mmWave Joint Radar/Communication (JRC) Testbed

Vehicle-to-everything (V2X) communication and vehicular radar imaging technologies have become the key enablers of Intelligent Transportation Systems (ITS) to promote safety, automation, and coordination of vehicle traffic. Up to 4 GHz of contiguous bandwidth is allocated as the vehicular radar spectrum that is dedicated solely to vehicles in the 76-81 GHz millimeter-wave (mmWave) band. The objective of Joint Radar/Communication (JRC) system is to perform both data transmission and radar imaging using the same joint waveform and transceiver hardware, enabling a plethora of new, bandwidth-intensive applications. The testbed is the proof-of-concept implementation, operating at 24GHz.

MIMO OFDM-based JRC Architecture at 24GHz

mmWave JRC Architecture

  • Fully-digital and software-defined architecture operating in the 24 GHz band
  • Achieves 200 MHz bandwidth with 4 TX and 2 RX channels
  • 2 × Software-defined radios (SDRs): USRP N320/1
    • Baseband → 5 GHz Intermediate Frequency (IF) band
  • Custom-built mmWave front-ends:
    • 5 GHz IF → 24 GHz RF band

Front-End Details

  • Host PC: MIMO OFDM generation + Baseband digital signal processing
    • Real-time MIMO radar imaging and communication with 10 Gbps Ethernet links
  • Phase-coherent TX/RX channels:
    • LO sharing for IF (5 GHz): Between USRP N320 and N321
    • Implemented a LO distribution system (19 GHz): Amplifiers + power splitters
  • Band pass filters → Lower sideband cancellation

Demo

Transmitter beam tracks the receiver with radar feedback

Open-Source Software Platform for ML Control of the mmWave MIMO-OFDM Testbed

The hardware setup for the mmWave MIMO-OFDM testbed in the DEMO, as well as the directed beam method that leverages radar images to obtain beamforming angle information for communication with user equipment, is based on [1].
The Gaussian Process Multi-Armed Bandit (GP-MAB) method in the DEMO, which uses radar images as contextual information, applies the GP-MAB algorithm to make decisions on beamforming angles. This approach is inspired by [2] from the AI-EDGE Institute.
[1] C. D. Ozkaptan, H. Zhu, E. Ekici and O. Altintas, “A mmWave MIMO Joint Radar-Communication Testbed With Radar-Assisted Precoding,” in IEEE Transactions on Wireless Communications, vol. 23, no. 7, pp. 7079-7094, July 2024, doi: 10.1109/TWC.2023.3337282.
[2] Y. Deng, X. Zhou, A. Ghosh, A. Gupta and N. B. Shroff, “Interference Constrained Beam Alignment for Time-Varying Channels via Kernelized Bandits,” 2022 20th International Symposium on Modeling and Optimization in Mobile, Ad hoc, and Wireless Networks (WiOpt), Torino, Italy, 2022, pp. 25-32, doi: 10.23919

Sources:

Source Code: https://github.com/ceyhunozkaptan/gr-mimo-ofdm-jrc

Publications:

  1. Ceyhun D. Ozkaptan, Haocheng Zhu, Eylem Ekici, and Onur Altintas “A Fully Digital MIMO Joint Radar-Communication Testbed with Radar-assisted Precoding“ to appear in IEEE Transactions on Wireless Communications, November 2023.
  2. C.D. Ozkaptan, H. Zhu, E. Ekici, O. Altintas, “Software-Defined MIMO OFDM Joint Radar-Communication Platform with Fully Digital mmWave Architecture,“ Proceedings of  IEEE Symposium on Joint Communication and Sensing, Online, March 2023.
  3. Ceyhun D. Ozkaptan, Eylem Ekici, and Onur Altintas Adaptive Waveform Design for Communication-Enabled Automotive Radars, IEEE Transactions on Wireless Communications, vol.21, no. 6, pp. 3965-3978, June 2022.
  4. C. D. Ozkaptan, E. Ekici, C-H. Wang, O. Altintas Optimal Precoder Design for MIMO-OFDM-based Joint Automotive Radar-Communication Networks, Proceedings of WiOpt 2021, Online, October 2021.

Supported by the
National Science Foundation