Study On DV-distance Based Collaborative Localization Algorithm In Wireless Sensor Networks

Node localization technology is one of the key supporting technologies inwireless sensor network (WSN), so it gets more and more attention and research inrecent years. In the traditionallocalization algorithm, only distance or angleinformation between the unknown node and the anchor node is used, so some nodesmay not be able to be located. In the collaborative localization algorithm, thecollaborative idea is used to solve the problem of wireless sensor network nodelocalization. Not only the distance or angle information between the unknown nodeand the anchor node is used, but also the distance or angle information between theunknown node and unknown nod is used too. Because the collaborative localizationalgorithm uses more redundant information, it usually has higher positioning accuracyand betterrobustness.DV-Distance algorithm uses multiplehops collaboration in thelocalization process,is a one of the typical collaborative localization algorithms. Represented byDV-Distance algorithm, this paper mainly studieson how to improvethe performance ofthe collaborative localization algorithm. The main research contentsin this paper are asfollows.①In this paper the concept, background,and existing problems of collaborativelocalization algorithm in wireless sensor network are introduced systematically andcomprehensively,then choosing the DV-Distance algorithm as the main research object,we study on how to improve the algorithmperformance. We analyze the localizationprocess of the DV-Distance algorithm, and find that the algorithm localization error ismainly from three parts: error caused byRSSI ranging, error caused by using thecumulative hop distance to instead the Euclidean distancebetween the unknown nodeto the anchor node, error caused by calculation method. According to different errors,two improving algorithmsare presented in this paper.②Aiming at the error caused by using the cumulative hop distance to instead theEuclidean distancebetween the unknown node to the anchor node, this paper proposesa dynamic weighted DV-Distance localization algorithm forwireless sensor networks.This algorithm uses the correction mode on the basis of unknown node to ensure eachunknown node in the localization network has different correction factor. Also, in orderto improve the location accuracy, a dynamic weighted correction model is introduced to correct the accumulated hop distance between the unknown node and anchor node.The dynamic weighted correction model makes use of the distance, hops and otherinformation between the anchor nodes to calculate the correction factor, and integratesthe correction factors in different directions via the method of dynamic weighted. Thesimulation and experimentresults show that the algorithm performs better by using theabove two improvements.③Aiming at the error caused byRSSI ranging and error caused by calculationmethod, this paper proposes a DV-Distance improved algorithm based on particleswarm optimization. The RSSI range, due to the influence of environmental factors,has big range error. This paperuses the link quality indicator (LQI) to inertia filter themeasurement value of the RSSI, and proposes a RSSI ranging model based on LQI. Atthe same time, in order to reduce the error caused by calculation method, this paperuses particle swarm optimize algorithm to optimize the node localization.Throughiterative collaboration, the particle swarm algorithm makes full use of the effectiveinformation in the network. Combining with the characteristics of DV-Distancealgorithm, the fitness function andalgorithm parameter of the particle swarm optimizealgorithm are confirm in this paper.