关键词: 非视距传播, 误差分布模型, 卡尔曼滤波器, 混合定位

Abstract: A novel TDOA（Time-Difference-Of Arrival）/AOA（Angle-Of-Arrival） wireless position scheme in NLOS（Non-Line-Of-Sight） environment which uses two Kalman filters is proposed. According to a thought that the type of normal distribution density curve is the optimal fitting for exponential distribution density curve in the least-squares sense, TDOA error model is established. First, a Kalman filter is used to preprocess the TOA（Time-Of Arrival）measurements for eliminating the NLOS errors. Then these preprocessed measurements are input to the TDOA/AOA hybrid location which uses another Kalman filter. The simulation results show that the method of positioning error is better than pure TDOA location method and the TDOA location which error is exponential distribution model.

Key words: Non-Line-Of-Sight（NLOS） propagate, error distribution model, Kalman filter, hybrid positioning