Single-carrier Frequency Sub-multiple Access Cellular System, Multi-user Detection Technology

As core physical layer technique for uplink of LTE systems, single-carrier frequency-division multiple access (SC-FDMA) has the main advantage in terms of peak-to-average power ratio (PAPR). Its capacity performance can be further enhanced if we combine SC-FDMA with other advanced MU-MIMO techniques such as SDMA. Currently, most research of MU-MIMO SC-FDMA focuses on its signal design and detection. Especially when considering the virtual MIMO structure of LTE uplink system, multiuser detection method at the base station can be a key factor which could affect the system performance. In this dissertation, we mainly research on the performance of some frequency multiuser detection algorithms in MU-MIMO SC-FDMA system. Different from the existing research, we propose a method to combine the multiuser detection methods with user pairing in visual MIMO system. Finally, we verify the performance of all the algorithms by computer simulation. We arrange the structure of the dissertation as follows.In chapter one, we briefly introduce the related technical background and current development trends of MIMO SC-FDMA system, and the main content of the dissertation.The pairing technology will affect the performance of virtual MIMO in the uplink, so in chapter two, firstly, some criteria for user pairing for V-MIMO in the uplink of LTE are introduced. Then we investigate the frequency domain equalization (FDE) algorithms used for MU-MIMO SC-FDMA system. This includes frequency domain linear equalization (FD-LE), decision feedback equalization (DFE), successive interference cancellation (SIC) and Turbo equalization. Finally we choice two classic detection algorithm as a follow-up study, the MMSE and SIC detection, also taking into account the combination with the various pairing criteria. Furthermore, the comprehensive computer simulation was done in different cellular environment and fading channels. Simulation results show that the iterative detection algorithm is significantly superior to the conventional frequency domain linear equalization based receiver in terms of bit error rate (BER) performance, the effective iterative algorithms which may bring about 1-5 dB improvement over the linear FDE method. The Turbo equalization has the best equalization performance by adding complexity at the receiver through an iteration process, in which feedback information obtained from the decoder is incorporated into the equalizer at the next iteration. The iterative processing allows for reduction of the ISI, multi-stream interference, and noise by exchanging of extrinsic information between the equalizer and the decoder.In 3GPP LTE Release 10, it is envisioned that compatible UE may support up to four transmit antennas, so spatial multiplexing (SM) and space time coding (STC) are considered to improve the uplink peak rate in such cases. So in chapter three, we extend the content of the second chapter and study the MMIO SC-FDMA system. In the second, we consider the pairing criteria for MIMO SC-FDMA system and propose an improved scheduling algorithm. As each UE with multi-antenna transmission in MU-MIMO system will increase the complexity of the receiver detection algorithm, the average uncoded BER of the linear MMSE and non-linear SIC detection is given via simulation. The performance of the SIC receiver is better than the linear MMSE receiver with added complexity in MIMO SC-FDMA systems.In chapter four, we extend the system model from a single-cell to multi-cell and investigate the frequency multiuser detection in coordinated multiple point (CoMP) reception scheme which proposed in the uplink of LTE-Advanced system. The CoMP scheme in conjoint with joint detection at the eNodeB side can reduce ICI and improve the cell capacity. To guarantee fairness of the cell-edge users, proportional fair (PF) scheduling method is adopted. Simulation results for uplink show that the CoMP scheme can significantly improve the average cell throughput and reduce ICI for the cell-edge users. Considering CCI is becoming the dominant limiting factor in the current cellular system, we propose an iterative multiuser detection method to suppress it in SC-FDMA system. The simulation results show that the iterative detect process can improve the edge user's performance, and to some extent, suppress the CCI in the multiuser systems.Finally, some conclusions of the whole dissertation are drawn, and the further research issues and possible direction are pointed out.