Research On Localization Algorithm Based On Location Fingerprint

With the rapid development of pervasive computing,mobile Internet and Internet of things technology,indoor localization technology has received extensive attention from academic and industrial researchers.WIFI indoor location algorithm based on localization fingerprint is a relatively recognized indoor location method in academic circles,but the crux of restricting its application in real scene is that the WIFI indoor location algorithm based on the index intensity of received signal(RSS)is easy to be disturbed and fluctuated,and the fingerprint Library in the room is complex and working.And the localization accuracy is relatively low during continuous localization.How to avoid the interference caused by the multipath and attenuation caused by the wireless signal transmission,the high precision depicting the location of the user in the indoor environment has become a problem that the scholars need to solve urgently.In view of the existing problems of the location fingerprint algorithm.this paper mainly does the following work:the fingerprint localization method and the principle of compression perception are deeply studied and analyzed.Through theoretical analysis and experimental simulation,the WIFI indoor location algorithm based on the received signal exponential strength is easily disturbed.In the off-line stage,the theory of compressed perception is used to construct the off-line database to reduce the complexity of off-line acquisition.In the online localization stage,the vector similarity theory is used to improve the accuracy of the location.The experimental results show that the average error of the proposed algorithm is 1.4117 m,which is 20% higher than that of the WKNN Horus.The algorithm can effectively improve the accuracy and anti-jamming performance of indoor static localization problem.when the static localization problem is extended to the dynamic tracking problem,aiming at the poor tracking effect of the indoor location algorithm based on sparse representation and position correlation,the static localization algorithm is combined with the inertial navigation algorithm to achieve trajectory tracking.The inertial sensor data information is used to compress the sensing localization process,so that the system can adapt to the dynamic localization process.The experimental results show that the tracking degree of the simulation tracking path is relatively high,and the accuracy of the localization error is 98% in the range of 2.5 m.The algorithm is feasible.Experimental results and simulation results show that the indoor location algorithm based on sparse representation and position correlation improves the localization accuracy and anti-jamming performance of indoor localization.The dynamic fingerprint algorithm assisted by inertial measurement unit is used for target tracking.It also achieves a good fitting effect on the pedestrian track.The paper meets the requirements of the subject and has certain research value and practicability.