Dynamic Fingerprint Based Indoor Localization Algorithm In Wireless Networks

Indoor localization is one of the key issues in wireless sensor networks,at present,there are many methods used for indoor localization,like special equipment based localization technology and Wi-Fi signal ranging based localization technology,This type of localization technologies have the advantages of high accuracy,but it need to deploy special hardware device and the cost is high,so,the research at home and abroad mostly apply fingerprint localization to implement indoor localization.This technology has the advantages of low cost,easy implementation and high accuracy,therefore,fingerprint localization technology is a relatively popular and widespread indoor localization technology at present.In location fingerprint based indoor localization system,the received signal strength(RSS)from a set of Wi-Fi access points is used as a unique fingerprint to identify a specific location.However,indoor environment is much more complex and changeable compared to outdoor environment,which leads to larger localization error brought by outdated RSS fingerprints.And re-measuring RSS fingerprints for all locations to maintain a dynamically changing fingerprint database will cause high cost and complexity.Dynamic fingerprint could adapt the complex environment,and could be able to solve the localization problem effectively.The paper focuses on the problem that traditional static fingerprint which could not adapt to complex and changeable indoor environment,studying dynamic fingerprint indoor localization,designed to enhance the accuracy of indoor localization algorithm and reduce the localization cost.The main contributions of the paper are as follows:Focusing on the problem that the classic fingerprint is static but the indoor environment is dynamic,lead to the error of matching algorithm at executing possess,this paper proposes Feedback mechanism based dynamic update and adjustment of fingerprint,considering real time RSS of part of reference node,adjust fingerprint to dynamic change along with surrounding environment and node information.This method can solve the problem that the localization errors result of complex and changeable environment,thereby,enhances the localization accuracy.Simulation results show that the localization accuracy and the stability of this algorithm are higher than traditional CS based localization algorithm and fingerprint localization algorithm in dynamic indoor environment.Furthermore,this paper imports the Monte Carlo to study indoor tracking algorithm.Focusing on the large number of sample of traditional Monte Carlo localization algorithm,and the effect of many sample to location of unknown node is too small,this lead to low localization and high cost.To solve this problem,this paper proposes dynamic fingerprint based sampling optimizing Monte Carlo tracking localization algorithm,this algorithm optimizing the sample weight;make the sample node in the important region have larger weight,at the same time,feedback the information of this sample to fingerprint to update the fingerprint dynamically.This algorithm could reduce the sample that localization needs effectively,meanwhile,enhance the localization accuracy and reduce the cost.Simulation results show that this algorithm not only reduce sampling time effectively,but enhance the accuracy of mobile node localization in wireless sensor networks.