New Soft Transition Dual Mode Type Algorithms For Blind Equalization

A dual mode algorithm (DMA) employs a blind acquisition algorithm, e.g., the Godard algorithm, during the startup mode and switches to a tracking algorithm, e.g., the decision directed algorithm, when the errors have been sufficiently suppressed. Six types of soft transition DMAs in the literature, e.g., the stop-and-go algorithm, are first briefly reviewed. Subsequently, a new class of DMA (nDMA) that uses a symbol error rate (SER) driven combination parameter is presented. Since the transition is smooth and optimal in the SER sense, the nDMA achieves the fastest convergence in addition to the lowest steady state error simultaneously (verified by simulations) without compromising computational complexity.