Enhanced Deep Learning-Based Optimization Model for the Coverage Optimization in Wireless Sensor Networks

A key component of wireless sensor networks (WSN) is optimal coverage. For WSN to have network lifetime and optimal resource utilization, maximum coverage must be guaranteed. The unequal distribution of sensor nodes in densely populated regions contributes to the build-up of network coverage. This research suggests a revolutionary intelligent deep learning-based optimization approach for WSN coverage optimization to tackle this problem. Here, the optimal network coverage of WSN is performed by the novel enhanced deep Q-network (EDQN) algorithm, where the parameters of DQN are tuned by nature inspired optimization algorithm called hippopotamus optimization (HO) with the intention of attaining the fitness function. In order to maximize coverage, the node position is changed in the developed methodology to represent the spatial properties of the network. Along with lowering latency, the increased coverage also increases throughput and network longevity. The outcome of several network coverage optimization trials is computed, and the EDQN-HO’s impact on network coverage optimization is also determined by adjusting the parameters. The network coverage optimization studies’ simulated findings demonstrate that the suggested EDQN-HO may be effectively utilized in a variety of settings. The proposed EDQN-HO for the WSN coverage model returns superior outcomes with 18.31%, 78.95%, 10.65%, 87.5%, 83.33%, and 38.20% than the existing methods in terms of coverage rate, energy consumption, computing time, positioning error, network lifetime, and average moving distance respectively.