Joint communication and radar at the millimeter-wave band
Millimeter-wave (mmWave) communication and radar are key technologies for many demanding applications, such as autonomous driving and smart connected devices. The combination of these two technologies into a single joint communication-radar (JCR) enables hardware reuse and a common signaling waveform. This leads to significant benefits in power consumption, spectrum efficiency, and market penetrability. Prior work has proposed a mmWave wireless local area network-based JCR, which provides a good baseline for designing a future mmWave JCR standard. These JCR systems, however, incurred limited radar estimation performance.
In this dissertation, we present advanced adaptive JCR waveform and beamforming designs to achieve superior radar performance, at the cost of a small reduction in the communication rate. These contributions are foundational for the development of futuristic standard-based mmWave JCR systems in automotive applications. The summary of our contributions is as follows.
First, we propose a virtual waveform design for an adaptive mmWave JCR to enhance velocity estimation accuracy without reducing the communication data rate much. We formulate three different minimum mean square error-based optimization problems for the adaptive JCR waveform design. We solve the JCR optimization problems for a uniform waveform as well as for nested and Wichmann virtual waveforms. Numerical results demonstrate that optimal virtual waveforms achieve significant performance improvements over a uniform waveform, especially at low SNR and high target density.
Second, this dissertation presents a low-complexity fully-digital multiple-input-multiple-output measurement platform, named JCR70, to evaluate and demonstrate the JCR performance in the 71-76 GHz band. Additionally, an experimental proof-of-concept is also performed for low-resolution analog-to-digital converters by emulating quantization effect on the collected data. The JCR70 platform demonstrated higher resolution capability than the Radarbook, which is a leading automotive radar evaluation platform at 77 GHz.
Finally, this dissertation develops an adaptive combined waveform-beamforming design for automotive mmWave JCR that uses a phased-array architecture. To estimate the radar channel in the Doppler-angle domain with a wide field-of-view, all the transmit antennas are used to generate a narrow coherent beam for communication and distribute the remaining energy uniformly along other radar sensing directions. By applying partial Fourier compressed sensing technique, the framework is shown to estimate radar channels with high accuracy, at the cost of a small reduction in the communication rate.