Wireless sensor network (WSN) is a key technology for a wide range of military and civilian applications. Limited by the energy resources and processing capabilities of the sensor nodes, reliable and efficient communications in wireless sensor networks are challenging, especially when the sensors are deployed in hostile environments. This research aims to improve the reliability and efficiency of time-critical communications in WSNs, under both benign and hostile environments.We start with wireless sensor network with mobile access points (SENMA), where the mobile access points traverse the network to collect information from individual sensors. Due to its routing simplicity and energy efficiency, SENMA has attracted lots of attention from the research community. Here, we study reliable distributed detection in SENMA under Byzantine attacks, where some authenticated sensors are compromised to report fictitious information. The q-out-of-m rule is considered. It is popular in distributed detection and can achieve a good trade-off between the miss detection probability and the false alarm rate. However, a major limitation with this rule is that the optimal scheme parameters can only be obtained through exhaustive search. By exploiting the linear relationship between the scheme parameters and the network size, we propose simple but effective sub-optimal linear approaches. Then, for better flexibility and scalability, we derive a near-optimal closed-form solution based on the central limit theorem. It is proved that the false alarm rate of the q-out-of-m scheme diminishes exponentially as the network size increases, even if the percentage of malicious nodes remains fixed. This implies that large-scale sensor networks are more reliable under malicious attacks. To further improve the performance under time-varying attacks, we propose an effective malicious node detection scheme for adaptive data fusion; the proposed scheme is analyzed using the entropy-based trust model, and has shown to be optimal from the information theory point of view.Next, we observe that: while simplifying the routing process, a major limitation with SENMA is that data transmission is limited by the physical speed of the mobile access points (MAs) and the length of their trajectory, resulting in low throughput and large delay. To solve this problem, we propose a novel mobile access coordinated wireless sensor network (MC-WSN) architecture. The proposed MC-WSN can provide reliable and time-sensitive information exchange through hop number control, which is achieved by active network development and topology design. We discuss the optimal topology design for MC-WSN such that the average number of hops between the source and its nearest sink is minimized, and analyze the performance of MC-WSN in terms of throughput, stability, delay, and energy efficiency by exploiting tools in information theory, queuing theory, and radio energy dissipation model. It is shown that MC-WSN achieves much higher throughput and significantly lower delay and energy consumption than that of SENMA.Finally, motivated by the observation that the number of hops in data transmission has a direct impact on the network performance, we introduce the concept of the N-hop networks. Based on the N-hop concept, we propose a unified framework for wireless networks and discuss general network design criteria. The unified framework reflects the convergence of centralized and ad-hoc networks. It includes all exiting network models as special cases, and makes the analytical characterization of the network performance more tractable. Further study on N-hop networks will be conducted in our future research.
Read
