Joint angle and delay estimation (JADE) for fading multipath signals in α-stable noise
Balakrishnan Kannan, WJ Fitzgerald · 2003
In wireless communications, transmitted signals arrive at the base station via multiple paths. Estimating each path's angle of arrival (AOA) and propagation delay is necessary for several applications such as mobile localisation for emergency services and advanced handover schemes. In conventional JADE methods, the channel impulse response is estimated first from the received samples and then this impulse response is used to estimate AOAs and time delays of the multiple paths. We propose a Bayesian approach for joint AOA and time delay estimation of fading multipath signals in symmetric a-stable noise. Approximating /spl alpha/-stable noise by Gaussian mixtures, we derive a posterior probability density for the AOAs, time delays and the fade amplitudes given the received samples and the model order (number of multipaths). This posterior probability is then used with the Metropolis-Hasting algorithm to derive the samples for the AOAs, time delays and the fade amplitudes. Unlike the conventional subspace-based methods, this algorithm can be used with significantly lower number of samples.