Research And Implementation On Acoustic Self-Calibrating Indoor Localization System Via Smartphones

The needs of indoor localization and navigation proposed by large underground parking,shopping mall,airport,subway station and smart factory,bring new and higher requirements for indoor localization systems,such as high accuracy,hardware compatibility,low cost for commercial application,instantaneity and fast data update rate etc.Indoor localization technologies have become a new and hot research topic both at home and abroad.The result shows that if we want to put indoor localization into practice,we have to figure out several problems including avoiding the clock synchronization under the premise of high localization accuracy and instantaneity;finding an adaptive method to overcome multipath in different scenarios;achieving self-calibration of anchor networks not affecting the localization accuracy.To solve the problems above,the main work and contributions of the paper include the following aspects:1)We propose a new TOA localization method,which can avoid clock synchronization and improve the instantaneity.The method only needs once acoustic signal interaction including twice signal broadcasting,then we can get all TOA measurements between users' smartphone and each anchor node,which can calculate the users'position in high accuracy.2)We propose a robust TPSN ranging measurement to solve the problems including the frenquency offset caused by the bad frequency response and the weak consistency of anchor nodes' microphone,the doppler effect caused by the movement of users' phone and the multipath effect.We infer the causes mathematically in detail and analyze the influence on the time delay estimation.Finally,we propose a parameter estimation method combining with a new multipath mitigation method based on FrFT,which improves the robustness of the system.3)We propose a new self-calibrating approach based on acoustic TPSN ranging and Semidefinite Programming(SDP)algorithm.Through the interaction of every anchor node in the network,the new approach helps to solve the problem of self-calibrating in large-scale anchor network,and also helps to remove the heavy maintenance requirements afterwards without extra hardware devices.4)We implement the system with commercial smartphone platform,which consists of a special-designed anchor network,smartphones installed with real-time app inside,and a backend server for processing.The system can support the Android and IOS at the same time.Simulations and experiments have been performed,and the results show that the system has nice indoor localization performance,which can ensure the mean error inside the 26cm in office scence and inside the 39cm in underground parking scence under 90% probability.