Skip to Main Content
A novel maximum likelihood sequence estimation (MLSE) equalizer is presented. The method does not require the explicit modeling of the channel and it belongs to the class of cluster based sequence equalizers (CBSE). The novelty of the method is that the required clusters of the received data are estimated in the one dimensional space via a technique that exploits the underlying symmetries in the structure of the received clusters. This gives a two-fold advantage; the computational load is drastically reduced, and at the same time the convergence speed is improved compared to both LMS based channel estimators and previously suggested cluster based equalizers.