A general framework for blind identification and equalization of time-varying channels

In this paper, a general framework is proposed to encompass several major blind approaches that have been proposed for identification and equalization of time-varying channels most recently. We outline the basic ideas of each method, discuss the assumptions, identifiability conditions and the characteristics of respective approach. We also present the corresponding remarks which devote to compare them with each other and point out the differences between them. The proposed general framework, especially the decomposition idea maybe useful for develop new blind approaches for the time-varying channels, for example, the classical Karhunen-Loeve expansion method can be borrowed for researching the time-varying channels problem.