On Least Squares Weighted Recursive Estimation of Clock Skew and Offset in Wireless Sensor Networks

Three clock synchronization algorithms for Wireless sensor networks (WSNs) in Pairwise broadcast synchronization (PBS) mechanism are derived. They include the joint Least squares estimator (LS), joint Least squares weighted estimator (LSW) and joint Least squares weighted Recursive estimator (R-LSW). For these estimators, the corresponding algorithms are derived and described by assuming a Gaussian random delay model. Unlike PBS, these estimators can achieve the Cramer-Rao lower bound (CRLB) for both listening node and active node without knowledge of the deterministic delay. The purpose of considering R-LSW is to reduce the use of storage space with the method of estimating while observing. Simulation and analytical results verify that the estimators are efficient.