Research On Indoor Pedestrian Positioning And Tracking Algorithms With Sparse Anchors

With the development of science and the improvement of communicationtechnology, location-based services (LBS) have drawn more and more attention.Indoor positioning is an important part of LBS, playing an important role in disasterrelief, shopping guide, hospitals monitor, mine personnel positioning etc. And it willhave more positive influence on people’s life and work.For indoor sensor deployment was very hard accomplished leaves no dead angle,and increasing density of sensors will increase the cost. Also because of thecomplexity of the indoor environment, signal’s diffraction and reflection is serious,even lead to the signal is unacceptable and difficult to locate. Traditional indoorpositioning algorithm mostly assume the area have full anchors or only consider thenormal use circumstances, which means the monitoring area’s each position has3ormore anchors’ signal. But in real life, we need to consider the special circumstancessuch as damage of some anchors in fire and earthquake or the deployment fails tofully cover the area. And obtain enough location tracking precision. Under thisbackground, the traditional trilateral localization algorithm accuracy will decreaseseriously when the number of anchors reduced, Considering the limitations of theexisting work, this thesis mainly solve intermittent positioning failure problem underthe sparse anchors deployment, make the location tracking algorithm is more adapt todifferent indoor location environment.First, to solve the problem of indoor positioning under sparse anchors deployment,put forward a method to solve the problem of positioning under the condition ofsparse anchors, based on the position fingerprint’s KNN algorithm estimate theposition and use particle filter, then through some facts and indoor topologyconstraints algorithm to modify the results. The experimental results show that themethod can effectively improve the indoor positioning accuracy.Second, under the background of the present indoor positioning is more and moretend to use multiple sensor fusion positioning., proposed a algorithm of inertial sensorassisted indoor positioning and tracking, which blend the anchors information, inertialsensor information’s dead reckoning method and the indoor topology through wallconstraints and corners constraints information to locate. And in full anchors region,insufficient anchors region and without anchor node region respectively take different tracking methods, using the different combination of available information, achievethe goal of seamless positioning. The experiment results show that the method caneffectively solve the location tracking problem in sparse anchors region, and can get abetter accuracy.In all, this thesis presents algorithms of indoor pedestrian positioning and trackingunder sparse anchors deployment, and the proposed algorithm is verified through thesimulation and experiments, the results show that the algorithms achieve goodlocalization accuracy. At last, the insufficiency of this thesis and the future work aregiven.