Research On Localization Of High Speed Moving Node Based On Wireless Sensor Network

Wireless sensor network integrates embedded computing technology,sensor technology,wireless communication,micro electromechanical system,which is a new type of network developed rapidly since 21 s Century.WSN is composed of a large number of sensor nodes in the monitoring area.Firstly,the nodes can monitor,sensor and collect the target information through cooperation.Secondly,the collected data will be transmitted by wireless communication technology.Finally,the data will be transmitted to the user terminal through the way of multi hop self-organization among nodes.In recent years,WSN has the advantages of low power consumption,low cost,distributed and self-organization with the research of the wireless sensor and improvement of the WSN network protocol,so it is widely used in military,national security,environmental monitoring,target tracking,smart home,health care and space exploration and other fields.Recently,the WSN applications mainly focus on static target localization,but the research is few on the mobile WSN localization.If add a few of high-speed mobile nodes into static WSN,we can't only increases the application range and strain capacity,but also satisfy our needs about position information in modern times.Therefore,node localization is very great value and theoretical significance both mobile node localization in the general environment and the node localization in high speed environment.The dissertation researches the existing classical moving node localization algorithms based on WSN,mainly for the algorithm's problems for further study.According to the range,localization accuracy and energy consumption and other issues,combining of all kinds of mobile node algorithms' characteristics,this dissertation proposes an algorithm for mobile node localization in high speed environment,using the fixed sampling range and the the characteristics of high positioning accuracy and independence in a short time,so gets the accurate positioning information.The main contents of dissertation are as follows:(1)Aiming at phenomenon of Monte Carlo localization(MCL)algorithm,which the sampling range is too large,the dissertation puts forward a new kind of algorithm model which is combined the linear deployment with a deployment of the equilateral triangle.The model gets a refined overlapping area by using the characteristic that the sensing range of the anchor nodes can be overlapped,replacing the traditional MCL to the radius of the composition of the sampling range.(2)According to the MCL algorithm of particle degeneracy problem caused by resampling and dead reckoning(DR)error accumulation problem,the dissertation proposes an improved DR algorithm.The unknown node with high speed collects the anchor nodes information as the reference points of correcting the positioning error.In the discrete time period,the reference points are used to correct the positioning error,and the trajectory of the moving unknown node can be depicted from the positioning error.(3)Taking the problems of signal node invalid linear deployment and mobile node localization in high speed environment,the dissertation proposes data backhaul transmission technology and abnormal data eliminating technology,and constructs a data processing center in the algorithm model.When the high-speed moving unknown nodes collect information,the data will be transmitted back to the data processing center directly,which solves the problem of one-way transmission of data in linear deployment.After comparing the correction factor and the threshold value,the algorithm will exclude error which is greater than threshold,and get the more accurate positioning.(4)The simulation is carried out with research on localization algorithm of high speed moving node based on WSN,and compares and analyzes with different uniform motion,different speeds and different reference points.Finally,by comparing and analyzing the algorithm of position error in the one-dimensional space and the two-dimensional space,the result verifies that the localization error in one-dimensional space is less than that in two-dimensional space,and the algorithm is also suitable for the localization of high speed mobile nodes in one-dimensional space.