Event Details
Reduced Rank Channel Estimation and Channel Prediction for Wireless Communication Systems
Presenter: Farnoosh Talaei
Supervisor:
Date: Tue, December 12, 2017
Time: 11:00:00 - 00:00:00
Place: Engineering Office Wing Room 430
ABSTRACT
Wireless communication for time-variant channels becomes more important with the fast development of intelligent transportation systems. This motivates us to propose a reduced rank channel estimator for time-variant frequency-selective high speed railway (HSR) systems and a reduced rank channel predictor for fast time-variant flat fading channels.
The proposed channel estimator exploits the correlation of the MIMO-OFDM HSR channel in time, frequency and spatial domain for representing it based on a 4D basis expansion model (BEM) which is spanned by limited number of time-concentrated and band limited GDPS sequences. A reduced rank LMMSE estimator has been derived for estimating the basis coefficients without requiring any information about the channel's correlation matrix. The simulation results demonstrate the robust performance of the proposed estimator for different delay, Doppler and angular spreads.
We also have developed a new sub-frame wise channel tracking scheme which considers a low-dimensional DPS-BEM for exploiting the variation of the channel inside each sub-frame and a Q-order AR model for tracking the variation of the basis coefficients through the whole frame. The optimal sub-frame length, number of basis and AR model order are derived for minimizing the prediction error of the whole frame. The simulation results validate the better performance of the proposed channel predictor compared to the minimum energy DPS based channel predictor and the conventional AR tracking scheme.