Multi-objective fixed set search for lifetime and connectivity optimization in WSN

Abstract

Wireless Sensor Networks (WSNs) have witnessed massive development but still face significant challenges, primarily energy limitations and network lifetime. To address these issues, clustering and optimal Cluster Head selection are highly effective solutions. This thesis proposes the use of an advanced metaheuristic algorithm, the Fixed Set Search (FSS), as an alternative to probabilistic (LEACH) or purely geometric (K-Means) methods. The proposed algorithm relies on a multi-objective fitness function that balances throughput, compactness, connectivity, and residual energy. Simulation results conducted in Python over 30 random network topologies demonstrated the clear superiority of FSS. It significantly prolonged the network stability period and maximized data throughput compared to benchmark algorithms, proving its efficiency in optimizing WSN performance.