SWIFT: Opportunistic mm-Wave Receivers

University Of Southern California, Los Angeles CA

Investigators

Abstract

Traditional wireless systems operate in the radiofrequency (RF) frequency range, which are typically below 10 GHz. The increased available bandwidth at higher frequencies has incentivized the use of millimeter-wave (mm-wave) frequency range that are above 10 GHz for wireless communications (e.g., 5G standard, satellite communications), radar (e.g., automotive radar), and imaging (e.g., airport scanners) applications. The coexistence of these emerging applications with traditional applications including radio astronomy at the same frequency bands creates challenges in terms of interference management. The coexistence issue is particularly challenging as many of the mm-wave spectrum users are based on legacy technologies that lack proper interference mitigation and filtering. On the other hand, the cost of developing, testing, and deploying dedicated mm-wave solutions for each application and frequency band is high. This proposal focuses on development of mm-wave systems that can opportunistically operate any-where within the important frequency spectrum of around 18 - 54 GHz. The first key step in it is the development of transceiver hardware that can adapt operation. Furthermore, these receivers are array based, which allows them to identify and operate at the optimum spatial dimension as well. Tightly related to this is the development of sensing and adaptation algorithms that can determine available frequency bands and directions in which transmission may occur. These algorithms are based on novel machine learning approaches and consider the special propagation conditions of millimeter-wave channels. This multidisciplinary proposal offers a holistic study of millimeter-wave systems, emphasizing coexistence, with elements that include system architecture, channel propagation, algorithms, and integrated circuit design. The innovative integrated circuits can operate across a broad frequency spectrum while providing selectivity (interference mitigation) that is currently only achieved at lower radio frequencies. Furthermore, the proposed receiver architecture allows concurrent multi-band frequency selectivity to enable for intra- and inter-band carrier aggregation. The machine-learning based algorithms start from a formulation that takes real-world constraints, such as directional dispersion and partial sparsity of the propagation channels, as well as imperfect hardware into account, and develops new machine learning approaches to tackle them. Extensive data of directional interference characteristics will be made available to all US researchers. The outcomes of the proposed study benefit current and future millimeter-wave systems and applications spanning licensed and unlicensed commercial wireless communications, satellite communications, radio astronomy, and wireless power delivery. Research results will impact undergraduate and graduate curriculum. The PIs will build on their extensive track record of involving Undergraduate students in the proposed research, and of broadening participation of students in research positions related to this project. Research results be transitioned to industry via the USC Steven's Center for Innovation and spin-off startup companies. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.