Abstract
In wireless sensor networks, target localization has been the focus of considerable research effort, and the compressive sensing (CS) based localization method is of particular interest. However, most existing works usually assume that the positions of sensors are known exactly, while in practice, they may not be accurate. When the assumption is violated, the localization performance will deteriorate dramatically. In this article, we propose a novel CS-based multiple target localization method using the superimposed received signal strength, measured by sensors with inaccurate positions. In order to address such issues, we regard the known but inaccurate sensor locations as adjustable parameters. Accordingly, the sensor positions can be refined through the adjustment of parameters. As a result, the problem is reformulated as a joint sparse signal estimation and parameter optimization task. Then, the variational expectation-maximization (EM) algorithm and the subspace trust-region method are applied to iteratively estimate the unknown target locations and refine inaccurate sensor locations. Simulation results are included to confirm the superior localization performance of the proposed algorithm by comparing with the state-of-the-art methods.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call