Evolutionary based Multi objective Routing in Diverse Wireless Sensor Networks

Abstract

To reduce energy consumption, wireless sensor networks (WSNs) require efficient routing newlinealgorithms as communication is the cause of 60% usage of total energy. Routing protocols in WSNs newlineaim to conserve sensor node energy to elongate network lifetime and maintain connectivity. This newlineresearch focuses on optimization of routing in diverse WSNs using the theoretical foundations of routing algorithms. The primary focus is to develop a comprehensive framework for optimal routing in terrestrial and underwater wireless sensor networks by selecting optimal cluster head and path to maximize the network performance and throughput by considering multiple conflicting objectives. The proposed algorithms are analyzed by simulating it in various conditions and comparing it with the existing algorithms using NS2.35, a comprehensive open-source network simulator. This research proposes routing algorithms to enhance data transmission, energy efficiency, and network lifespan in two environments: terrestrial and underwater. The first approach, optimal multiobjective genetic swarm optimization (MO-GSO), addresses network partitioning and connectivity issues in terrestrial WSNs. The proposed algorithm outperforms the other state-of-the-art methods with a packet delivery ratio (PDR) of 96.8%, energy consumption of 0.19J, packet loss ratio (PLR) of 3.2%, and end-to-end delay (E2ED) of 30ms. The second algorithm focuses on underwater WSNs, which is an integration of elite opposition-based fuzzy logic for multi-criteria decision-making (EOFL-MCDM) in cluster head selection, and a hybrid auction-based wild horse optimization algorithm (HAWHO) for path selection. This approach significantly improves throughput, E2ED, energy consumption, network lifetime, and PDR, outperforming the existing algorithms by up to 46%. Both algorithms were simulated using an NS2.5 simulator and extensive simulation analysis was carried out to validate the performance of the proposed algorithms. Future research could extend this work by addressing Quality ....