Improved coherent accumulation power spectrum detection technology

Abstract

The most basic method to detect line spectrum signal from background noise is spectrum analysis, such as direct method [1]. In order to reduce the variance of the spectral component of background noise, Welch method is derived. However, Welch method only considers the non coherent information so that it will lead to the loss of gain. The existing coherent average detection method which is based on phase compensation can enhance the processing gain of line spectrum signals. In this paper, an improved coherent average power spectrum detection technique is proposed by combining polynomial fitting technique with coherent average detection method. The simulation results show that this method can reduce the fluctuation of background noise further, and its performance is better than the coherent average power spectrum method.