Joint Time Synchronization and Localization of Multiple Source Nodes in Wireless Sensor Networks

Abstract

Time synchronisation and localisation should be jointly conducted for time-based source location estimates in an asynchronous network. In this paper semidefinite programming (SDP), complexity-reduced SDP and linear least square (LLS) estimator are proposed for time synchronisation and localisation of multiple source nodes. The proposed algorithms provide joint estimates for the source locations and clock parameters and avoid the shortcoming of maximum likelihood (ML) estimator that requires an initial solution. Then, a location refinement (LR) technique is designed to improve the accuracy performance of the estimated parameters. The simulations show that the original SDP algorithm provides better accuracy performance than the complexity-reduced SDP. However, the complexity-reduced SDP runs faster than the original SDP. Although the complexity of LLS estimator is lowest among three proposed algorithms, the convex optimisation algorithms, including the original SDP and complexity-reduced SDP, have more robust performance compared with the LLS estimator.