it's calculated for the car. 6.2.1.3 Body Loss Body loss, also referred to as head loss, is the degradation of the RF signal strength due to the close proximity of the subscriber handset antenna to the person’s body which is indicated for mobile calculation as 3db manly constant for all types of service. 6.2.1.4 Ambient Noise: The ambient noise defines the environmental noise that is in excess of KTB for the sector. This noise could be generated from automobiles, factories, machinery, and other man made noise. The ambient noise margin parameter can be added to the link budget to allow for an adjustment to the thermal noise value. 6.2.1.5 RF Feeder Losses: RF feeder losses include all of the losses that are encountered between the base station cabinet and the base antenna. 6.2.1.6 Antennas: Antennas can be either omni or directional. Omni antennas provide approximately the same amount of gain throughout the entire 360° horizontal pattern. Directional antennas, sometimes referred to as sector antennas, have a maximum gain in one direction with the backside being 15 to 25 dB below the maximum gain this for BTs antenna where we took it 18 dBi. I22 the other types of antennas called sectorial antenna which sometimes be 3 sectors or 6 sectors where we use 3 sectors antenna with sectionastion gain factor Gs = 2.63. And for mobile antenna gain we apply 2dBi where we used at data only and it's zero for voice transmission. 6.2.1.7 Interference Noise Rise: In determining RF coverage in CDMA systems, the effect of interference generated from other users on the serving cell as well as the neighboring cells must be considered. The interference margin is needed in the link budget because the loading of the cell, the load factor, affect the coverage. The more loading is allowed in the system, the larger is the interference margin needed in the uplink, and the smaller is the coverage area. For coverage-limited cases a smaller interference margin is suggested, while in capacity-limited cases a larger interference margin should be used. In the coverage-limited cases the cell size is limited by the maximum allowed path loss in the link budget, and the maximum air interface capacity of the base station site is not used. Typically values for the interference margin in the coverage-limited cases are 1.0- 3.0 dB, corresponding to 20-50% loading 6.2.1.8. Eb/NO: Eb/No corresponds to energy per bit over interference plus noise density for a given target Frame Erasure Rate (FER, typical voice FER target is 1%).The Eb/NO requirement depends on the bit rate, service, multipath profile, mobile speed, l23 receiver algorithms and receiver antenna structure. For mobile systems, The Eb/NO target varies dynamically as the subscriber moves around. In our case study we use these different values for these different services: Table 6.1 Different values for Eb/NO for different speed of services Speed of service Eb/NO Value (dB) Voice 12.2 kbps 5 dB Real data time 144 kbps 1.5 dB 384 kbps non real data time 1 dB Mixed type of service 2.5 dB 6.2.1.8. Fast fading margins (power control headroom): Some headroom is needed in the mobile station transmission power for maintaining adequate closed loop fast power control. This applies especially to slow-moving pedestrian mobiles where fast power control is able to effectively compensate the fast fading. Typical values for the fast fading margin are 2.0-5.0 for slow-moving mobiles where it's decreasing with increasing the speed of subscriber. 6.2.1.9. Product Specific Parameters Product specific parameters are those items in the RF link budget which can vary based on the product (base station and subscriber) chosen. l24 6.2.1.10. Product Transmit Power: The transmit power is typically referenced by the power output of the piece of equipment prior to the RF transmission lines and antennas. The maximum transmit power of the base station and the mobile is a product specific parameter. Its value is dependent on hardware components and design. For the Subscriber Unit using directive antennas for transmission will have a class dependent limit. We here use for mobile transmitter 0.l25 W for voice 12.2 kbps, but for 144kbps real time data is 0.25 W and for non real data 384kbps is 0.25 W, which is important for uplink calculation. [6] 6.3. Comparison between different scenarios by definitions of Quality of Service parameters: 1- ) Bad quality calls: Defined as call having an average frame error rate FER exceeding a threshold (usually 5% for speech). The minimum call duration is set to 7 seconds in order to increase the confidence of the averaging. Statistical data of these calls are recorded such as coordinates, start and end time and the call duration. 2- ) Dropped calls: i.e. calls that have consecutive fiame errors that exceed a threshold (usually 50 frame errors). Usually dropped calls are considered as severely poor quality calls. So bad quality and dropped calls can be taken as one measure whose percentage is referred to the number of started calls after the warm-up period. 