Waveform Design and System Verification of Millimeter Wave Radar

Abstract

In modern radar applications such as autonomous driving, the range resolution of the radar system is of great importance. The key factors for improving the range resolution involve smart waveform and radar sensing design in addition to the signal bandwidth. Another concept when it comes to waveform design, is the idea of designing radar waveforms that are compatible with communication, which could provide efficient spectrum and hardware usage. In this thesis, we demonstrate the waveform generation and radar sensing process for a linear frequency modulated (LFM) and orthogonal frequency-division multiplexing (OFDM) based radar waveform compatible with communication. The functionality and radar performance of both the radar systems are evaluated through simulations and real-world measurements at a radio frequency of 80 GHz. It is found that the range resolution of both the LFM and OFDM radar systems is 13.2 cm at 1 GHz bandwidth, when using their conventional radar sensing algorithms. With the intent to improve the range resolution without increasing the bandwidth, a super-resolution algorithm known as multiple signal classification (MUSIC) is implemented for the OFDM radar which is able to improve the range resolution down to 6 cm, with the cost of a higher computational complexity than the conventional OFDM and LFM algorithms.