Fifth-generation (5G) wireless systems are expected to provide enormous improvements in data rates available to users, as well as much improvement overall user experience. Massive multiple-input multiple-output (MIMO) arrays consist of hundreds of antenna elements, serving many users and are considered to be a cornerstone of 5G wireless systems, and are expected to dramatically improve both the radio spectrum utilization and user experience. At the same time, the use of millimeter-wave (mm-wave) frequencies is supposed to provide additional spectrum for new services in the years to come, and small physical antenna separation makes mm-wave attractive for massive MIMO. While there has been substantial progress in the development of the theoretical concepts associated with the design of massive MIMO systems, very little work has been done to actually design a mm-wave massive MIMO system and on the network techniques needed to scale these systems to dozens of simultaneous spatial streams. This proposal addresses the key challenges in the development of signal processing algorithms, network protocols, and a prototype hardware design to enable scalable low-latency mm-wave MIMO networks with high degrees of spatial multiplexing. It will provide a path to a hundred-fold improvement in user data rates.
By integrating the theoretical system aspects with its practical development, this proposal addresses critical challenges for the development of mm-wave massive MIMO technologies. In particular, this project aims to achieve: (1) a mm-wave massive MIMO array architecture suitable for low-cost and energy-efficient deployment at massive scale, (2) an optimized scalable signal processing approach to massive MIMO array processing, which includes hybrid beamforming, distributed channel estimation and distributed beamforming, (3) a medium-access control (MAC) technique suitable for low-latency, low-coherence time applications by leveraging a full-duplex frame to enable rapid user acquisition, synchronization, tracking, and paging, (4) practical in-band full-duplex operation, realized through a combination of antenna array design, spatial filtering, and adaptive analog and digital cancellation, (5) practical front-end circuits with linearity and phase noise suitable for a large number of simultaneous spatial streams, and (6) a mm-wave massive MIMO test bed designed in a modular manner to enable future development and performance measurements of signal processing and MAC techniques. In addition to cross-disciplinary training of students involved in this project, interaction of project members with industry leaders will dramatically accelerate the penetration of 5G wireless communications.
