High-accurate range acquisition for LFMCW radar with optimized maximum likelihood estimation towards Internet of Everything

Linear frequency modulated continuous wave (LFMCW) radar, as an emerging key sensor, facilitates the realization of Internet of Everything (IoE) covering a tremendous range of areas such as healthcare, smart homes, and industrial automation. Among the applications towards IoE, the priority is high-accurate range acquisition of LFMCW radar. For the LFMCW radar, the beat frequency estimation directly determines the ranging accuracy. In this paper, we study the high-accurate beat frequency estimation using the optimized Maximum Likelihood Approach (MLA). The Nelder–Mead algorithm is employed during the optimization process. The numerical results reveal that the estimated beat frequency severely deviates from the theoretical value across various signal-to-noise ratio (SNR) levels without optimization. In contrast, our proposed method enables extremely accurate estimation of the beat frequency, with relative errors less than 1.6 Hz compared to the theoretical value. Accordingly, a high-accurate range with relative errors below 4.4 mm can be gained. Furthermore, it is concluded that both the estimated beat frequencies and ranges closely approach the Cramér–Rao bound across various SNR levels, which means that the optimized MLA has formidable noise immunity in practical applications, enabling highly accurate and robust range acquisition.