Coefficient adjustment matrix inversion approach and architecture for massive MIMO systems

Thanks to hundreds of antennas, spectral efficiency of massive multiple-input multiple-output (MIMO) systems has drastically increased. However, the resulting huge dimension of matrices involved in massive MIMO MMSE detection causes prohibitive complexity. Although large scale matrix inversion with Neumann approximation achieves good tradeoff between complexity and accuracy for i.i.d. massive MIMO channel, its convergency speed degrades seriously for correlated massive MIMO channel. To this end, in this paper the matrix inversion approach based on coefficient adjustment (CA), which is more adaptable to correlated channel with higher throughput, is proposed. The corresponding hardware architecture is also given. FPGA results have shown that for 4 × 32 MIMO system, the proposed architecture can achieve 69.4% higher frequency with only 49.1% hardware cost compared to Cholesky decomposition method. CA approach can also achieve 37.9% higher throughput than Neumann scheme for correlated channel on average.

[1]  Xiaohu You,et al.  Efficient matrix inversion architecture for linear detection in massive MIMO systems , 2015, 2015 IEEE International Conference on Digital Signal Processing (DSP).

[2]  Rohit U. Nabar,et al.  Introduction to Space-Time Wireless Communications , 2003 .

[3]  Erik G. Larsson,et al.  Scaling Up MIMO: Opportunities and Challenges with Very Large Arrays , 2012, IEEE Signal Process. Mag..

[4]  Fredrik Rusek,et al.  Hardware efficient approximative matrix inversion for linear pre-coding in massive MIMO , 2014, 2014 IEEE International Symposium on Circuits and Systems (ISCAS).

[5]  Fredrik Tufvesson,et al.  Linear Pre-Coding Performance in Measured Very-Large MIMO Channels , 2011, 2011 IEEE Vehicular Technology Conference (VTC Fall).

[6]  Preben E. Mogensen,et al.  A stochastic MIMO radio channel model with experimental validation , 2002, IEEE J. Sel. Areas Commun..

[7]  Aravindh Krishnamoorthy,et al.  Matrix inversion using Cholesky decomposition , 2011, 2013 Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA).

[8]  Poras T. Balsara,et al.  VLSI Architecture for Matrix Inversion using Modified Gram-Schmidt based QR Decomposition , 2007, 20th International Conference on VLSI Design held jointly with 6th International Conference on Embedded Systems (VLSID'07).

[9]  Joseph R. Cavallaro,et al.  Approximate matrix inversion for high-throughput data detection in the large-scale MIMO uplink , 2013, 2013 IEEE International Symposium on Circuits and Systems (ISCAS2013).

[10]  Xiaohu You,et al.  Efficient iterative soft detection based on polynomial approximation for massive MIMO , 2015, 2015 International Conference on Wireless Communications & Signal Processing (WCSP).