Modified Rat Swarm Optimization Based Localization Algorithm for Wireless Sensor Networks

A huge number of sensor nodes collect information about the environment around them in wireless sensor networks (WSNs), but this information is not valuable until the precise location where it was collected is revealed. No infrastructure exists to estimate the locations of deployed nodes, since global positioning system (GPS) receivers are too expensive to be included with every sensor. Hence, localization of sensor nodes plays a key role in a number of WSN applications, such as health, whether, industrial and military. Sensor node localization is one of the most significant challenges in WSNs, that aims to determine the coordinates of unknown nodes based on the coordinates of anchor nodes. The researchers are designing new localization schemes that are suitable for WSN implementation, as traditional localization algorithms (eg., GPS) are not suitable. There are a variety of meta heuristic algorithms used to solve optimization problems in WSNs. Rat swarm optimizer (RSO) is a recently developed algorithm with competitive performance and remarkable different results from other meta-heuristic algorithms. In this work, we propose a modified rat swarm optimizer (MRSO) based nodes localization problem in wireless sensor networks (WSNs). To evaluate the proposed work comparative study is done with the original RSO and other meta-heuristic based approaches. The proposed MRSO outperforms the original RSO algorithm and other existing optimization algorithms in terms of different localization error metrics. The proposed MRSO reduces the ALE by 68.52%, 71.75%, 70.58% and 66.81% comparing to RSO, bat optimization algorithm (BOA), BOA variant 1 and BOA variant 2, respectively.