Energy Efficient Clustering in Wireless Sensors Network using Adaptive Levy-Flight Firefly Algorithm

Abstract

In Wireless Sensor Networks (WSNs), nodes have minimal energy autonomy. Accordingly, designing routing protocols that reduce the total network energy usage is one of the most challenging tasks in this field. Recent deployments of energy-efficient routing protocols for WSNs have used clustering mechanisms . In this regard, choosing optimal placements of cluster heads is an NP-hard problem that can be solved using a variety of biomimetic meta-heuristic algorithms. The Firefly Algorithm (FFA) is considered as one of the most promising and effective algorithms already used for addressing nonlinear optimization problems in general, and the energy-aware clustering for WSNs in particular. However, when solving complex optimization problems, FFA has a high risk of becoming trapped in the local optimum. Since the randomization operator plays a crucial role in updating particle positions and enhancing its global search (exploration) and convergence (exploitation) behaviors, Levy flight-based random ´ walk has been deployed to improve the firefly algorithm’s searching capability and prevent it from the premature convergence, thereby preventing it from trapping in the local optimum. This paper proposes an Adaptive Levy-Flight Firefly Algorithm-based ´ Protocol (ALFFAP) to increase the energy efficiency in WSN. MATLAB 2018a is used to simulate and assess the proposed approach, and its performance is compared to that of the classical Firefly algorithm (FFA)-based clustering protocol, LEACH, and LEACH-C. ALFFAP outperforms other protocols regarding the number of surviving nodes, total energy consumption, death of the first node, death of half node, death of the last node, stability period, and the number of data packets forwarded to the Base Station.