A List Sphere Decoding Algorithm with Improved Radius Setting Strategies

作者：Wang Ping; Tho Le Ngoc^*

来源：Wireless Personal Communications, 2011, 61(1): 189-200.

DOI：10.1007/s11277-010-0018-4

• Sphere decoding (SD)
• List Sphere Decoding (LSD)
• MIMO
• Maximum-likelihood (ML)
• Integer Least-square (ILS)

Complexity of different sphere-decoding (SD) algorithms is heavily influenced by their own distinctive computational features. Sphere radius setting for initialization and reduction during the search process, is one of the most important aspects affecting the SD efficiency. This paper examines radius setting strategies for list sphere-decoding (LSD) algorithms to calculate the soft-output information for coded systems, and proposes strategies to improve both the radius initialization and reduction in the original LSD algorithm. The proposed strategies can avoid the instability problem in the radius initialization of the original LSD algorithm and significantly increase its search efficiency as confirmed by simulation for both full-column-rank and underdetermined MIMO channels.

• 出版日期2011-11
• 单位McGill