Practical MIMO Decoder Based on Tree Search Approaches: Algorithm and VLSI Architecture

Monday, May 17, 2010 - 10:00 a.m. to Tuesday, May 18, 2010 - 10:55 a.m.
Engineering Gateway 3161
CPCC Seminar



Featuring Chung-An Shen

Ph.D. Candidate, Electrical Engineering and Computer Science

The Henry Samueli School of Engineering, UC Irvine



Abstract:

In spatial multiplexing MIMO communications, the straight implementation of Maximum Likelihood (ML) decoding incurs enormous complexity and power consumption, which makes it infeasible for practical wireless communications. Recently tree search based decoding schemes such as sphere decoder and/or K-best decoder have been proposed to achieve (close to) ML MIMO decoding with manageable power consumption and complexity. This talk presents an algorithm and VLSI architecture for K-Best decoding that combines the benefits of radius shrinking commonly associated with sphere decoding and the architectural benefits associated with K-Best decoding approaches. The proposed algorithm is to decompose a tree structure into multiple lower-dimensioned tree structures and process iteratively by a K-Best decoder. This algorithm requires much smaller K and possesses the advantages of adaptively updated pruning threshold such that the search space is greatly reduced. The proposed VLSI architecture of the decoder is based on a pipelined sorter-free scheme such that the decoding throughput is maximized and the overhead (area as well as power) incurred by the sorter block is minimized. The proposed K-Best decoder is designed to support a 44 64-QAM system and is synthesized with 65nm technology at 158 MHz clock frequency and 1 V supply. The synthesized decoder can support an average decoding throughput of 285.8 Mbps at 25 dB SNR with 210 K gates area at 12.8 mW power consumption.



About the Speaker:

Chung-An Shen received a B.Sc. degree from National Taiwan University of Science and Technology in 2000 and an M.Sc. degree from The Ohio State University in 2003, both in electrical engineering. He worked as a DSP system application engineer in Texas Instruments, Taiwan from 2004 to 2007. He is currently pursuing a Ph.D. degree at the University of California, Irvine. His research interests include the development and design of algorithm and VLSI architecture for wireless communication systems.