3- ) Power outage: - for speech, this is taken from active terminals. Therefore it is slightly distorted. So the actual outage for terminals that are actively "talking," is 125 higher. Rough value is twice than that of the output. There is no discrepancy for data. 4- ) Eb/No targets: is taken from all active terminals including those in SHO. So all factors regarding the channel and diversity are taken into account. In reality the base stations antennas are not installed at equal height and thus the optimization of the base stations should be performed site by site. however the benefit of optimization is rather small when the system is strongly downlink limited and thus the uplink sensitivity improvement is not so beneficial. We have to note that the QoS can be improved in the uplink direction in lightly loaded networks. And it's known that How much of the uplink capacity improvement can be utilized in the downlink direction planning & depends naturally on the current downlink loading situation and the admission and load control strategies implemented in the network. The results of this study also clearly show that the higher sectorisation offers more capacity to the network but to achieve this antenna selection is very crucial to effectively control the interference and soft handover overhead. For each sectorisation case an optimum beam width exists. l26 Chapter VII Conclusion 7.1. Conclusions l- Network planning is notjust frequency planning, but a much broader process. The network planning process includes things like traffic estimation. figuring the proper number of cells, the placement of 883, and in terms of frequency planning estimation for code spreading has been done. The propagation model is used in conjunction with the RF link budget to obtain an estimation of the cell radius based on the allowable path loss from the link budget. WCDMA is a multiple access scheme based on spread-spectrum communication techniques. It spread the message signal to a relatively wide bandwidth by using a unique code that reduces interference. Indoor propagation models are based on the characteristics of the interior of the building, building materials, and other factors and are described in terms of various zone models. Multipath and the Doppler Effect contribute to short-term fading, and multiple reflections lead to long-term fading. ln CDMA spread spectrum technique, a Wideband carrier does not increase the capacity of the allocated BW, but it increases the ability of offering mixed services. l27 7- ln WCDMA system, the capacity is typically not limited by the exact number of channels elements, but by the amount of interference in the air interface (soft capacity). 8- In CDMA systems, it is possible to reuse the same frequency in adjacent cells. This means that the frequency reuse factor is one, while in typical GSM systems the value is between four to seven. 9- Efficient power control is very important for CDMA network performance. It is needed both in the UL and in the DL to minimize the interference in the system. l0- The UTRAN is the new radio access network designed especially for UMTS. The UTRAN consists of radio network controllers (RNCs) and Node 35 (base stations). A‘n RNC is comparable to a Base Station Controller (BSC) in (38M networks. ll- The operator may use hierarchical cell structures in certain places where traffic may need some kind of partitioning among cells with overlapping coverage. 12- The dimensioning developed program makes it easy to calculate initial and final cell radius and it provides easy interface for the users. This can be happening easily with applying the capacity calculation and coverage calculation simultaneously. l3- The link between user data rate, signal to noise ratio and system capacity has been demonstrated. The relationship between the parameters is dynamic in a way that is unique to CDMA systems. Link budgets are affected by fast fading margins and interference margins as well as by the higher noise levels that are due to the increased BW. l28 l4- The main calculations on the WCDMA network planning process is coverage calculation and capacity calculation, both have common factor (load factor my) and the best way to make the planning process is by doing both of them at the same time (online). And in terms of wanting both methods to be satisfied, the minimum should be selected. l5- As the main parameter in the planning process is Load Factor nu. . We should be accurate about its initial value to be close to the intersection point between the two calculation to make sure the optimum solution we got is within the active region of the calculation, for this reason graphical calculation needed before running any algorithm. 16- Looking for the convex sum technique from angle as modification of Self Organizing Map apply best optimization for positioning the base station on the area based on the distribution of the subscribers in the land. 17- For optimizing the base station positions, The regular SOM algorithm gives fair optimization and clustering where the weights will describe the probability density of subscribers and how frequently each weight has been updated. But, the modified SOM technique gives positions to the weights describejust the density of the users only, which more accurate. 7.2 Future Trends Even though 36 telecommunications systems are implemented a lot and people start thinking about the switch to ward next generation. But, WCDMA networks still under development and not yet in widespread operation throughout the world. In today’s fast moving information society, continuous improvement is essential. This 129 applies equally to 36 systems themselves, which have already evolved a long way from the first such systems in terms of services and capacities. And what comes after 30? Obviously, 4G. Fortunately, also here there seems to be some consensus on the features of such a system. Even higher bit rates will be supported, averaging perhaps around 2 Mbps with some peaks at 20 Mbps and in extreme cases up to 200 Mbps. As to trends in services, it will become more and more important to deliver the right information at the right time and to the right place. Content and applications become of high importance. Location-based services are enabling a variety of new applications and are complementing many existing applications with a new dimension. It is thought that many services will exist at the same time in different environments and with different resolutions. Many location-based services are already possible today provided that the location is input manually. The enhancement of those services with automatic location by the network will add value for all parties. Finally the market will decide on service profitability and on which location methods will succeed. Location-based services, as well as many other applications, are already well developed and implemented. 