Time Synchronization for Mobile Underwater Sensor Networks

Ying Guo, Yutao Liu
2013 Journal of Networks  
Time synchronization is very crucial for the implementation of energy constricted underwater wireless sensor networks (UWSN). The purpose of this paper is to present a time synchronization algorithm which is suitable to UWSN. Although several time synchronization protocols have been developed, most of them tend to break down when implemented on mobile underwater sensor networks. In this paper, we analyze the effect of node mobility, and propose a Mobile Counteracted Time Synchronization
more » ... , called "Mc-Sync", which is a novel time synchronization scheme for mobile underwater acoustic sensor networks. It makes use of two mobile reference nodes to counteract the effect of node mobility. We also analyze and design the optimized trajectories of the two mobile reference nodes in underwater environment. We show through analysis and simulation that Mc-Sync provides much better performance than existing schemes. Index Terms-time synchronization, mobile reference node, optimized trajectory, node mobility, underwater sensor networks I. INTRODUCTION Time synchronization is an important issue for many distributed applications, especially for sensor networks [1]. It requires collaboratively processing of time sensitive data in the applications of environment supervision, target tracking and so on [2][3]. Recently many time synchronization protocols for terrestrial wireless sensor networks have been proposed [4][5][6], which provide a high degree of precision. As terrestrial communication is based on Radio Frequency (RF) technology, all of these mechanisms assume that propagation latency is negligible and can be effectively factored out of design consideration. While underwater communication mainly uses acoustic communication technology. There are several fundamental differences between RF communication and acoustic communication, such as large propagation delay and node mobility [7][8]. Thus protocols for terrestrial sensor networks could not be directly applied to underwater acoustic sensor networks. Underwater sensor networks (UWSN) inherit many different features. Node mobility is one of the most important effects [9] . From empirical observation, nodes without any self-propelling capability can move with wind and ocean current typically at the rate of 0.83-1.67m/s, and existing Autonomous Underwater Vehicles (AUV) typically move at a rate of up to 2.9m/s. In addition, node movement also brings Doppler shift [10] [11] , which adds the difficulty to estimate propagation delay exactly. Thus time synchronization algorithm must be able to cope with sensor node mobility. Moreover, UWSN has long propagation delay, limited bandwidth, limited transmission rate, high bit error rate and so on [12] . The propagation speed of underwater acoustic channel is five orders slower than radio waves, thus resulting in significantly longer propagation delay [13] , which makes relatively large Doppler Effect and inter symbol interference. Due to strong attenuation in high frequency band, acoustic communication has limited available bandwidth and low carrier frequency. Because of low cost hardware and limited power supply, most of sensor nodes have limited transmission rate. In addition, limited bandwidth and transmission rate make it suffers from high bit error [14] . All of these make terrestrial time synchronization different from underwater. Recently, several underwater sensor network time synchronization algorithms have been proposed. TSHL [15] is the first time synchronization algorithm designed for high latency networks specifically. It uses one-way communication to estimate the clock skew and two-ways communication to estimate the clock offset. MU-Sync [10] runs two times of linear regression to estimate the clock skew and clock offset for cluster based UWSN. TSHL assumption of constant propagation delay, thus it cannot handle mobility issues. MU-Sync assumes that the one-way propagation delay can be estimated as the average round trip time which causes extra errors, and has relative high message overhead. Mobi-Sync [7] and D-sync [8] consider more about node mobility. Mobi-Sync need the help of surface buoys with GPS and synchronized super nodes. Ordinary nodes use a correlation model to estimate their velocity and launch time synchronization. In order to improve time synchronization precision, D-sync exploits Doppler shift to provide an indication of the relative motion between
doi:10.4304/jnw.8.1.116-123 fatcat:ck27ucwmgnfqfjp3r745jhwmwi