Adaptive Hybrid Firefly Algorithm for Cluster Head Selection in Gaussian-Grid Based Wireless Sensor Networks

Abstract

A novel hybrid algorithm combining the Firefly Optimization Algorithm (FOA) with nature-inspired enhancements is proposed for cluster head selection in wireless sensor networks (WSNs) deployed over Gaussian virtual grids. The study analyzes three established protocols Least Energy-based Gaussian Network (LEGN), Threshold Energy-based Gaussian Network (TEGN), and the FOA-based approach—through node-level energy and packet statistics. The performance of the proposed hybrid method is compared against these benchmarks using metrics such as Packet Delivery Ratio (PDR), Energy Efficiency, and Residual Energy. A comprehensive literature survey on bio-inspired clustering algorithms reveals significant research gaps in adaptive parameter tuning and multi-objective optimization. The proposed Adaptive Hybrid Firefly Algorithm (AHFA) addresses these limitations by incorporating dynamic threshold mechanisms, social behavior patterns from dragonfly algorithms, and energy-aware fitness functions. Simulation results demonstrate significant improvements in energy balancing (15-20% better residual energy retention), throughput (12% higher PDR), and network lifetime (21% extension) over conventional methods. The algorithm maintains computational complexity of O(n²) while achieving superior convergence properties and robustness against node failures.