State space estimation of a wireless frequency-selective fading channel in Gaussian noise

Abstract

This paper investigates the possibility of using the state space optimal estimation techniques to estimate the inphase, quadrature and square envelope of a wireless multipath fading channel when the transmitted signal is wideband. These models are derived from the Doppler power spectral density using factorization and realization techniques. The simulations illustrate that tracking of the inphase and quadrature components of each path is possible, when the received signal-to-noise ratio is -3 dB. As the received signal-to-noise ratio increases to 20 dB, the inphase, quadrature and square envelope estimators of each path show the good tracking with respect to the mean square errors.