Research On Multi-sensor Network Based Indoor Positioning And Tracking

Location-based Service(LBS) has application requirements in many fields such as life, production, medical and military. In LBS, how to realize low cost, low complexity and seamless positioning system which has robustness in any environment and condition is one of the problems to be solved. Global Positioning System(GPS) has a good positioning performance in outdoor place under good weather, but the satellite Positioning System cannot use in bad weather places and indoor places. Indoor positioning and tracking system, therefore, has become a hot area of research for the recent years.This paper has researched two aspects of the target location and tracking problems in wireless muti-sensor network(Cooperation Network).The first part has researched correction problem of the attenuation coefficient in lognormal model, based on Zig Bee network as the experimental platform, and presented a dynamic parameter correction algorithm of RSSI indoor localization. This research mainly focuses on wireless channel, and the premise for the problem is prior probability statistics. The application scenario for this algorithm is in places which anchor nodes’ density is large, their location fixed for long-term, and target node sensors moving slowly. The second part has researched the indoor location tracking problem of fusing the zero velocity detection-inertial navigation positioning and ranging positioning, based on the inertial measurement unit(IMU) simulating data, and proposed a new algorithm of using Kalman filter as a fusion tool. This research mainly focuses on information fusion, and the hypothesis is the premise of posterior probability statistics. The application scenario for this algorithm is in places which anchor nodes’ density is small, their location can deploy in a short time, and target node sensors’ moving speed isn’t limited. These two parts have inner link, the former part of the study results can be used as the initial estimates of the latter part’s filter state vector position variable. In addition, the third part further designs the software and hardware implementation scheme of the localization system in wireless sensor networks, and a lot of locating and tracking experiments have been done in the real application scenario with the prototype system.Field experimental results show that the method of fixed channel attenuation coefficient proposed in the first part can change channel parameters follow the environment changes in real-time, and obtain steady target location estimate compared with the traditional method.Next, numerical experiments show that the information fusion method proposed in the second part can effectively combine the continuity of inertial navigation system positioning results and ranging positioning results without cumulative error, can obtain steady target tracking results compared with the non-fusion method. Finally, according to the experimental results of the prototype system in the third part, the system has good practical effect when used in the squared-off spaces, and it can be used as the experiment platform of further study the information fusion algorithm as the second part presented.