A hybrid particle swarm optimization with a variable neighborhood search for the localization enhancement in wireless sensor networks

摘要

Localization accuracy and development costs are key and substantial issues considered when operating and managing a wireless sensor network (WSN). This study presents a modern and high-efficiency algorithm based on a new optimization technique for localization processes in an outdoor environment. The new optimization technique combines particle swarm optimization (PSO) with variable neighborhood search (VNS) and is called hybrid particle swarm optimization with variable neighborhood search (HPSOVNS). The objective function, which utilized by HPSOVNS for optimization, is the last mean squared range error of all neighboring anchor nodes. The interior distances between WSN nodes are calculated using a received signal strength indicator (RSSI) function. HPSOVNS is a hybrid optimization technique showing elevated performance in finding the best solution that rapidly affirms the minimization of an objective function without being stuck in local optima. The proposed algorithm can increase localization accuracy because it combines the positive features and effective capabilities of PSO and VNS with RSSI. Simulation results show that HPSOVNS performs better than other algorithms based on basic PSO and even state-of-the-art localization algorithms, such as GEPM, NLLE, and RSSI-LSSVR. The performance of HPSOVNS is demonstrated in several evaluation metrics, such as localization accuracy, localization rate, and localization time.