Code design for MIMO-OFDM(A) systems

Achieving higher link rates in support of newer services and increasing range are the two main goals for next generation indoor wireless LAN networks. On the other hand, outdoor fixed broadband wireless access (FBWA) is required to provide high-data-rate access comparable to wireline cost/performance considerations. For these high-rate, low mobility applications, MIMO-OFDM is regarded as an attractive solution. However, most of the existing space-time schemes are suboptimal in exploiting space-frequency diversity available in MIMO frequency-selective fading channels. In addition, MIMO designs for WLANs and WMANs ideally should maximize link throughput and diversity gain with a low decoding complexity receiver; there does not exist as yet a single desirable solution that achieves all three simultaneously. Therefore, MIMO-OFDM actually calls for new code designs. In the first part of this dissertation, two kinds of space-time coding methods are proposed for point-to-point MIMO-OFDM without the CSI requirement at the transmitter, i.e. rate-11 space-frequency block codes and rate-M space-frequency block codes. Both codes proposed are delay optimal codes which only need one OFDM block time for encoding and decoding. Rate-one SFBC can exploit the maximum diversity gain in MIMO-OFDM systems over frequency-selective channels. Rate-M space-frequency block codes is a full rate code which achieves a desirable trade-off among rate, diversity and complexity.; In the second part of the dissertation, we moved onto the multi-cell multiuser communication scenario in the presence of co-channel interference . With the emphasis on the downlink for future data-intensive applications, we propose a kind of MIMO scheme for TDD OFDM/TDMA that utilizes multiple antennas at the base station (BS) to avoid causing intercell co-channel interference through only local BS coordination and therefore can retain the high peak rate achievable for point-to-point single-cell communications. We describe several different levels of CSI availability at both the transmitter and the receiver that lead to different system architecture choices. The performance of rate-1 SFBC code developed in [1] with channel estimation is investigated via simulation; our results show that with pragmatic channel estimation schemes, multi-cell MIMO-OFDM is a good candidate for future high-rate applications. Moreover, using similar idea for multicell MIMO-OFDMA/TDM systems, we simplify the multi-cell resource allocation problem into multiple independent single cell resource allocation problems. Combined with transmit beamforming and adaptive modulation, we demonstrate that our cross-layer design scheme can effectively exploit multiuser diversity to improve aggregate system throughput.; 1rate-one is defined as transmitting one independent symbol per time slot or per subcarrier.