Publication

Abstract : Wireless Sensor Networks(WSN) are becoming more crucial in contemporary applications and thus energy efficiency is crucial for extended system durability. This work focuses on improving the accuracy of energy consumption prediction in WSNs using nature-inspired optimization strategies. It explores the application of Particle Swarm Optimization (PSO), Salp Swarm Optimization (SSO) and Sparrow Search Algorithm (SSA) to fine-tune the weights of a linear regression model for enhanced forecasting performance. Through the reduction of prediction errors, such algorithms will have the capability of delivering a reliable solution for the estimation of energy consumption in sensor nodes. Comparative studies feature each approach’s convergence rate, predictability, and computational cost. Results confirm the ability of PSO to outperform conventional optimization significantly in the regression training, particularly in challenging and dynamic WSN scenarios. This work identifies bio-inspired algorithms as promising tools for developing intelligent and energy-efficient models for WSNs. This approach contributes to the advancement of more sustainable and optimized network infrastructures through the application of predictive analysis driven by nature-inspired optimization techniques.