Adaptive Mixing Probability Scheme in Mixed Gibbs Sampling MIMO Signal Detection

Kenshiro Chuman, Yukitoshi Sanada

研究成果: Article査読


This paper proposes an adaptive mixing probability scheme for mixed Gibbs sampling (MGS) or MGS with maximum ratio combining (MRC) in multiple-input multiple-output (MIMO) demodulation. In the conventional MGS algorithm, the mixing probability is fixed. Thus, if a search point is captured by a local minimum, it takes a larger number of samples to escape. In the proposed scheme, the mixing probability is increased when a candidate transmit symbol vector is captured by a local minimum. Using the adaptive mixing probability, the numbers of candidate transmit symbol vectors searched by demodulation algorithms increase. The proposed scheme in MGS as well as MGS with MRC reduces an error floor level as compared with the conventional scheme. Numerical results obtained through computer simulation show that the bit error rates of the MGS as well as the MGS with MRC reduces by about 1/100 when the number of iterations is 100 in a 64 × 64 MIMO system.

ジャーナルIEICE Transactions on Communications
出版ステータスPublished - 2023 12月

ASJC Scopus subject areas

  • ソフトウェア
  • コンピュータ ネットワークおよび通信
  • 電子工学および電気工学


「Adaptive Mixing Probability Scheme in Mixed Gibbs Sampling MIMO Signal Detection」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。