An Efficient Localization Approach in Wireless Sensor Networks Using Krill Herd Optimization Algorithm

Abstract

Arbitrary deployment of mobile or static sensor nodes has been a challenging issue in wireless sensor networks (WSNs) over the years. It is difficult to identify the accurate location for mobile sensor nodes due to their time-variant movement. Global position system (GPS) for each sensor node costs high power consumption, mainly in large scale WSNs. Hence, designing an energy-efficient localization algorithm is very crucial for sensor nodes. In this article, we adopt a meta-heuristic algorithm known as Krill Herd (KH) inspired by the behavior of krills. KH is selected to find the location of nonanchor nodes by using mobile anchor nodes. Mobile anchor nodes have GPS units and broadcast their exact locations at regular time intervals to find the location of nonanchor nodes. The time-variant location of the nonanchor node is calculated by movement generated by the existence of all sensor nodes in the neighborhood range, foraging motion of anchor nodes, and random diffusion of all the sensor nodes. Two genetic operators namely crossover and mutation are added to the algorithm to analyze the behavior of mobile anchor nodes. The experimental results are carried out for minimizing localization error, propagation error, roo-mean-square error, and mean absolute error. The results demonstrate the outperformance of KH-based localization algorithm with minimum error rate when compared with particle swarm optimization algorithm and improved amorphous localization algorithm.