Research On Channel Feedback For Multiuser Diversity MIMO Systems

Future wireless communication systems require higher transmission rate and guaranteed Quality-of-Service (QoS), however, power limitation, bandwidth and complexity restrictions are the major challenges for its realisation. Therefore, the main research topics in 4G are technologies providing higher performance and greater spectral efficiency. Muliple-input multiple-out (MIMO) is considered as one of key techniques for 4G system as it has proven its protential to offer high spectral efficiency as well as link reliability without additional power and frequency expenses. To date, a lot of research have focused on the single-user point-to-point scenario where the transmitter and receiver each has antenna arrays. More recently, attention has shifted to multiuser MIMO systems especially for multiuser diversity. Our research concentrates on channel feedback problems for multiuser diversity MIMO systems.Several work/research have already proved that multiuser diversity introduced by the random dispersion of users in a network can radically increase the performance or capacity of MIMO systems。However, in the past literature, the multiuser diversity is mostly increased by enhancing adaptive resource allocation algorithms, yet neglecting or simply ignoring to investigate the channel feedback problem. Most importantly, channel feedback is absolutely essential for achieving adaptive ressourve allocation, so there could not be any multiuser diversity gain if lacking feedback information.First of all, we explore the limited channel feedback scheme applied to Orthogonal Space Time Block code(STBC) MIMO systems. Multiuser diversity can then significantly improve the wireless communication system performance by exploiting the feedback information that conveys the channel quality. However, while using multiple subcarriers(OFDM), feeding back all the Channel Quality Information (CQI) per user, per subcarrier would waste a substantial amount of uplink resource and therefore decrease the spectrum efficiency. In order to reduce the feedback overhead, a quantization method is proposed by quantized value used to indicate the modulation level instead of the full values of CQI in the second chapter, and the achievable spectrum efficiency shows no loss compared with the perfect case. This quantization method is then separately combined to three different feedback schemes (Threshold feedback, Best feedback and their combination called Combined feedback) and the respective close-form expression of the average spectrum efficiency is derived. Finally, The calculation of the optimal feedback parameters is obtained from two aspects of the feedback channel capacity and the capacity relative loss. Extensive simulations are carried out in order to evaluate these proposed strategies and the results satisfactorily match with the numeral analysis very well. Under the same performance requirements, the proposed limited feedback schemes can greatly reduce the feedback overload when the feedback parameters are choosen accordingly. Among these methods, the Combined feedback scheme depicts the best performance and therefore can be (partially) applied in the practical system design.STBC-MIMO can only obtain space diversity gain, but the use of beamforming can acquire space diversity gain and antenna array gain. Unfortunately, optimal performance requires either complete channel information or knowledge of the optimal beamforming vector; both are hard to realize. In this correspondence, a quantized maximum signal-to-noise ratio (SNR) beamforming technique is proposed where the receiver only sends the label of the best beamforming vector in a predetermined codebook to the transmitter. By analyzing the distribution of the optimal beamforming vector in independent and identically distributed Rayleigh fading matrix channels, D.J. Love proved for the first time the equivalence between the codebook design and the Grassmannian line packing problems. After applying several approximations, the density of Grassmannian line packings is derived and used to develop the codebook size upper and lower bounds given a SNR or capacity loss. Improvements are made on the vector quantization LBG(Linde,Buzo and Gray) algorithm, the specific steps for producing the initial codebooks based on split method(SM) are also presented and their comparison is obtained by means of simulation. Spatial multiplexing is a common technique for precoded MIMO systems where independent information streams are sent over different transmit antennas. However, spatial multiplexing is sensitive to ill conditioning of the channel feedback information. In chapter III, we propose a quantized precoding system where the optimal precoder is chosen from a finite codebook known to both user and basestation. The index of the optimal precoder is conveyed over a feedback link. Criteria are presented for selecting the optimal precoding matrix based on the error rate and mutual information for different receiver designs. Codebook design is proposed for each selection criterion by minimizing a bound on the average distortion. The codebook design criteria are shown to be equivalent to packing subspaces in the Grassmann manifold using the projection two-norm or Fubini–Study distances.The multiuser diversity gain increases not only with the number of users, but also with the dynamic range and rate of the fluctuations and is thus limited in environments with little scattering and/or slow fading. In such environments, we propose the use of multiple transmit antennas to induce large and fast channel fluctuations so that multiuser diversity can still be exploited. The scheme can be interpreted as opportunistic beamforming. Chapter IV focuses on channel feedback for opportunistic beamforming MIMO systems. There is a fundamental tradeoff between quantization errors and the orthogonality property in designing codebook. Giving priority to quantization errors, our results reveal useful design guidelines for the split of feedback bits for CQI quantization and beamforming vector quantization. On the other way, super-codebook is proposed giving priority to orthogonality property, and its size is limited according to the probability of allocation outage.Channel feedback is essential for realizing multiuser diversity and multiuser scheduling directly influences the multiuser diversity gain. In chapter V, by appling the effective bandwidth theory of wireline network on wireless domain, a multiuser scheduling algorithm for MIMO-OFDM time-frequency structure frame is introduced. This algorithm do not only concern the channel state information(CSI) in the physical layer, but also pay attention to priority, fairness and delay QoS at the date link layer, which would be more effective compared to the conventional layered structure. At last, another proposed scheme utilizes the property of the proportion fair scheduler and opportunistic feedback. The numerical results show that the feedback overload is minimized without capacity loss.From the dissertation we can find that, if correct channel feedback algorithms are chosen for specific transmission strategies on MIMO systems, multiuser diversity inherent in environment will be utilized effectively to provide system-level improvement.