Improved Accurate Localization using PSO and the Weighted Dijkstra Algorithm in Software Defined Wireless Sensor Networks

Abstract

Wireless Sensor Networks (WSNs) are crucial in various fields, including monitoring the environment, surveillance, and healthcare. They rely on localization services for accurate data transfer and optimal network performance. Traditional WSN techniques struggle to adapt to dynamic environmental changes beyond the intended task scope. A synergy between Software-Defined Networking (SDN) and WSN has been suggested to address this issue. This research paper presents proposed approach for machine learning-based localization in Software Defined Wireless Sensor Networks (SDWSNs) using Particle Swarm Optimization (PSO) technique and the Weighted Dijkstra algorithm. PSO technique is used for clustering, the weighted Dijkstra algorithm (WDA) for finding the shortest path and sending data packets, and machine learning algorithms like AdaBoost and Naïve Bayes for data classification. The effectiveness of the proposed approach is measured using energy consumption, throughput, network lifespan, and packet delivery ratio, outperforming existing models like OEERP, LEACH, DRINA, and BCDCA. The machine learning algorithms' performance is also evaluated, with Naïve Bayes achieving the highest accuracy of 78.24% and AdaBoost 98.90%.