Performance of Discrete Wavelet Transform-based deconvolution applied to multipath channel estimation

In this paper a technique for Discrete Wavelet Transform (DWT)-based deconvolution is used to estimate the impulse responses of multipath communications channels. The performance of this new technique is evaluated for the cases of five different multipath channel Power Delay Profiles (PDPs). Both slow and fast-fading variations of all five channels are considered, along with Additive White Gaussian Noise (AWGN) resulting in E_b/N₀ ratios from 0 dB to 30 dB. Estimation errors are obtained from Monte Carlo simulations. The results show that channel estimation using DWT-based deconvolution is comparable in performance to that obtained when discrete timedomain deconvolution is used.