Research On TOA Estimation Algorithm For Wireless Localization

The time of arrival(TOA) estimation method is very suitable for application of complex indoor multipath propagation environment. Main factors affecting the estimation precision of time of arrival(TOA) signal are noise, multipath interference, out of synchronization between transmitting and receiving system and non-line-of-sight, etc. Now, a fundamental problem is how to accurately estimate the time delay of direct path signal.This thesis mainly aims at TOA estimation algorithm in wireless localization to carry out in-depth research. Firstly, commonly used TOA estimation algorithms are studied by establishing appropriate multipath channel model. Then, based on the analysis of the existing TOA estimation algorithms, several new improved TOA estimation algorithms are proposed. It respectively is time of arrival estimation based on improved Teager-Kaiser operator, the maximum search slope(MSS) algorithm and the hybrid TOA estimation algorithm based on correlation.The improved Teager-Kaiser operator algorithm reduces the effects of noise on output signal by accumulating the high power of correlation curve and estimates TOA combing with the Teager-Kaiser operator. This algorithm analyzed the complexity of the algorithm and put forward the corresponding scheme to reduce the complexity by limiting the range of data. The rationale of MSS algorithm is that the rate of change of correlation curve corresponding to the location of TOA is the biggest,and the algorithm realizes TOA estimation by searching the corresponding "mutations" position. Considering the effects of noise on the TOA estimation, the hybrid TOA estimation algorithm is proposed. The algorithm combined the wavelet denoising with correlation function to realize the TOA estimation. A plurality of differential curves can be obtained according to the correlation curve. The wavelet coefficients corresponding to the TOA value can be implemented by means of the wavelet decomposition; besides, the other coefficients are set to zero. After the reconstruction of wavelet, it can estimate TOA using the method of superposition average. It is shown from the simulation results that the performance of the various improved algorithms is better than that of traditional TOA estimation algorithms. Especially under the condition of serious noise, the proposed algorithm has improved TOA estimation precision greatly.