This proposal aims to develop an open-source platform called M3 to facilitate research in 5G vehicular networking and automotive sensing. M3 enables experimental research on millimeter-wave (mmWave) technologies--the cornerstone for 5G wireless communication and automotive radar sensing. mmWave radio/radar electronically steerable directional beams, generated by large antenna arrays, as communication/sensing medium. Programmability is critical for mmWave experimental research, especially in real-time vehicular networking/sensing. Yet, to date, programmable mmWave devices are either too costly, or lack a reasonably-sized antenna array which is critical for real-time beam-steering operations. M3 will fill this gap with a low-cost software radio/radar featuring a large antenna array. By designing a novel radio/radar architecture, M3 brings the per-node cost down by an order of magnitude, and increases the phased-array size by an order of magnitude, compared with the state-of-the-art. The research team will deploy an open-access experimental testbed on the UCSD campus comprised of the M3 radios/radars. The researchers will also bring M3 to the broader research community, through online Q&A forum, hands-on workshops/tutorials, remote access, and hardware loans/replication services.
This proposal aims to develop a programmable open-source massive Multiple-Input/Multiple-Output (MIMO) millimeter-wave (mmWave) platform called M3 to facilitate experimental research in 5G vehicular networking and automotive sensing. The project comprises three major research thrusts: (i) Develop software/firmware to enable a partially programmable 802.11ad mmWave radio with 288-element phased-array, allowing users to reconfigure the codebook entries and beam patterns, and access real-time per-beam channel state information (CSI). (ii) Develop a mmWave massive MIMO software-radio, comprising 4 radio frequency (RF) chains and 144 antenna elements in total. The software radio will integrate the OpenAirInterface 5G physical layer and core network stack. (iii) Develop a programmable MIMO mmWave phased-array radar, with 4 RF chains and 144 antenna elements, used for exploring high-resolution automotive sensing. M3's cost is an order of magnitude lower compared with the state-of-the-art mmWave software-radio, but its phased-array is an order of magnitude higher. The surprisingly low cost is attributed to a novel radio architecture design, which repurposes a commodity phased-array antenna as a programmable phased-array. This project will bring M3 to the broader research community through hands-on workshops/tutorials, and provide user services including hardware loans, replication, and restricted remote access to a testbed comprised of M3 radios.
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.