EAGER: A Beamforming Optimization Framework for Large Scale mmWave Networks

Old Dominion University Research Foundation, Norfolk VA

Investigators

Abstract

In recent years we have witnessed remarkable proliferation of intelligent wireless devices. According to leading market research firms, there will be 50 to 100 billion wireless devices by 2025. At the same time, mobile-broadband services will continue driving the demand for higher consumer data rates. As a result, the wireless industry is anticipated to meet a 1000x growth challenge. The vast majority of today's wireless communications systems operate in the microwave spectrum below 3 GHz, which is experiencing severe spectrum shortage and has become a crowded and precious resource. The millimeter wave (mmWave) band, operating at frequencies between 20 and 300 GHz, is a promising candidate for next-generation wireless communications systems. The massive underutilized spectrum at the mmWave band provides great potential to support multiple gigabit-per-second user data rates and thousand-fold increase in total mobile broadband data rate. Beamforming is a fundamental enabling technology for wireless communications at the mmWave band. This project explores a framework for beamforming optimization in large scale mmWave networks. If successfully completed, this exploratory research is expected to advance and affect the design, deployment, and operation of future mmWave communication systems, and result in significant scientific, economical, and societal impacts. The project will broaden the participation of underrepresented groups, such as underrepresented minority and female students, in the state-of-the-art research, as well as promote STEM education through outreach activities such as summer camps. To compensate the high path loss at the mmWave band, the beamforming technology concentrates energy in a sharp beam toward the receiver to achieve a high channel gain. However, the existing beamforming approaches are prohibitively costly, both technically and economically, and cannot be scaled to beamforming for large scale mmWave networks. This project explores a framework for beamforming optimization in large scale mmWave networks, with multiple base stations, a large number of mobile stations or users, and a vast, multi-dimensional solution space for various beam formation, user association, and beam scheduling configurations. The proposed research is exploratory and consists of four tasks: 1) Clustering and Resource Allocation, 2) Power Control, 3) Multi-base station Oriented Clustering, and 4) Beamforming and Interference. The four tasks are intertwined and multidisciplinary. The innovative optimization formulations, various heuristic algorithms and schemes based on multiple techniques such as set covering and bin packing will be explored, to form a comprehensive framework for user clustering, beam resource allocation, power control and beam scheduling in large scale mmWave networks. The optimization models, algorithms and schemes developed for the framework make it practical for beamforming optimization in large scale mmWave networks. This will significantly help to pave the roads for large scale mmWave networks. The techniques developed in this project will have significant impacts on a broad spectrum of applications, including 5G, mmWave WLAN, and public safety communication networks.