Channel shortening and equalization based on information rate maximization for evolved GSM/EDGE

In this paper, we study Rusek´s channel shortening receiver based on the maximization of the lower bound of the information rate for inter-symbol-interference scenarios. We introduce the necessary simplifications which enable the use of the channel shortening algorithm for Evolved GSM/EDGE receivers for 2G cellular networks. We derive the computational complexity, evaluate the performance and compare both to state-of-the art Evolved GSM/EDGE receivers. A performance gain of up to 1 dB is achieved over state-of-the-art channel shortening receivers, while being less complex in terms of multiplications. To further substantiate the concept, a VLSI implementation of the core part of the channel shortener is presented which achieves a three times higher area-throughput product than prior art.