| Node self-localization technology is one of the main supporting technologis for Wireless Sensor Network.In most applications,determining the physical positions of sensor nodes is the basic requirements.At present,the existing node self-localization algorithm has greatly impacted by circumstances and anchor deploying,high algorithm complexity and high power consumption problem and so on.It is obviously that the technology of Wireless Sensor Network will go-ahead very quickly and its applications will be wider and wider in the coming future.The research as described in this dissertation is of significance in both theoretical and application areas.This thesis is for the purpose of making some related discussions and research of self-localization technology based on wireless sensor network.(1) To address the problem that anchor ratio have a strong impact on localization error and coverage in Euclidean algorithm,This paper proposes a improved distributed localization algorithm.This method uses high localization accuracy nodes as new anchor nodes.According to the information,other nodes raise the localization accuracy by iterative refinements.And it controls the circulation times with the variance of coordinates.Simulation results show that,compared with Euclidean algorithm,improved localization algorithm can enhance the localization accuracy efficiently when the anchor ratio is lower.It is obvious that the influence of anchor ratio on the localization accuracy and coverage is less.(2) We apply the Support Vector Machine(SVM) theory to the sensor node self-localization algorithm.Then a range-free localization algorithm based on SVM classification regions is proposed.First,SVM constructed a binary decision tree classifier via learning of training data.Then the classifier determined the certain classification region where the unknown nodes located in.Finally,we use the region's center point as the estimated position of the unknown node.The proposed algorithm required mere connectivity information(i.e.,hop counts only),so as to reduce the network cost and communication loads.The simulation results show that this algorithm alleviates the coverage holes and border problem significantly while certain location accuracy was assured.In the end,it is the conclusion of the thesis and the prospect for the future research. |