Distributed Localization Algorithms In Uncertain Network Environment

Since the 1990 s,the Wireless Sensor Network(WSN)has drawn great attention worldwide.Such networks have the characteristics of low cost and flexible deployment,and own great practical value in the field of environmental monitoring and military applications.Among them,the location information of the sensors is crucial to the application of WSN.Taking into account the limited usage of GPS in a variety of occasions,developing WSN self-localization algorithm has become a very significant research direction,and how to deal with the impact of uncertainties in the network environment on algorithms has become a hot spot of scholars.This thesis mainly focuses on the distance-based distributed localization algorithm in wireless sensor network with communication delay,data packet loss and quantization noise.It extends the distributed localization algorithm DILOC proposed by Khan U.A.et al.to complex environment.While maintaining the DILOC network topology conditions unchanged,the thesis studies the localization problem in the condition of time-varying communication delay and data packet loss,including:1.a synchronous localization algorithm in the case that each node can realize the local clock synchronization according to the clock synchronization protocol,and the convergence of which is rigorously proved with graph theory,which means the sensors to be located can be globally converged to their exact locations;2.an asynchronous localization algorithm in the case that the unknown sensors' update time instants are different with each other,which can save the consumption of suppressing the local clock drift,and the global convergence of which is proved using consensus theory for switching topology;3.asynchronous localization algorithms in the case when sensor nodes are equipped with uniform quantizer and logarithmic quantizer respectively,as well as discussion on the influence of quantizer on localization accuracy and convergence speed of the localization algorithms,according to mathematical models and a number of simulation experiments.In addition,the distance measurement error and distance estimator are also added in the simulation.The simulation results show that the distributed localization algorithms proposed in this thesis is still applicable in environment with distance noise.It is verified that under environment with communication delay and data packet loss,the asynchronous distributed localization algorithm is fault tolerant for both quantization noise and distance error.