Blind Separation of Multi-Gaussian Sources using B-Distribution
There is an essential assumption in Blind Source separation (BSS) that the sources have no more than one Gaussian signal. This paper introduces a new BSS approach exploiting the difference in the timefrequency (t-f) spectrum of the signals to be separated. The approach uses the B. Boashash distribution to obtain t-f distribution of observed signals, then localizes the signal energy by Hough transform and obtains the estimated signals based on the diagonalization of a combined set of t-f distributions. The proposed approach allows the separation of Gaussian sources with identical spectral shape but with different t-f localization properties. Finally, the performance simulation is provided highlighting its effectiveness.
Gaussian sourcess blind source separation Bdistribution Hough transform
Jing Guo Ying Deng
School of Electronic and Information Engineering Southwest University Chongqing 400715, P.R. China C Yucai College Southwest University Chongqing 401524, P.R. China
国际会议
2010 International Conference on Signal and Information Processing(2010年IEEE信号与信息处理国际会议 ICSIP2010)
长沙
英文
304-308
2010-12-14（万方平台首次上网日期，不代表论文的发表时间）