Joint Sparse Recovery in Precision Agriculture WSN and IoT Applications

Abstract

Joint sparse recovery is a problem in which com pressed measurements at multiple sensing nodes are jointly reconstructed. Such problem is common in Wireless Sensor Network (WSN) and Internet of Things (IoT) applications, such as environment monitoring for precision agriculture. In this paper, we propose a joint sparse recovery algorithm, the Joint Fast Matching Pursuit (JFMP), considering the JSM-I sparsity model commonly encountered in WSN and IoT applications. JFMP iteratively estimates the support of the common component of the sparse signals, and then estimates the sparse signals based on such support. In each iteration, the estimated support is further refined. Our experiments, using both random and real-life data, show that JFMP achieves perfect reconstruction of the measured attribute in a short time with fewer sensors and measurements compared to related algorithms. This is due to JFMP optimum selection strategy, pruning strategy, and avoidance of large matrix inversion during signal estimation.