Consensus Algorithms For Wireless Sensor Networks:Design And Analysis

With the rapid development of sensor technology, wireless communication technology and MEMS, wireless sensor networks (WSNs) have become an important bridge connecting cyber world and physical world, which have changed production and life styles of humans. Consensus algorithms, as a class of significant distributed control and optimization algorithms, have been extensively studied and applied in WSNs due to their strong robustness and scalability. The typical applications include time synchronization, sensor fusion, localization and so on. However, the research on consensus algorithms often ignores the fundamental and practical constraints of sensor networks, which include constrained energy, communication delays, security and privacy issues and etc. Thus, the applications of consensus algorithms in WSNs are limited. Based on the previous works, considering the energy efficiency, robustness to delay and security and privacy of existing algorithms, this thesis further investigates the consensus algorithms design and analysis in wireless sensor networks. The main contributions of this thesis are summarized as follows.1. For the convergence rate of average consensus in WSNs, a mobile node is exploited to ac-celerate the convergence of average consensus and promote its energy efficiency. Firstly a sufficient condition of the state of the mobile node is given, and then an interaction proto-col between the mobile node and static nodes is designed. The matrix eigenvalue is utilized to analyze the performance of the proposed algorithm in expectation, which is theoretically proven to converge faster. Finally, to deal with the final value, an average preserving scheme is designed to realize mobility assisted average consensus.2. For the distributed time synchronization of sensor nodes under bounded noise model, a high-ly accurate distributed consensus-based time synchronization algorithm is proposed. The principle that a bounded monotonic sequence must possess a limit and the concept of max-imum consensus are exploited to design a novel relative skew update algorithm, which is proven to converge in probability. Meanwhile, skew compensation and offset compensation are also designed to realize accurate and fast time synchronization.3. For privacy issues of sensor fusion in WSNs, a privacy preserving maximum consensus algorithm is proposed. The noise sequence is introduced to preserve the identity of the maximum state owner and the initial states of all nodes except for the maximum state owner. Then, the convergence time of the proposed algorithm is analyzed in expectation and it is proven that it converges in finite time. The probability of being inferred serves to evaluate the degree of privacy, and the analysis on the privacy preserving performance is carried out theoretically.Finally, the thesis is concluded and the future works are discussed.