Acquisition of detailed and accurate information from a brain-morphic wireless sensor network by reservoir computing

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K11867
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 60060:Information network-related
• Research Institution: Osaka University
• Principal Investigator: Wakamiya Naoki 大阪大学, 大学院情報科学研究科, 教授 (50283742)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 無線センサネットワーク / リザバ計算 / 温度分布推定 / イベント検知

Outline of Final Research Achievements

To establish a new architecture called "brain-morphic wireless sensor network" which is highly energy-efficient and highly robust by performing all of sensing, information processing, and communication in the form of pulses, we investigated the optimal design of input, communication, and output. By an appropriate combination of device configuration, input current, and a regression algorithm, the accurate estimation of spatial distribution of temperature was accomplished. Regarding event detection by an actuator, the probability of event detection was as high as 82.81% when an actuator was located at the center of the region.

Free Research Field

バイオ情報通信工学

Academic Significance and Societal Importance of the Research Achievements

デバイスの構造や情報処理ならびに通信のアルゴリズムが極めてシンプルであるため低コスト、低消費電力でありながらも干渉や故障に対して非常にロバストな無線センサネットワークを容易に構築できることから、特に多数または高密度な環境でのセンシング情報収集を行う多様なアプリケーションの基盤技術として広く応用、展開が可能である。あわせて物理リザバ計算に関する研究の観点でも新たな知見を得ており当該分野の今後の発展に寄与するものである。