Fast Time-Varying Dispersive Channel Estimation and Equalization for 8-PSK Cellular System

Sang-Yick Leong
Jingxian Wu
Chengshan Xiao, Missouri University of Science and Technology
Jan C. Olivier

This document has been relocated to

There were 40 downloads as of 27 Jun 2016.


In this paper, a novel channel-estimation scheme for an 8-PSK enhanced data rates for GSM evolution (EDGE) system with fast time-varying and frequency-selective fading channels is presented. via a mathematical derivation and simulation results, the channel impulse response (CIR) of the fast fading channel is modeled as a linear function of time during a radio burst in the EDGE system. Therefore, a least-squares-based method is proposed along with the modified burst structure for time-varying channel estimation. Given that the pilot-symbol blocks are located at the front and the end of the data block, the LS-based method is able to estimate the parameters of the time-varying CIR accurately using a linear interpolation. The proposed time-varying estimation algorithm does not cause an error floor that existed in the adaptive algorithms due to a nonideal channel tracking. Besides, the time-varying CIR in the EDGE system is not in its minimum-phase form, as is required for low-complexity reduced-state equalization methods. In order to maintain a good system performance, a Cholesky-decomposition method is introduced in front of the reduced-state equalizer to transform the time-varying CIR into its minimum-phase equivalent form. via simulation results, it is shown that the proposed algorithm is very well suited for the time-varying channel estimation and equalization, and a good bit-error-rate performance is achieved even at high Doppler frequencies up to 300 Hz with a low complexity.