Study on Integration of Belief Propagation and Deep Learning for Large Multiuser MIMO Detection

2020 Fiscal Year Final Research Report

• Project/Area Number: 19K23516
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 0302:Electrical and electronic engineering and related fields
• Research Institution: Osaka University
• Principal Investigator: Takahashi Takumi 大阪大学, 工学研究科, 助教 (40844204)
• Project Period (FY): 2019-08-30 – 2021-03-31
• Keywords: 大規模MIMO / 大規模信号検出 / 確率伝搬法 / 深層展開 / 深層学習 / データ駆動型チューニング

Outline of Final Research Achievements

Large MIMO is one of the most promising technologies in the fifth generation and beyond (5G+) and sixth generation (6G) networks, in order to achieve high spectral efficiency and massive connectivity. To achieve this, low-complexity signal processing is indispensable for the base station to process multi-dimensional information. In this project, we focused on the signal separation in the uplink scenarios, and aimed to develop a low-complexity and high-accuracy large-scale multi-user detection (MUD) method by integrating belief propagation (BP) and deep learning. The proposed framework can optimize the BP-based detector via data-driven tuning with appropriate loss functions according to various communication parameters to improve the detection capability in practical MUD.

Free Research Field

無線通信　信号処理

Academic Significance and Societal Importance of the Research Achievements

IoT (Internet of Things) を情報基盤とするSociety 5.0の実現には，高速大容量・大規模同時接続・高信頼低遅延など，様々な要件が求められる．限られた周波数資源でこれらを達成するため，無線通信システムで扱う信号はますます大規模化・多次元化しており，多次元信号を高速かつ省電力で処理する低処理量な信号処理の開発が急務である．しかし，多くの低処理量な信号処理手法は扱う信号のモデル誤差に対して脆弱であり，実用化する上での大きな障壁となっている．本研究の本質は，この理論と実践の間にある隔たりを機械学習によって埋め合わせるものであり，未来のIoT社会を下支えする基盤技術となる．