Research On Space-time Signal Processing And Channel Estimation For MIMO Broadband Wireless Communications

With the predominant advantage to combat channel fading and enhance data rates, Space-time coding becomes the promising technique for the next generation mobile communications. In the frequency- or time-selective fading channels, it is of theoretical and realistic signification to take full advantage of space-time coding to obtain the system full diversity gains with high bandwidth efficiency. The dissertation focuses on the space-time signal processing and channel estimation technique for the multiple input multiple output (MIMO) broadband mobile wireless communications. some innovative results have been gained as follows:Firstly, the dissertation lays the stress on the full diversity space-frequency coding applicable to the time-varying Doppler rapidly fading channels. Based on the basis expansion model, the nonlinear time-varying Doppler fading channel is dually converted into the linear virtual channel via time-frequency duality. Following this, two kinds of full diversity space-frequency coding schemes are proposed, both of which achieve full diversity gains including the inherent channel Doppler diversity. The first one of them is a concatenated space-frequency coding directly constructed from the general orthogonal space-time coding through repeating, which effectively combat the channel Doppler fading. Though the coding ratio is less than 1, it is even more easy to construct from the traditional space-time coding; Grouping the subcarriers in the Orthogonal frequency division multiplexing (OFDM) through the block matrix method, and generalized complex orthogonal designed on each block subcarriers, space-frequency-Doppler coded OFDM for the general MIMO system is constructed, in which the coherent decoding error only lies on the channel frequency impulse response differences between the consecutive adjacent subcarriers in each block subchannels and thus significantly alleviates the influences of the time-varying Doppler fading on the coding transmission. Furthermore, without increasing any transmit redundancy, it achieves the system full diversity gains with high bandwidth efficiency. Simulation results on the two space-frequency coding schemes approve that both of them have prominent performance to combat the channel Doppler fading.Secondly, based on the signal constellation size and channel impulse response order, grouping the input data sequences according to the single-carrier system inherent maximal diversity, the dissertation proposes a constellation-based single-carrier full diversity space-time block coding transmission scheme. By means of single-carrier transmission, separating and correspondingly processing the channel equalization and receive decision in the frequency and time domain respectively, the inverse discrete Fourier transformation (IDFT) and discrete Fourier transformation (DFT) in OFDM only appear in one terminal of the wireless links, which significantly simplifies the wireless mobile hardware requirement, and the inherent limitations in OFDM, such as high peak-to-average power ratio (PAPR) and more sensitivity to carrier frequency offset, are canceled or significantly alleviated with lower complexity equivalent to OFDM system. it also has no requirement for the traditional linear redundant precoding, and thus significantly decreases the computational complexity with even high bandwidth efficiency.Thirdly, the dissertation discusses the single-carrier differential space-frequency coding transmission over the time-varying rapidly fading channels. Based on the essential principle of the base square differential space-frequency complex orthogonal coding matrix, and taking use of the relation between the nonsquare and square complex orthogonal coding matrix, single-carrier single-block differential space-frequency coding transmission over the time-varying rapidly fading channel is proposed. It is able to differential space-frequency coding transmission with arbitrary number of transmit antennas and be effectively decoded without any channel state information estimation. Furthermore, it transmits the encoded signal matrix only within one OFDM symbol period regardless of the number of transmit antennas, which significantly alleviates the influences of the channel time-varying on the differential coding transmission and can thus robustly and reliably supports high data rates.Finally, analyzing from the informational theoretical entropy opinion and the maximized minimal theorem, based on the basis expansion model, optimal pilot assisted channel estimation sequence design criteria and scheme for the MIMO time-selective fading channel is presented. Relating the channel estimation parameters with the time-invariant Doppler bases, the proposed scheme achieves the excellent estimation performance in the sense of the maximized minimal system capacity with lower complexity and bit error ratio. The dissertation also proposed a jointly space-time-frequency channel estimation algorithm by facilitating the pilot symbols space-frequency block coding. Space-frequency coding transmission on the input known pilot data sequences and directly estimating the channel state information in the receiver through the inverse space-frequency decoding, each block subchannels frequency response is effectively estimated. Following this, Winner filtering interposing according to the channel autocorrelation matrix decomposition, it finally leads to the jointly space-time-frequency channel estimation. The proposed algorithm has no requirement of the high-level matrix inverse computation, which significantly decreases the channel estimation computational complexity. Furthermore, it makes the best of the channel time-frequency correlation, and thus has even robust and effective channel estimation performance.