Enhanced Energy Efficient Congestion Control in Wireless Sensor Networks

Abstract

Wireless Sensor Networks (WSNs) play a pivotal role in monitoring physical and environmental conditions, yet face challenges such as energy constraints and network congestion, impacting their lifespan and performance. This research presents a novel approach that aimed at mitigating congestion issues and maximizing the energy efficiency of WSNs. newlineInitially, this research introduces an energy-efficient Ultra Scalable Ensemble Clustering mechanism, complemented by the Flamingo Search Algorithm based Fuzzy Inference System for Cluster Head (CH) selection. Optimization factors for CH selection include the distance to the Base Station (BS), node density and nodes residual energy. Rat Swarm Optimization is employed to select routes between CH and BS. Next, it addresses congestion mitigation through an Adaptive Neuro-Fuzzy Inference System (ANFIS) based path determination approach, coupled with the Black Widow Optimization (BWO) algorithm. The proposed method begins by establishing a congestion mitigation framework, forecasting buffer occupancy using exponential smoothing. ANFIS is then applied to determine paths by considering hop count, buffer occupancy and remaining energy as input factors. Simulation results demonstrate superior quality of service, high energy levels, low delay and a high packet delivery ratio, along with an increasing number of alive nodes, outperforming existing methods. newlinevi newlineFurthermore, a clustering-based routing protocol is introduced in the context of Power-Efficient Gathering in Sensor Information Systems (PEGASIS). This protocol employs a double cluster head (PDCH) approach with an artificial neural network (ANN) to analyze overall network lifetime. The methodology comprises four phases: clustering network nodes using the firefly algorithm, CH selection via ANN, chain formation through PDCH and secondary CH (SCH) selection using the grey wolf optimizer (GWO).