Robust Clock Synchronization Methods for Wireless Sensor Networks

Wireless sensor networks (WSNs) have received huge attention during the recent\nyears due to their applications in a large number of areas such as environmental\nmonitoring, health and traffic monitoring, surveillance and tracking, and monitoring\nand control of factories and home appliances. Also, the rapid developments in the\nmicro electro-mechanical systems (MEMS) technology and circuit design lead to a\nfaster spread and adoption of WSNs. Wireless sensor networks consist of a number of\nnodes featured in general with energy-limited sensors capable of collecting, processing\nand transmitting information across short distances. Clock synchronization plays an\nimportant role in designing, implementing, and operating wireless sensor networks,\nand it is essential in ensuring a meaningful information processing order for the data\ncollected by the nodes. Because the timing message exchanges between different\nnodes are affected by unknown possibly time-varying network delay distributions, the\nestimation of clock offset parameters represents a challenge. This dissertation presents\nseveral robust estimation approaches of the clock offset parameters necessary for time\nsynchronization of WSNs via the two-way message exchange mechanism. In this\ndissertation the main emphasis will be put on building clock phase offset estimators robust with respect to the unknown network delay distributions.\nUnder the assumption that the delay characteristics of the uplink and the downlink\nare asymmetric, the clock offset estimation method using the bootstrap bias\ncorrection approach is derived. Also, the clock offset estimator using the robust Mestimation\ntechnique is presented assuming that one underlying delay distribution is\nmixed with another delay distribution.\nNext, although computationally complex, several novel, efficient, and robust estimators\nof clock offset based on the particle filtering technique are proposed to cope\nwith the Gaussian or non-Gaussian delay characteristics of the underlying networks.\nOne is the Gaussian mixture Kalman particle filter (GMKPF) method. Another\nis the composite particle filter (CPF) approach viewed as a composition between\nthe Gaussian sum particle filter and the KF. Additionally, the CPF using bootstrap\nsampling is also presented. Finally, the iterative Gaussian mixture Kalman particle\nfilter (IGMKPF) scheme, combining the GMKPF with a procedure for noise density\nestimation via an iterative mechanism, is proposed.