Channel Estimation And Data Detection Technologies In MIMO-OFDM Systems

MIMO systems, having multiple antennas at transmit and/or receive ends, can increase channel capacity and provide very high spectral efficiency while using the same total power and bandwidth as SISO systems. In ideal conditions, the capacity of the MIMO channel can be improved linearly with the number of transmit/receive antennas. So MIMO becomes one of the key technologies for high data rate and performance in 3G mobile communication systems and beyond. The performance of detection algorithm for a MIMO system is one of the key criteria to judge whether it can be used in practice. And the research of detection algorithms with high performance and low complexity is important in MIMO field.In the future broadband MIMO wireless communication systems, the transmitted signals suffer from frequency selective fading, which makes the traditional equalizer become extremely complex. Being a special case of multi-carrier transmissions, the OFDM technology has the ability of overcoming the multi-path fading and reducing the ISI in wireless channels. Because the future mobile wireless communications will use wider wireless bandwidth to transmit diversified signals, OFDM will inevitably become one of the key technologies in data transmission. Therefore, MIMO-OFDM as a promising technique for high bandwidth efficiency and easier realization obtains great interest from many researchers and companies.Usually, there are two methods to combine OFDM with MIMO. One is that spatial division multiplexing is realized by multi-antennas to increase the data rate. The other is that spatial diversity is realized by multi-antennas in special ways to improve transmission reliability and get better BER performances. Channel estimation and detection arithmetic with high performance and low complexity are research hotspots in MIMO-OFDM systems.OFDM systems are sensitive to frequency offset, so coherent detection is applied. Therefore,channel estimation is important to get better performance. The channel estimators are usually classified into two types. One is decision feedback estimation; the other is pilot symbol assisted estimation. The latter is in common use while considering the complexity of the receivers and performance of the systems. According to some criteria such as minimizing certain error, least square (LS) and minimum mean square error (MMSE) estimators are mostly used. In practice, it's hard to design an estimator with low complexity and high channel track performance, especially in MIMO-OFDM systems. In Chapter 3, beginning from the design of optimal pilots, the channel pulse response estimators in MIMO-OFDM systems are investigated on time-variety frequency selective fading channels.Space-time codes combine transmit diversity in space-domain and channel coding in time-domain. Among them, space-time block code is in wide use for its low complexity. There are two ways to apply this technique in OFDM systems. One is to encode in both space and time domains, the other in both space and frequency domains. In Chapter 4, a sort of channel frequency response estimation is investigated based on the special structures of the transmitted signals. BER performances of STBC/ SFBC-OFDM systems under ideal channel estimation are also investigated in this chapter.Under flat fading channels there are mainly two classes of detection techniques: optimal and suboptimal ones. Although maximum-likelihood detecting (MLD) is the optimal detecting algorithm, its extremely high complexity excludes its application in practical multi-antennas systems, especially when high order modulation and large number of antennas are employed. So detectors with resemble BER performance and lower complexity are research hotspots, such as sphere decoding (SD) and suboptimal detectors. In the beginning of Chapter 5, we investigate the MLD and SD in VBLAST-OFDM systems on time-variety and frequency selective fading channels. Then suboptimal detectors are presented, includingzero-forcing (ZF) detector, MMSE detector and classical successive interference cancellation (SIC) and ordered successive interference cancellation (OSIC) techniques based on ZF and/or MMSE.