Coverage Optimization and Prediction in Wireless Sensor Network Based on Enhanced Decisive Red Fox Black-Winged Kite With Multistrategies

Abstract

Wireless sensor networks' (WSNs) coverage optimization and prediction are critical for improving the efficiency of WSNs in various applications, which aim to maximize the area of interest while minimizing the number of sensors to balance energy consumption and network lifespan. More specifically, coverage optimization focuses on ensuring maximum coverage in WSNs through the strategic placement of resource-constrained sensor nodes. However, existing approaches often reach local optima and exhibit poor performance in optimization. Consequently, this leads to suboptimal coverage, where certain areas remain unmonitored or excessive overlap occurs among multiple sensors. To address this, an enhanced Decisive Red Fox and Black-winged Kite with Multistrategies (DRFBK-MS) is proposed for optimizing WSN coverage while ensuring initial value distribution across the search space to enhance diversity. The proposed DRFBK-MS approach lies in its hybrid integration of Decisive Red Fox and Black-winged Kite optimization strategies, enhanced with Sobol sequence initialization for better population diversity, simulated annealing, and dynamic search steps to escape local optima. Additionally, it uniquely incorporates a Reinforcement Convolutional Network (ReConvNet) for accurate and low-complexity prediction of WSN node status. This unified optimization–prediction framework significantly improves coverage performance, search efficiency, and energy utilization, achieving a coverage rate of 96.24%, coverage efficiency of 98.1%, with an execution time of 10 s, making it a robust and efficient solution for WSN coverage optimization.