Abstract:

In this study, spatial modulation (SM), which is an interesting and new approach for 5G and beyond communication systems, and deep neural network (DNN), which have also received great attention recently, are discussed, and a DNN-based receiver architecture for SM systems is proposed. Since the DNN will not be retrained until the channel change after training, it requires less processing, so it will be a potential receiver architecture for next-generation wireless communication and therefore SM systems. In this paper, a new DNN-based SM receiver is proposed to detect the transmitted symbols and the activated antenna index at the same time, and its performance is examined. As can be seen from the computer simulations, the DNN-based receiver offers low error performance with a small number of hidden layers and a low number of neurons in these layers. At the same time, even when the data rate is increased, the same DNN structure (without increasing the processing load) shows better/same performance than the receivers in the literature.

Keywords: