Advanced Signal Processing Enabled Massive MIMO With NOMA

点击翻译按钮获取中文摘要

The capacity of current wireless networks is becoming a major bottleneck as most of the radio frequency bandwidth has already been allocated for specific services. In contrast, millimeter-wave (mmWave) communication operating within the 30-300GHz frequency band is a promising technology for the 5th generation (5G) and beyond (B5G) wireless networks. Due to its abundant bandwidth and short wavelength, mmWave enables much higher capacities and transmission rates than conventional microwave communications (e.g., Sub-6 GHz). Moreover, its short wavelength allows a large number of antennas to be deployed in the base station and multiple antennas to be integrated in a small hand-held device, making massive multiple-input multiple-output (MIMO) systems go hand-in-hand with mmWave networks. On the other hand, non-orthogonal multiple access (NOMA), as a potential technology to boost system spectral and energy efficiencies, has recently received a great deal of attention. NOMA allows multiple users to access networks at the same time and frequency resource block, and separates the signals of different users by allocating different power levels according to channel state information (CSI). Despite many advantages of mmWave communication using massive MIMO and NOMA, however, there are technical challenges in implementing such a complex system. These include requiring extensive CSI, which is difficult to acquire under massive MIMO, and many radio-frequency (RF) chains for the antenna elements. Moreover, mmWave signals experience severe path loss, high penetration loss and harsh atmospheric absorption as compared to micro-wave signals, and therefore directional transmission with sharp beam is required to overcome the severe path loss of mmWave signals. In addition, NOMA inevitably introduces interferences between multiple users that share nonorthogonal resources at the same time and frequency. In this project, we investigate and develop novel signal processing approaches to tackle the fundamental challenges in massive MIMO-NOMA systems operating at mmWave frequencies for future B5G networks. The long-term objective of this proposal is to design enabling signal processing technologies for massive MIMO-NOMA systems in support of B5G wireless communications. The short-term objectives include (1) design of novel channel estimation techniques for mmWave massive MIMO systems; (2) optimization of transceiver beamforming and beam-tracking; (3) NOMA user clustering and power allocation; and (4) performance evaluation of mmWave MIMO-NOMA systems. The proposed project also involves significant training activities for both undergraduate and graduate students as well as post-doctoral research fellows that will take place at Concordia University.