Low-complexity soft-output K-Best MIMO decoder in 2×2 spatial multiplexing system

This paper presents algorithm of low-complexity K-Best MIMO decoder algorithm applied in 2×2 spatial multiplexing system with 64-QAM modulation. Low computational complexity in the proposed algorithm is achieved by separating real and imaginary computation. We also discovered that some complex-value multiplications can be reduced into two real-value multiplications without affecting performance. These improvements increase the suitability of K-Best algorithm to be applied in real-time VLSI implementation. Finally, we verified the proposed K-Best algorithm performance and compared it with the performance of conventional algorithm in term of Bit Error Rate (BER) using IEEE 802.16-2009 WiMAX system as an implementation platform. The proposed algorithm can achieve computational reduction in Partial Euclidean Distance (PED) calculation as high as 57.78% for real-value adders and 74.93% for real-value multipliers compared with conventional K-Best.