Performance Enhancement in Distributed Sensor Localization Using Swarm Intelligence

Wireless Sensor Networks (WSNs) consist of distributed autonomous devices which sense the environmental or physical conditions cooperatively and pass the information through the network to a base station. Sensor Localization is a fundamental challenge in WSN. Location information of the node is critically important to detect an event or to route the packet via the network. In this paper localization is modeled as a multi dimensional optimization problem. This problem is solved using bio inspired algorithms, because of their quick convergence to quality solutions. Distributive localization is addressed using Particle Swarm Optimization (PSO) and Comprehensive Learning Particle Swarm Optimization (CLPSO). The performances of both algorithms are studied. The accuracy of both algorithms is analyzed using parameters such as number of nodes localized, computational time and localization error. Comparison of both the results is presented. A simulation was conducted for 100 target nodes and 20 beacon nodes, which resulted in CLPSO being 80.478% accurate, and PSO 61.48% accurate. The simulation results show that the PSO based localization is faster and CLPSO is more accurate.