Study On 3D Node Localization Methods For Wireless Sensor Network Based On Compressive Sensing Under Single Sampling

Wireless Sensor Network(WSN)has the advantages of low cost,wide coverage and high precision measurement.Node localization technology of wireless sensor networks is one of the key technologies of WSN,which has been applied to health care,environmental monitoring,military reconnaissance and other fields.Compared with the traditional location algorithms based on time of arrival(TOA),time difference of arrival(TDOA),angles of arrival(AOA),the algorithm based on RSSI for WSN localization does not need time synchronization between nodes,and it is convenient to configure the wireless transmission module without additional hardware,which reduces the cost and is suitable for large-scale wireless sensor networks with more nodes.At present,some RSSI-based localization methods for wireless sensor network are proposed,however,these traditional methods still have some shortcomings:(1)these traditional methods only consider the localization in two-dimensional space,however the actual scene is mostly three-dimensional space,so these methods have some limitations in practical application;(2)it is impossible to achieve accurate node location in the case of fewer samples,especially single sampling;(3)the traditional location methods are not effective in low signal-to-noise ratio environment.In view of the shortcomings of traditional algorithms,the three-dimensional localization problem of wireless sensor network nodes is deeply studied based on compressed sensing theory under the condition of single sampling.Based on the indoor empirical propagation model,a three-dimensional localization model of wireless sensor network is established.Using compressed sensing theory and Bayesian theory,the location of unknown target nodes in three-dimensional space under single sampling condition is realized,which has high positioning accuracy.The work is supported by the Natural Science Foundation of Jilin(No.20180101329JC).The innovative work of this paper is as follows:In view of the fact that most of the traditional algorithms are only used in two-dimensional space,this paper expands WSN nodes localization in two-dimensional space to three-dimensional space.Combined with compressed sensing theory,a three-dimensional dictionary is constructed to realize the accurate location of WSN nodes,which is more suitable for practical application scenarios.In order to overcome the limitation of Nyquist sampling theorem that traditional localization algorithms need sufficient sampling number,a novel three-dimensional localization method for WSN nodes based on orthogonal matching pursuit(OMP)algorithm under single sampling is proposed.This method uses indoor experience propagation model as perception matrix and utilize OMP algorithm based on compressed sensing theory to reconstruct sparse signals.On the one hand,it reduces the hardware storage space needed for positioning,on the other hand,it has higher accuracy.Aiming at the problem of low anti-noise ability of the OMP algorithm,a new three-dimensional localization method for WSN nodes based on Bayesian Compressive Sensing(BCS)under single sampling is proposed in this paper.Combining Bayesian theory with compressed sensing theory,the signal can be reconstructed according to the prior information of the signal,and the signal can be reconstructed and positioned accurately in the case of low signal-to-noise ratio.The method of correlation vector machine can greatly improve the recovery speed of the algorithm,reduce the complexity of signal processing system and the power consumption of sensor nodes,and prolong the network life.