Blind channel estimation : multivarate approach

Abstract

In the case of fading communication links, like the mobile radio channels, blind channel equalization techniques can potentially eliminate the need for periodic training of the receiver. This results in increased band width efficiency. This work addresses the problem of blind channel estimation na rising in the context of digital communications. It demonstrates that by using certain precoding and interleaving techniques at the transmitter, and certain principles of Multirate filter theory, it is possible to induce cyclo stationarity in the transmitted symbols. This cyclo stationarity property can be exploited at the receiver by considering the poly phase decomposition of the received channel output. The received symbols are arranged interms of their polyphase components using the principles of Multirate theory, and processing this arrangement with the help of an adaptive algorithm, gives a fairly accurate estimate of the channel. These estimates are then used for the procedure of equalization.