A novel joint parameter estimation method of single-channel and time-frequency overlapped multi-component signal

摘要

In this paper, a novel algorithm is proposed, which focus on estimating the period of independent source and the source number of the multi-component signal that recorded by single antenna and overlapped both in time and frequency domain. The method adopts the technique of changing single-channel recording vector into virtual multi-channel measurement matrix, and then using Singular Value Decomposition (SVD) method decompose this matrix. By changing the virtual multi-channel measure matrix, singular value ratio (SVR) spectrum can be got. By searching the peaks of the SVR, the period of the independent components can be estimated effectively, and based on the correlation coefficient between each estimated components obtain useful SVD, then the number of original source will be got. The simulation results demonstrate the effectiveness of the algorithm. What is more, the high estimated accuracy of this method under low signal-noise-ratio (SNR) condition also illustrated by the simulation.