2018 Fiscal Year Final Research Report
Analysis of distributed load balancing algorithms on inhomogeneous and dynamic networks
| Project/Area Number |
17H07116
|
|---|
| Research Category |
Grant-in-Aid for Research Activity Start-up
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Theory of informatics
|
|---|
| Research Institution | Chuo University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2017-08-25 – 2019-03-31
|
| Keywords | 負荷分散 / マルコフ連鎖 |
|---|
| Outline of Final Research Achievements |
Given a distributed network where each vertex has discrete loads, we consider the problem of minimizing the discrepancy between the maximum and minimum loads among all vertices. For this problem, diffusion based algorithms are well studied because of its simplicity. In a diffusion based algorithm, at each synchronous and discrete time step, each vertex is allowed to distribute its loads to each neighbors.
This work is concerned with the ability of natural diffusion based algorithms. We presents a new randomized diffusion algorithm like multiple random walks. Our algorithm achieves a sub linear upper bound of the maximum degree for any graphs.
|
| Free Research Field |
アルゴリズム理論
|
| Academic Significance and Societal Importance of the Research Achievements |
昨今のネットワークの大規模化, P2P ネットワークのような分散型ネットワークの出現により, ネットワーク上の各プロセッサにかかる負荷を均等に配分しなおすことを目指す「ネットワーク上の負荷分散」アルゴリズムは近年その需要が増している. しかし, 単純なアルゴリズムに対しても, その解析の困難さより, 既存研究では正則グラフをはじめとした限られたグラフ構造上での解析に留まっている.
本研究では主にマルコフ連鎖の過渡解析技法を適用し, 既存のモデルを非正則なグラフ上でも動作するように拡張・更にはその誤差に対する制度保証を与えた. 非正則なグラフ上での成果はほぼ初であり, その意義は大きい.
|
