Research On Precoding Techenologies For MIMO System

The multiple-input multiple-output (MIMO) system can significantly improve the system's spectral efficiency and the link reliability by using of the multiple antennas at the transmitter and receiver. In MIMO system, the cochannel interference (CCI) is the main obstacle to acquire the tremendous performance gain. On the basis of the channel state information at the transmitter (CSIT), the linear precoding and decoding scheme can mitigate the CCI and achieve the system performance gain.We begin with the brief introduction to the precoding in single-user MIMO and multi-user MIMO system in chapter two and chapter three, respectivly. After that, in chapter four and chapter five, we concentrate on the innovative parts. In chapter four, we introduce the maximum signal-to-leakage ratio (SLR) and maximum signal-to-leakage-and-noise ratio (SLNR) related schemes. However, the original scheme allocates every user the same transmit power and takes the matched filter to decode the receive signals, which results in the limited bit error rate (BER) performance. With the antenna correlation at BS and the channel estimation error for every user, we design a modified matrix by minimizing the system Mean Square Error (MSE) after maximizing the SLNR under the total transmit power constraint. At each user, a linear decoder is calculated based on the MMSE criteria in the presence of imperfect channel state information (ICSI). Due to the dynamic power allocation during the parallel data streams and the linear MMSE (LMMSE) receiver, the proposed scheme can mitigate the residual interference induced by ICSI and improve the system BER performance efficiently.In chapter five, we proposed an effective and unified transmit power allocation (TPA) scheme for the Singular Value Decomposition (SVD) -Assisted multi-user multiple-input multiple-output (MU-MIMO) downlink (DL) and uplink (UL) transmissions in the presence of ICSI. The existing power allocation policies for the SVD-Assisted MU-MIMO system, such as equal power allocation (EPA) in the DL and maximum signal-to-noise ratio (MSNR) policy in UL have been developed under the perfect CSI case. However, the EPA scheme doesn't exploit the CSI and ignores the BER difference among all singular values meanwhile the MSNR method neglects the noise enhancement caused by the decoding operation at Base Station (BS). Aiming at solving these problems, the proposed TPA scheme takes the smallest singular value and the noise enhancement into account and derives an unified expression for DL and UL. Under the ICSI, the proposed scheme can exploit the power allocation operation to mitigate the residual interference induced by ICSI and improve the system BER performance efficiently.