earticle

논문검색

Quantifying AS Path Inflation by Routing Policies

초록

영어

A route in the Internet may take a longer AS path than the shortest AS path due to routing policies. In this paper, we systematically analyze AS paths and quantify the extent to which routing policies inflate AS paths. The results show that AS path inflation in the Internet is more prevalent than expected. We first present the extent of AS path inflation observed from the RouteView and RIPE routing tables. We then employ three common routing policies to show the extent of AS path inflation. We find that No-Valley routing policy causes the least AS path inflation among the three routing policies. Prefer-Customer-and-Peer-over-Provider policy causes the most AS path inflation. In addition, we find that single-homed stub ASes experience more path inflations than transit ASes and multi-homed ASes. The AS pairs with shortest AS path of 3 AS hops experience more path inflations than other AS pairs. Finally, we investigate the AS path inflation on the end-to-end path from end users to two popular content providers, Google and Comcast. Although the majority of the shortest AS paths from end users to the two providers consists of no more than three AS hops, the actual end-to-end paths that the traffic will take are longer than the shortest AS paths in many cases. Quantifying AS path inflation in the Internet has important implications on the extent of routing policies, traffic engineering performed on the Internet, and BGP convergence speed.

목차

Abstract
 1. Introduction
 2. AS Path Inflation
  2.1. Data Sets
  2.2. Multiple Vantage Points
  2.3. AS Path Inflation Measurement
  2.4. Impact of Missing Links on AS Path Inflation
 3. AS Path Inflation by Routing Policies
  3.1. Routing Policies
  3.2. Three Common Routing Policies
  3.3. Computing Policy-conforming Paths
  3.4. AS Path Inflation due to Routing Policy
  3.5. Accuracy of Inferred Policy-conforming AS Paths
  3.6. Path Inflation Changes
 4. Path Inflation at Google and Comcast
 5. Conclusions
 References

저자정보

  • Qixin Gao Department of Computer Science, Northeastern University, China
  • Feng Wang School of Engineering and Computational Sciences, Liberty University, USA
  • Lixin Gao Department of Electrical and Computer Engineering, University of Massachusetts, Amherst, USA

참고문헌

자료제공 : 네이버학술정보

    함께 이용한 논문

      ※ 원문제공기관과의 협약기간이 종료되어 열람이 제한될 수 있습니다.

      0개의 논문이 장바구니에 담겼습니다.