The Research Of Indoor Localization Algorithm Based On AoA/PDR Fusion In WLAN Environment

With the improvement of life standards as well as the development of technology,people’s demand for LBS(Location Based Services)is increasingly strong.As a complement for outdoor satellite positioning technology,indoor positioning technology has always been attracting significant attention.Research findings from all kinds of indoor positioning solutions have proved the fact that a single positioning system is difficult to satisfy the accuracy requirements of LBS in indoor environment.In addition,the fact also indicates that what makes product development more cost-efficient is to make use of existing resources without additional hardware devices.Based on above reasons,this thesis designs an indoor integrated positioning system that combines WLAN angle estimation positioning method and multi sensor navigation technology,which is more robust and accurate compared with single technology based positioning system,and more costefficient compared with traditional WLAN/MEMS fused localization system.First of al,combining the knowledge of channel state information and orthogonal multi carriers,the proposed system constructs a large array of virtual antennas in WLAN environment which reduce the requirement for deployed antennas in complex environme nt.Based on two-dimensional spatial smoothing,super-resolution parameter estimatio n,and joint angle of arrival of incoming wave and time of flight,the interference problem in coherent signal source is effectively solved.In addition,this thesis studies the weight-based direct path detection algorithm,establishes the corresponding path judgment model,and adopts the affine propagation clustering algorithm to provide the basis for direct path detection.Meanwhile,Joint localization algorithm with multiple access points has been intensively studied.in this thesis.Secondly,the pedestrian dead reckoning with multi sensor is studied,and the accelerometer information is utilized to detect and estimate pedestrians’ gait and step length.Based on the course angle estimation by quaternion method,the state equation is established with gyroscope,and the observation equation is constructed with accelerometer and magnetometer.Finally,by considering two separate positioning systems comprehensively,the useful information of inertial navigation and WLAN angle estimation is integrated into the proposed navigation system,through extended Kalman filter.Results show that the positioning accuracy of proposed algorithm is higher than that of two separate systems.In addition,compared with the WLAN fingerprint and MEMS fusion system,the proposed approach can achieve satisfying localization performance in a cost-efficient manner without site survey and has a higher promotional value.