New Modified Adaptive Learning Neural Network for Coverage and Connectivity Issues in WSN Using Energy Efficient Routing Algorithm

Abstract

Wireless Sensor Networks (WSNs) are increasingly being employed in a variety of applications ranging from military to infrastructure monitoring. However, with the ever-increasing demand of higher coverage, better connectivity, and accuracy, optimization of such networks is becoming a complex task. To address this challenge, this paper proposes a novel Modified Adaptive Learning Neural Network (M-ALNN).This new method is designed to address both coverage and connectivity issues within a given WSN. The MALNN adaptively optimizes WSNs by taking advantage of state-of-the-art reinforcement learning algorithms. The proposed algorithm uses a combination of two different algorithms, viz., the Supervised Neuron Algorithm (SNA) and the K-Means Clustering Algorithm, to achieve optimal coverage and reconfiguration of network topology. Both SNA and K-Means Algorithm are trained through a reinforcement learning algorithm to adaptively adjust their parameters as per the current application requirements. After optimization, the proposed M-ALNN algorithm improves the coverage and connectivity of WSNs by adjusting the network parameters. Simulation experiments are conducted to verify the effectiveness of the proposed method. Results show that our proposed network outperforms the conventional WSN in terms of coverage and connectivity when tested under various scenarios.