FPGA design of fixed-complexity high-throughput MIMO detector based on QRDM algorithm

This paper presents a field-programmable gate array (FPGA) implementation of an unbiased minimum mean square error (MMSE) metric based QR-decomposition M (QRDM) algorithm for the multiple-input multiple-output (MIMO) systems. Two advanced techniques, namely the merge-sort (MS) based and winner path expansion (WPE) based sorting schemes have been implemented and validated on an FPGA platform for a 4×4 16-QAM MIMO system. The results show that the MS-QRDM is advantageous in the simplified control circuits and leads to less logic resource use, whereas the WPE-QRDM is able to achieve the minimum use of the computational units and results in fewer multipliers. Furthermore, it also shows that both schemes can support up to 1.6Gbps decoding throughput when they are implemented in a fully pipelined parallel architecture.

[1]  Andreas Peter Burg,et al.  K-best MIMO detection VLSI architectures achieving up to 424 Mbps , 2006, 2006 IEEE International Symposium on Circuits and Systems.

[2]  Tharmalingam Ratnarajah,et al.  FPGA implementation of MMSE metric based efficient near-ML detection , 2008, 2008 International ITG Workshop on Smart Antennas.

[3]  Ahmed M. Eltawil,et al.  Architectural Optimizations for Low-Power $K$ -Best MIMO Decoders , 2009, IEEE Transactions on Vehicular Technology.

[4]  S. Lakshmivarahan,et al.  Parallel Sorting Algorithms , 1984, Adv. Comput..

[5]  A. Burg,et al.  VLSI implementation of MIMO detection using the sphere decoding algorithm , 2005, IEEE Journal of Solid-State Circuits.

[6]  John S. Thompson,et al.  Fixing the Complexity of the Sphere Decoder for MIMO Detection , 2008, IEEE Transactions on Wireless Communications.

[7]  Stephan ten Brink,et al.  Achieving near-capacity on a multiple-antenna channel , 2003, IEEE Trans. Commun..

[8]  Nariman Moezzi Madani,et al.  High-throughput low-complexity MIMO detector based on K-best algorithm , 2009, GLSVLSI '09.