80 right now we have to start with : l- The need for the information about the area need to be more developed to get more accurate results. Depending only on small number of information resources makes the planning process more attractive but not too much accurate. So, some work related to collecting the data about the target region should be done. l30 2- More real experiments should be done toward signing suitable values for the link budget parameters and loses power will help on the dimensioning process. 3- The planning program can be improved by using subprograms to estimate the density and positions of traffic from digital maps based just on very important information about the area. For example Satalite image for the region, population of the area and architecture information about buildings. The way that used by the program to solve the BTS position can be modified to apply to fixed communication field for more ability of com. 13] References: [1] J. Yang, M. Aydin, J. Zhang, and C. Maple, “Umts base station location planning: a mathematical model and heuristic optimization algorithms,” Communications, lET, vol. 1, no. 5, pp. 1007 —1014, oct. 2007. [2] S. Ting and T. T. Chai, “Wedma network planning and optimisation,” aug. 2008, pp. 317 -322. [3] T. Janevski and V. Nikolic, “Applicative wcdma mobile network planning and optimization,” oct. 2009, pp. 1 —6. [4] F. Adelantado, J. Perez-Romero, and O. Sallent, “An analytical model for the reverse link of wcdma systems with repeaters in nonuniform traffic distributions,” Vehicular Technology, lEEE Transactions on, vol. 58, no. 5, pp. 2180 —2190, jun. 2009. [5] Y. Gao, X. Zhang, Y. J iang, and J. woo Cho, “System spectral efficiency and stability of 3g networks: A comparative study,” jun. 2009, pp. 1 —6. [6] Haykin, Simon and Michael Moher, “Modern Wireless Communications,” Upper Saddle River: Pearson Prentice Hall, 2005. [7] Haykin, Simon. “Neural Networks,” Englewood Cliffs, Prentice Hall, 1999. [8] M. Rahman and P. Emstrom, “Repeaters for hotspot capacity in ds-cdma networks,” Vehicular Technology, IEEE Transactions on, vol. 53, no. 3, pp. 626 — 633, May 2004. [9] F. Gu, H. Liu, and M. Li, “Evolutionary algorithm for the radio planning and coverage optimization of 3g cellular networks,” in Computational Intelligence and Security, 2009. C18 ’09. lntemational Conference on, vol. 2, 2009, pp. 109 -1 I3. [10] Lee, Jhong and Leonard Miller, “Cdma Systems Engineering Handbook,” Boston: Artech House, 1998. [I I] Walke, Bernhard and M. Althoff, “Umts,” New York, Wiley, 2003. [12] Bannister, Jeffrey et.a1, “Convergence Technologies for 3g Networks,” New York: Wiley, 2004. [13] Ray E.sheriff&Y. fun Hu. “Mobile Satellite Communication Networks,” Wiley & Sons Publisher, 2001. [14] Yang, Samuel, “3g Cdma2000 Wireless System Engineering,” City: Artech House Publishers, 2004. [15] Ham Holma and Antti Toskala , “ WCDMA FOR UMTS Radio Access For Third generation Mobile Communications,” Wiley & Sons Publisher, 2007. 132 [l6] Castro, Jonathan, “All 1p in 3g Cdma Networks,” New York, Wiley, 2004. [17] Yang, Samuel, “Cdma Rf System Engineering,” Boston, Artech House, 1998. [18] Laiho, Jaana et.al, “Radio Network Planning and Optimisation for Umts,” New York, Wiley, 2006. [19 ] J. Laiho, A. Wacker, T. Novosad, and A. Hamalainen, “Verification of wcdma radio network planning prediction methods with fully dynamic network simulator,” in Vehicular Technology Conference, 2001. VTC 2001 Fall. lEEE VTS 54th, 2001. [20] Z. Dawy, S. Davidovic, and l. Oikonomidis, “Coverage and capacity enhancement of cdma cellular systems via multihop transmission,” in Global Telecommunications Conference, 2003. GLOBECOM ’03. lEEE, vol. 2, 2003, pp. 1147— 1151 Vol.2. [21] F. Gu, H. Liu, and M. Li, “Evolutionary algorithm for the radio planning and coverage optimization of 3g cellular networks,” in Computational Intelligence and Security, 2009. C18 ’09. lntemational Conference on, vol. 2, 2009, pp. 109 —I I3. [22] R. Schoenen, W. Zirwas, and B. Walke, “Capacity and coverage analysis of a 3gpp-lte multihop deployment scenario,” in Communications Workshops, 2008. ICC Workshops ’08. lEEE lntemational Conference on, May 2008, pp. 31 —36. [23] F. Adelantado, J. Perez-Romero, and O. Sallent, “An analytical model for the reverse link of wcdma systems with repeaters in nonuniform traffic distributions,” Vehicular Technology, IEEE Transactions on, vol. 58, no. 5, pp. 2180 —2190, jun. 2009. [24] Y. Gao, X. Zhang, Y. Jiang, and J. woo Cho, “System spectral efficiency and stability of 3g networks: A comparative study,” jun. 2009, pp. 1 '—6. [25] J. Zhang, J. Yang, M. Aydin, and J. Wu, “Mathematical modeling and comparisons of four heuristic optimization algorithms for wcdma radio network planning,” in Transparent Optical Networks, 2006 lntemational Conference on, vol. 3, 2006, pp. 253 -257. I33