Distributed drift estimation for time synchronization in lossy networks

摘要

Two distributed asynchronous drift estimation algorithms for time synchronization in networks with random communication delays and measurement noise are proposed. The algorithms are superior to similar existing methods. Convergence of the algorithms to consensus in the mean square sense and w.p.1 is proved. Rate of convergence of the algorithms is analyzed. It is also shown that the algorithms can be applied as flooding algorithms, with one reference node. Illustrative simulation results are also given.