Networking Technology for Peer-to-Peer service on Power Law Networks.
| Project/Area Number |
17560338
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Communication/Network engineering
|
|---|
| Research Institution | Kyoto University |
|---|
Principal Investigator |
ASUKA Takuya Kyoto University, Graduate School of Informatics, Associate Professor, 情報学研究科, 助教授 (30346043)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TAKAHASHI Tatsuro Kyoto University, Graduate School of Informatics, Professor, 情報学研究科, 教授 (30324677)
|
|---|
| Project Period (FY) |
2005 – 2006
|
| Project Status |
Completed (Fiscal Year 2006)
|
| Budget Amount *help |
¥3,500,000 (Direct Cost: ¥3,500,000)
Fiscal Year 2006: ¥1,800,000 (Direct Cost: ¥1,800,000)
Fiscal Year 2005: ¥1,700,000 (Direct Cost: ¥1,700,000)
|
|---|
| Keywords | Peer-to-Peer / Power-law / Network Topology / Search / Logical Network / Traffic Measurement / オーバレイ / ネットワーク / トラヒック制御 / オーバレイネットワーク |
|---|
| Research Abstract |
P2P(Peer-to-Peer) applications are becoming popular. We focus on Gnutella-like unstructured P2P systems, and propose two efficient controls for P2P networks based on diversified user's taste in contents.
The first proposal is "load balancing with content classification in P2P networks". Recent measurements of P2P networks for file sharing show that they follow power-law topology. P2P networks which follow power-law topology have a disadvantage that the overload burden on peers with high degree. Moreover file popularity in P2P also follows distributions close to power-law. Thus we propose an efficient cache replacement policy for unstructured-P2P networks with power-law distribution, where high degree nodes caches low popularity contents, and low degree nodes high popularity contents. The policy can significantly reduce the load of very-high degree nodes and improves hit rate of caches by using cache replacement in consideration of content classification according to degree which means number of neighboring peers.
The second proposal is "semantic Peer-to-Peer networks based on proximity". A lot of studies has been done to improve content search efficiency or to reduce download time in file sharing of very large files. The proposal tries to improve both efficiency and download time by reconstructing overlay networks using proximity of content taste of users and round trip delay between peers.
The performances of proposals are evaluated and proved by computer simulation.
|
Report
(3 results)
Research Products
(11 results)
