Improving the Performance of Wireless Sensor Networks Through Optimized Complex Field Network Coding

Abstract

Signal transmission and information fusion in wireless sensor networks (WSNs) are conventionally assumed to operate over orthogonal channels, which makes the network bandwidth and throughput inefficient. To remedy this inefficiency and improve the performance of the WSNs, we consider complex field network-coded (CFNC) relay-assisted communications, which operates over nonorthogonal channels and provides both spatial and temporal diversity. We derive the optimal likelihood ratio test-based fusion rule for the considered system. To provide robustness against the multiaccess interference, each sensor in the CFNC-coded system is assigned to a unique predetermined signature. Hence, the signature selection and the relay power allocation become crucial factors affecting the performance of the WSNs. We also develop an analytical method to jointly adjust the sensor signatures and the relay power utilizing the average symbol error rate bound of the network together with some information theoretical results. Finally, we evaluate the detection performance of the proposed scheme and compare it with that of the conventional method. The simulation results suggest that the proposed signature selection and relay power allocation method in the CFNC-coded relay-assisted WSNs considerably improves the network performance.