Indoor Positioning Algorithm Research And System Implementation

Human demand for location-based services is more and more strong, and get the location of the service object is one of the core and the key problem. GPS positioning system is the most widely used among positioning technologies, however, GPS positioning system is used in outdoor environment. Due to the influence of the wireless signal masking, fading and multipath effects, The performance of GPS positioning system in indoor positioning is not very good. Therefore, it is very important to provide accurate and reliable location-based services in the 1 the indoor environment.,This thesis focuses on two kinds of positioning methods:one is the sound source localization method based on microphone array the other one is Rssi fingerprint identification method based on Zigbee.In the method of sound source localization based on microphone array, the location method based on TDOA (Time Difference Of Arrival) is mainly studied. In the location method based on TDOA, the most important is to calculate the time difference between the sound wave reaching the reference microphone and the other microphones. In this paper, a new method of calculating time delay is presented: parameter identification method, then comparing with traditional cross-correlation method through simulation, and the result show that Two algorithms have their different advantages and can be used in different occasions according to different requirements.It is mainly divided into two stages for Rssi fingerprint identification method based on Zigbee, the first stage is to establish a fingerprint database stage, the second stage is positioning stage. In the establishing a fingerprint database stage, first you need to collect RSSI fingerprint data point by point in the location area, after filtering datas you can save datas to database. In the positioning stage, location experiment was carried out by respectively using nearest neighbor method, the weighted k nearest neighbor method, the specific area weighted K nearest neighbor method. In the end, useing the map matching method combined with fingerprint orientation method to complete online tracking process in the simulation indoor environment.