A new blind identification algorithm based on cyclostationarity and the bandlimited property of the communication channel

This paper addresses the problem of blind channel identification in the context of digital communications. The identification is based on the cyclostationarity properties that are obtained when the received signal is sampled at a rate higher than the baud rate. This approach has been used in previously published blind estimation methods, but here, the assumption that the received signal is approximately bandlimited is explored. Two previously proposed methods are modified to take into account the bandlimited property. The resulting algorithms always have a computational advantage and can also show better estimation quality in the case when the channel model order is not well defined, as often occurs in practical applications. Illustrative simulations are presented that corroborate the improved performance of this approach