Hidden Markov Model Based Localization Using Array Antenna

We present a hidden Markov model (HMM) based localization using array antenna. In this method, we use the eigenvector spanning signal subspace as a location dependent feature. The eigenvector does not depend on received signal strength but on direction of arrival of incident signals. As a result, the eigenvector is robust to fading and noise. In addition, the eigenvector is unique to the environment of propagation due to indoor reflection and diffraction of the radio wave. The conventional localization method based on fingerprinting does not take previous information into account. In our proposal algorithm with HMM, we take previous state of estimation into account by comparing the eigenvector obtained during observation with the one stored in the database. The database has the eigenvector obtained at each reference point according to setting in advance. In an indoor environment represented in a quantized grid, we design the transition probability due to previous estimated position. Because of this, target’s movable range is obtained. In addition, we use maximum likelihood estimation method based on statics of correlation values. The correlation value is an indicator of pattern matching in a fingerprinting method. The most likely trajectory is calculated by Viterbi algorithm with above mentioned probabilities. The experimental results show that the localization accuracy is improved owing to the use of HMM.