Hardware implementation of a multiuser detection scheme based on recurrent neural networks

In this paper we describe the hardware (HW) implementation of a discrete-time channel matrix computation and a multiuser detection (MUD) scheme. We propose a MUD scheme based on recurrent neural networks (RNN) for the TDD mode of UMTS Terrestrial Radio Access. This algorithm achieves a performance which is close to the optimum MUD, while keeping the computational complexity low. To reach the high real-time data throughput we implemented the algorithm with parallel multipliers on a field programmable gate array (FPGA).