Reconfigurable Design of Adaptive Radius Sphere Decoding for MIMO Systems.

Multiple Input-Multiple Output (MIMO) systems uses multiple antennas at the transmitter and the receiver to achieve high spectral efficiency. Sphere decoder (SD) has been widely used as the decoding algorithm in MIMO systems which achieve near Maximum Likelihood (ML) performance with reduced complexity. The complexity of SD increases considerably with increase in system configuration or modulation order. In this paper an efficient early pruning technique has been proposed which eliminates the nodes that are unlikely to contribute to near ML solution which results in reduced complexity. The simulation shows a reduction of 41 % of complexity at 18 dB SNR. The proposed algorithm for uncoded 4X4 MIMO system with 16-QAM modulation is synthesized using Xilinx Virtex 5 XC5VLX50T FPGA. The FPGA resource utilization and timing summary shows a considerable improvement with the existing one.