Network bandwidth reservation method combining machine learning and linear programming

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K11798
• 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: Nihon University
• Principal Investigator: GENDA Kouichi 日本大学, 工学部, 教授 (00564105)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: SDN / 帯域予約 / 機械学習 / 線形計画法

Outline of Final Research Achievements

Network bandwidth reservation is a representative service that utilizes the advantages of software-defined networks in which users directly reserve network resources on an on-demand basis. A bandwidth reservation method is proposed here to meet the requirements, instantaneous response to requests and high request acceptance ratio, by using machine learning (ML), particularly for unpredictable bandwidth demands in which the use time is indicated strictly. Simulation results indicated that the proposed method adopting a multi-layered neural network can achieve a high accuracy ratio within a 10% difference compared to the ideal solution. In addition, the bandwidth-reservation system combining the request judgment by ML and the network resource allocation by linear programing was proposed, where they are executed in a pipeline manner. We demonstrated that the proposed system can achieve adequate network resource allocation within a 10% difference compared to the ideal solution.

Free Research Field

情報ネットワーク

Academic Significance and Societal Importance of the Research Achievements

SDNの利点を活用した新たなサービスとして期待されるネットワークのオンデマンドサービスの１つであるネットワーク帯域予約サービスの実現に向けて、速やかな受付判定と高いリクエスト受付率を両立できる新たな予約受付判定方法を提案した。
本検討では、今後需要の高まりが期待されるネットワーク領域でのオンデマンドサービスを技術面から進展させたとともに、ネットワークと機械学習を融合した新たな活用法を提示した。