Hybrid TDOA/AOA Mobile User Location with Artificial Neural Networks

Abstract

This paper proposes a mobile user location scheme with artificial neural networks. The neural network is trained using the second order learning algorithm (Extended Kalman Filter) because of its superiority in the learning speed and mapping accuracy. And then, the location scheme combines the time difference of arrival (TDOA) measurements from the forward link pilot signals with the angle of arrival (AOA) measurement from the reverse link pilot signal. High chip rates in wideband code-division multiple-access (CDMA) systems facilitate accurate TDOA measurements, and a smart antenna used at the home base station (BS) can provide accurate AOA measurement in a macrocell environment. A two-step least square location estimator is developed based on a linear form of the AOA equation in the small error region. Numerical results demonstrate that the proposed hybrid TDOA/AOA location scheme gives much higher location accuracy than TDOA only location, when the number of base stations is small and/or when the TDOA measurements have a relatively poor accuracy.