Super-resolution range estimation in FMCW radar system

Abstract

This paper presents two new methods to estimate multiple ranges with super-resolution in Frequency Modulated Continuous Wave (FMCW) radar system. Conventionally, the range estimation is implemented through Fast Fourier Transform (FFT) processor. We note that the computational complexity of FFT processor is O(N log N). The purpose of this paper is to keep the complexity less than $O(N^2)$ so that the range-resolution can be improved with slight computational overhead. The first proposed method is using monopulse technique to estimate the beat frequency. The method is extended to multiple targets case. While the FFT processor is not needed, the complexity is $O(N^2)$ . The second method is using early-late gate to estimate beat frequency. The method is extended to multiple targets case. The complexity is O(N) after the measurement is converted from time to frequency domain by the FFT processor. The overall complexity of the method is O(N log N). Both of the methods require no variation of hardware configurations because the methods are implemented in digital signal processing. With slight computational overhead, the proposed methods improve the range resolution to the corresponding Cramer-Rao (CR) bound. In the simulation, we confirmed that Root Mean Square Error (RMSE) of each estimated range approaches to its CR bound.