The Optimization of Spectrum Analysis for Digital Signals

This paper provides a complete method to solve the defects of spectrum analysis for discrete signals. Much research has been issued to deal with the defects of the fast Fourier transform (FFT). Those methods may cause the characteristics of the original signal to be altered or may solve this problem in part only. The defects of spectrum analysis cannot be solved efficiently and completely by those methods. These defects result because frequency scales cannot match with signal characteristics. This paper is based on a concept to establish a complete solution divided into three steps. First, this paper analyzes the signal characteristics to be the basis of spectrum adjustment. A simple and accurate algorithm is used to find the frequency and amplitude of each component. Next, this paper finds optimum spectrum parameters to make the new spectrum match with signal characteristics. Finally, this paper takes the parameters to reanalyze the original signal. The method will make spectrum analysis reach optimization. Every procedure in this paper compares with traditional ones to prove its benefits. Moreover, we verify the theory feasible by analyzing actual signals.