Abstract

This chapter considers pilot symbol assisted linear mean square error channel estimation in the joint space-time-frequency (STF) domains, for multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) systems. Through the use of properly designed and placed pilot sequences, and three dimensional processing, the proposed STF channel estimator exploits the correlation that may exist in the three domains, yielding performance gains. Computer simulation results for a coded MIMO-OFDM system illustrate the performance advantages of our approach. As a specific example we show that in a MIMO-OFDM system employing a space-frequency orthogonal block code, such an STF channel estimation technique limits the bit error probability (BEP) degradation with respect to perfect channel knowledge to 2 dB at a BEP of 10-5 over the 3GPP fast-fading suburban macro environment with high transmit space correlation. Furthermore, we show that STF channel estimation reduces variability in BEP performance when the mobile speed varies.