WSNs data acquisition by combining expected network coverage and clustered compressed sensing

To tackle the challenges of extensive data transmission and high redundancy in wireless sensor networks (WSNs), this study proposes a novel data collection scheme based on expected network coverage and clustered compressive sensing (CS). First, the K-medoids clustering algorithm organizes nodes within the WSN coverage area into clusters. Combined with an optimized network coverage algorithm, a node scheduling strategy is introduced to focus on critical observation areas while minimizing overall energy consumption. Next, by analyzing the relationship between network clustering and node deployment, a weakly correlated observation matrix is designed to mitigate the impact of packet loss on data collection. Finally, the sparrow search algorithm is employed to enhance the accuracy of CS data reconstruction at the cluster head. Simulation results demonstrate that, compared to existing data collection schemes, the proposed approach significantly reduces WSN transmission overhead, ensures accurate recovery of raw data, decreases data reconstruction error, and extends network lifetime.