Improving MIMO Sphere Detection Through Antenna Detection Order - PowerPoint PPT Presentation

Improving MIMO Sphere Detection Through Antenna Detection Order Scheduling Michael Wu, Chris Dick, Yang Sun, Joseph Cavallaro December 1, 2011

MIMO Detection  Spatial Multiplexing H  Increases throughput  Used in many wireless standards  MIMO detector recovers                 ˆ y y r h r h r h x x n n 0 0 00 00 01 01 02 02 0 0 0 0                     original signal ˆ y y h 0 r h r h x x n n                 1 1 10 11 11 12 12 1 1 1 1                   ˆ               y y h 0 h 0 r h x x n n  Search problem 2 2 20 21 22 22 2 2 2 2     y y Rx Hx n n ˆ  2 y Rx ˆ min n 2  x Ω 2 12/1/2011

Tree-Search Based Detection  Depth-first tree search  Variable execution time  Fairly sequential, slow  Breadth-first tree search  More data parallel  Sort is the bottleneck  Number of comparisons + Memory requirement (KM) 3 12/1/2011

SSFE Detector  Selective Spanning with Fast Enumeration (SSFE)  Data parallel deterministic search  First antenna level  Enumerate all modulation points  Subsequent levels  Pick the best outgoing node QPSK 2x2 Example for each path. 4 12/1/2011

SSFE Detector  “Sort-free” MIMO Detector  Picking the best outgoing node does not require sort S k P =[s 0 , s 1 … s k ], y k+1 , h k+1 1 �̂ ��� � � ��� �� � ��� , ��, �� � � ���,��� s k+1 round() Schnoor-Euchner enumeration 5 12/1/2011

Flexsphere Implementation F lexsphere Design ˆ x H compute the optimal search detection order Preprocessing Resource Cost 9% RVD/QRD Hardware 16% Sphere Detector 11% Soft Output Generator 64% C. Dick, M. Trajkovic, S. Denic, D. Vuletic, R. Rao, F. Harris, K. Amiri, FPGA Implementation of a Near-ML Sphere Detector for 802.16e Broadband Wireless Systems , proceedings of SDR conference, 2009 6 12/1/2011

Flexsphere Implementation: V-BLAST Reordering 4x4 4x4 Matrix Inverse Norm Matrix Matrix Back Search/ multiply QRD multiply H Subst. Reorder 3x3 3x3 Matrix Inverse Norm Matrix Matrix Back Search/ QRD multiply multiply Subst. Reorder 2x2 2x2 Matrix Inverse Norm Matrix Matrix Back Search/ QRD multiply multiply Subst. Reorder  Complex block: 3 matrix inverses and 3 matrix multiplies 7 12/1/2011

Flexsphere Implementation F lexsphere Design ˆ x H compute the optimal search detection order Preprocessing Resource Cost 9% RVD/QRD Hardware 16% Sphere Detector 11% Soft Output Generator 64% Can we do better? C. Dick, M. Trajkovic, S. Denic, D. Vuletic, R. Rao, F. Harris, K. Amiri, FPGA Implementation of a Near-ML Sphere Detector for 802.16e Broadband Wireless Systems , proceedings of SDR conference, 2009 8 12/1/2011

N-Way MIMO Detector  Get rid of the V-BLAST channel reordering block  Duplicate search blocks depending on BER requirement.  Add permute block which enforces a detection order  Example: N = 2, two search blocks y, H Merge X 0  X 1  X 2 ˆ x X 1  X 2  X 0 9 12/1/2011

N-Way MIMO Detector: Merge Block  Simple block, performs max-log-map computation +1 L 0 Soft Output min Generator -1 LLR - +1 L 1 Soft Output Generator min -1 10 12/1/2011

BER Performance  Rayleigh fading channel  Soft Output MIMO Detector + Rate 1/2 WiMAX LDPC decoder  1 outer iteration + 20 inner iteration with early termination 11 12/1/2011

BER Performance (16QAM) 12 12/1/2011

BER Performance (64QAM) 13 12/1/2011

Implementation: N-Way MIMO Detector  Target: 83.768Mbps(WiMAX), Virtex 5 @ 225Mhz  T otal resource = N (RVD/QRD + Sphere Detector + Soft Output Generator) + Merge N Slices LUTs/FFs DSP48 Block RAM 1 5,658 9,437/15,990 78 41 2 11,274 19,018/32,525 156 82 3 16,827 28,743/49,117 234 123 4 22,832 38,515/65,381 312 164 Flexsphere 15,657 29,776/45,944 237 146 14 12/1/2011

Implementation: N-Way MIMO Detector  Target: 83.768Mbps(WiMAX), Virtex 5 @ 225Mhz 15 12/1/2011

Conclusion  Scalable data parallel detection algorithm  Search is cheap  Better performance/resource compare to Flexsphere  Target for software implementation?  Enumeration complexity doesn’t depend on modulation  Increase parallelism of the detection algorithm Parallelism: NM Parallelism: M 16 12/1/2011