An accurate clock offset estimation method between two moving objects based on maximum likelihood estimation

Abstract

We consider a problem of accurate clock offset estimation between two moving objects. To establish clock synchronization between objects, the principle of the Two Way Ranging (TWR) method has been commonly used in the Network Time Protocol (NTP), the IEEE 1588 standard and so on. In this paper, we first reveal that they do not perform accurately at all when two objects are in motion. Next, to solve the problem, we propose an accurate clock offset estimation method. In the proposed method, we utilize the location information on the two objects as well, and taking into consideration that the measured distance between them is not Gaussian-but Rician-distributed, we strictly derive the Maximum Likelihood estimation (ML) on the clock offset. Then, to reduce the computational complexity of the strict ML method, we introduce two approximated ML methods. In computer simulation, we show that the proposed method outperforms conventional methods, achieving the clock offset estimation accuracy of less than several tens of nanoseconds, which is required for current and future wireless communication tools among moving objects.