A novel method of power spectrum estimation in CW radar signal processing

The continued wave radar has been widely used in the industrial control, vehicle safety and the other related field. This paper aims to lower down the noise when we estimate the power spectrum. We use the difference of the autocorrelation function between the noise and the signal and we introduce a new method to improve SNR when estimating the power spectrum. According to the Wiener-Khintchine theorem, we combine the power estimation and the denoising and we build a structure of power spectrum estimation which can improve SNR. The structure can accumulate the power of useful signal and the SNRI can reach 20log(M)dB if accumulating the signal M times. This structure has limitations in dynamic range and times of accumulation. We use Matlab to simulate and analyze the performance of the structure at the end of this paper.