Research On Nlos Localization Algorithm Based On Intelligent Reflecting Surface Assistance

With the rapid development of mobile communication technology,location has gradually become an important function integrated in the communication system.The Internet,which is built on the physical layer signal transmission,is also moving toward the "Internet of Everything",so location information is an important attribute and label in the "Internet of Everything".Although the current positioning technology has made great progress based on the communication system.However,there are still various environmental factors that interfere and hinder the improvement of the positioning accuracy,making the bottleneck.Intelligent Reflecting Surface(IRS),as one of the research directions for future mobile communication,is an array composed of a large number of reflecting units,which can change the direction of the beam reflected by IRS to enhance the communication and positioning performance by using analog beamforming theory.Because of the characteristics of IRS,the electromagnetic wave signal transmission under Non-Line of Sight(NLOS)conditions produces great gain,and has the advantage of low power consumption and low cost.To better utilize the features and benefits of IRS,this paper firstly introduces the theoretical basis of electromagnetic reflection to derive the theoretical system of beamforming and channel representation of IRS technology,and then establishes a cascade channel model from the base station to IRS and then to the receiving terminal,takes the phase shift matrix of IRS reflection unit as an important part of the channel,proposes a new positioning architecture based on IRS and a signal reconstruction algorithm based on compressed sensing.Enables highprecision positioning in NLOS conditions and practical applications.Firstly,this paper transforms the channel matrix into a sparse representation on the angular domain,and proposes a hierarchical environment control-location solving joint localization architecture,which uses NLOS path signals for location estimation according to the logical relationships in the architecture combined with the proposed Complete Orthogonal Matching Tracking(COMP)algorithm.The simulation results show that the localization error is reduced by more than 50%compared to the conventional reflection model,and the solution success rate is always very high at the set transmit power,and the coverage of the beam blind area is enhanced so that all the localization areas can be kept with low errors.Secondly,for the scenario where the IRS location is unknown,the mobile terminal can still be localized using multiple IRS anchor points.The proposed architecture and method are further improved to adapt the scenario with unknown IRS location and derive the Cramer-Rao Lower Bound(CRLB)of the localization error to provide a reference for evaluating the performance of the proposed method.The simulation results show that the localization error can achieve close to the CRLB in the actual transmit power interval,and both performance indicators of localization error and solution success rate are significantly improved at low transmit power.Compared with the conventional goniometric algorithm,the proposed algorithm can still maintain a low error at low transmitting power.Finally,the effects of the dual IRS configuration and the parameters in the system on the localization results are analyzed to verify the wide applicability and robustness of the proposed method.