Low consumption dynamic time synchronization for mobile and high latency underwater acoustic communication networks

Abstract

Precise time synchronization is a foundation in the distributed wireless sensor networks in order to ensure the effective co-operation of work such as data fusion and time division scheduling. Due to the high latency caused by low sound speed and mobility between nodes in the underwater acoustic environment, it is difficult to apply mature radio time synchronization algorithms which are widely used in terrestrial networks in the underwater acoustic communication networks. In recent years, several time synchronization algorithms for underwater sensor networks have been developed. However, these algorithms are mostly in the stage of simulation research, which is based on the premise of relative static nodes which require large energy for data exchange and complex linear regression calculations. When taking the mobile platform as the time reference node, the time synchronization between nodes will lead to the problem of bidirectional delay inequality. To tackle these problems, a dynamic time synchronization algorithm based on relative speed compensation with lower energy consumption and higher reliability is proposed. Firstly, an high accuracy clock with low energy consumption is taken in the proposed algorithm to avoid the high energy consumption and large computation in the estimation processing of the clock frequency skew. Secondly, the Linear Frequency Modulation (LFM) pulse pair which can estimate the relative speed of motion is inserted during the exchanging of time information, which can estimate the link propagation delay. Thirdly, a decision mechanism with three times information interaction between two nodes is used to ensure the robustness of time synchronization. Finally, the lake experimental results show that the maximum time synchronization deviation of traditional TSHL algorithm is 18ms, whereas the maximum time synchronization deviation of the proposed algorithm is less than 6ms, and it has higher reliability when the relative speed is up to 6kt between the two nodes.