Research on distributed approximation algorithms with adaptive fault tolerance properties in dynamic networks

2018 Fiscal Year Final Research Report

• Project/Area Number: 26330015
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Theory of informatics
• Research Institution: Hiroshima University
• Principal Investigator: Kamei Sayaka 広島大学, 工学研究科, 准教授 (90434977)
• Project Period (FY): 2014-04-01 – 2019-03-31
• Keywords: 自己安定アルゴリズム / 安全収束性 / 分散アルゴリズム / 支配集合問題 / 相互排除問題 / 耐故障性 / 動的ネットワーク

Outline of Final Research Achievements

In this study, we designed fault-tolerant distributed approximation algorithms. They are sufficiently practical for networks with dynamically changing topology. In dynamic networks, transient failures such as disconnection of links, loss of messages, and alteration of memory contents due to disturbances occur frequently. We designed self-stabilized distributed algorithms with safety convergence properties. They are resistant to such failures and changes. In order to guarantee the quality of service in the legitimate configuration, it is necessary to guarantee the approximation rate for the optimal legitimate configuration. Safe convergence property is a property that converges quickly to a feasible configuration in which the minimum service can be guaranteed, and converges to a n optimal configuration with a good approximation rate while maintaining that safety property.

Free Research Field

分散アルゴリズム

Academic Significance and Societal Importance of the Research Achievements

本研究では動的ネットワークに適した分散アルゴリズムの構築を行った．これにより，モバイルアドホックネットワークやセンサーネットワークなどの動的ネットワークにおける通信経路の計算やクラスタリングに大きく貢献することが期待される．自己安定分散アルゴリズムの設計は一般的に容易ではない．しかし，自己安定分散アルゴリズムの有用性は大変大きく，ネットワーク上の基本的な問題であるクラスタリングや経路探索にその性質を適応させることは重要な問題であると考える．