Research On Indoor Positioning And Target Tracking Algorithm Based On MD5-RSSI Probabilistic Fingerprints

With the rapid development of Internet and Internet of things technology,Location-Based Service(LBS)plays an increasingly important role in human's life.As the core part of LBS,indoor positioning and tracking technology has been paid much attention by researchers.Due to the maturity of Wi-Fi related technology and the perfection of infrastructure,indoor positioning and tracking based on Wi-Fi has been a hotspot in resent years.This paper studies the indoor positioning and tracking technology based on the traditional technology.The main work and innovations are as follows:1.This paper proposes the indoor positioning algorithm based on MD5-dr GP in order to further reduce the storage space occupied by the reference point fingerprints' database and improve the efficiency of online location of the target node and analyzes the shortcomings of the current indoor positioning algorithm based on RSSI(Received Signal Strength Indication)probability fingerprint.The algorithm optimizes the reference point fingerprints' database.The global RSS map of the indoor positioning algorithm based on dr GP(distributed recursive Gaussian Process)was managed through the MD5(Message Digest 5 Algorithm)and was transform into MD5-RSSI probability fingerprints.Each fingerprint of the database was compressed as 128 bits and could be transformed into 32 hexadecimal numbers.The storage of the reference point fingerprints and the time consumed by matching algorithm are reduced at the same time,thus improving the efficiency of indoor positioning algorithm.The simulation results showed that,compared with the algorithm based on dr GP,the proposed algorithm improves the positioning efficiency without reducing the positioning accuracy.The performance of positioning algorithm is improved to some extent.2.A federated filter combined with RSS map was used in indoor target tracking in order to solve the problem of indoor target tracking based on UKF(Unscented Kalman Filter)and PF(Particle Filter).The proposed algorithm conducts a regional planning for indoor RSS map based on terrain environment.In order to realize an improvement on the efficiency of tracking on the premise of not reducing tracking precision,it proposes an idea that using the tracking algorithm based on improved UKF in the case of the target is in the area where it can only make approximate linear motion,and using the tracking algorithm based on loose-coupling PF in the case of the target is in the area where it can make nonlinear non-gaussian motion.The results of the indoor positioning algorithm based on MD5-dr GP are substituted into the observation of the filtering algorithm,and the status update method is changed accordingly.The simulation results showed that,compared with the tracking algorithm based on loose-coupling PF,the proposed algorithm reduced the calculation amount and increase the efficiency of tracking without reducing the tracking accuracy.3.This paper proposes the modified indoor multiple targets tracking algorithm based on MHT(Multiple Hypothesis Tracking).On the basis of the method that a federated filter combined with RSS map using in indoor target tracking,this algorithm solved the problem of data correction between observation and target tracks combining with the MHT algorithm and realized the multi-target tracking in complex indoor environment.The simulation results showed that the improved algorithm can track accurately when the target state is in nonlinear system compared with the traditional multi-target tracking method based on MHT.