Research On Key Technology Of Localization For Wireless Sensor Network

Wireless Sensor Networks(WSN)integrates wireless communication technology,MicroElectro-Mechanism System(MEMS)technology,sensors,distributed information processing technology,etc.As a bridge of the Internet from virtual world to real physical world,it changes the way how people get along with nature.WSN has many aspects of technologies worthy of studying,and node localization technology is one of the core technologies in WSN.Accurate and effective node location is the basis of providing various detection location information services.It is necessary for us to design effective low-power node location algorithms for specific scenarios where the density,computing,communication and storage capabilities of nodes are limited.The wireless sensor network positioning technology is put into practical benefits in this paper by reducing the error in node positioning,and one WSN radiation source positioning technology with source and capacity of nodes limited based on the current mainstream node localization algorithms is proposed.In this paper,some localization techniques are applied to radiation source positioning,and some meaningful results are obtained.Aiming at the problems existing in the current node positioning algorithm in wireless sensor network,the positioning technology is deeply studied from aspects of energy consumption and positioning accuracy to reduce positioning error.The main research contents of this paper are as follows:1.On the basis of consulting a large number of literatures,it is firstly introduced that the research status of wireless sensor network at home and abroad,analyzes its network structure and characteristics,and introduces the protocol stack of wireless sensor network.The research significance of node localization technology is proposed.The PSO particle swarm optimization algorithm and machine performance optimization algorithm are introduced to illustrate the existing node localization algorithm,and the existing node location algorithms and radiation source location technologies are explained.At last,the evaluation criteria of node location algorithm performance are given.2.The original DV-Hop algorithm is studied and the positioning principle of the algorithm is analyzed.At last,the specific positioning steps is summarized.According to the defects of the algorithm,three improvements are present and one algorithm is proposed that improves DVHop localization algorithm based on RSSI ranging technique(NRSSI?DV-Hop).Firstly,the localization method of unknown nodes in one hop between anchor nodes are improved by applying RSSI technology.Secondly,the unknown nodes that have been localized will be upgraded as anchor nodes for later localization process so that the the proportion of anchor nodes will be promoted.Finally,the calculation method of the average distance per hop is improved,so that the distance estimation of the algorithm is more accurate,thereby the node positioning accuracy is improved in the WSN.3.In this chapter,a multi-class SVM positioning algorithm model based on NRSSI?DVHop is established.The NRSSI?DV-Hop algorithm is applied to it and combined with the traditional multi-class SVM positioning algorithm to improve the positioning accuracy of wireless sensor nodes.The idea of the SVM positioning algorithm is mainly to divide the communication coverage of the wireless sensor network node into several equal-sized blocks,and the block number of the anchor node and the hop count information of all nodes are used as the training parameters of the SVM to train the mapping model.With the mapping Model,the position coordinates of unknown nodes can be predicted.Compared with the NRSSI?DVHop algorithm and the traditional multi-class SVM positioning algorithm,from the experimental results,it is shown that the improved multi-class SVM positioning algorithm has a good positioning effect.4.Aiming at the shortcomings of RSSI ranging method which is vulnerable to environmental impact and large positioning error,a localization algorithm(PSO?BP)combining particle swarm optimization(PSO)and BP neural network is proposed.Based on this,PSO particle swarm optimization algorithm is improved,which is called CA-PSO.We propose one algorithm called CA-PSO?BP for the optimization of the distance value measured by the RSSI at the ranging node in the 3D deployment environment.It mainly includes the improvement of PSO algorithm,the dynamic calculation of parameter inertia weight and trust coefficient in PSO algorithm,the inertia weight and the asynchronous trust coefficient based on concave function,so that the algorithm can better optimize and then make the improved PSO algorithm(CA-PSO)applied to the BP neural network to take advantage of the PSO algorithm in search optimization.Through training,the training time of the BP network can be shortened,and the convergence speed of the algorithm can be accelerated,so that the BP network obtains the optimal initialization weight and threshold.Finally,the CA-PSO?BP algorithm is used to locate the unknown nodes in the three-dimensional space in WSN.This algorithm can significantly improve the positioning accuracy of the network.5.In-depth study of the network protocol and positioning algorithm based on wireless sensor network radiation source location,design and implement an efficient,energy-saving WSN radiation source positioning application system,meantime obtain the geographical location and radiation value through the GPS and radiation sensors carried by the anchor node itself.According to the positioning model,realize the positioning of the radiation source.Based on the radiation source location model,the MAC protocol and routing protocol of the classic wireless network are deeply analyzed,and the multi-channel MAC protocol for WSN and compatible with the IEEE802.15.4 standard physical layer(MC?MAC protocol)and a secure routing protocol(SF-Routing)are adopted.In the whole radiation source positioning system,the MC?MAC protocol and the SF-Routing secure routing protocol can improve the positioning accuracy of the node,and can more effectively reduce the energy consumption of the network node and prolong the life cycle of the network.