互联网自治域商业关系推测算法

Inference Algorithms of Internet Autonomous Systems Business Relationships

随着互联网进入商业化发展阶段,不同管理机构的自治域通过商业合作关系来共同路由互联网中的流量.从工程上来看,获取全面准确的自治域商业关系可以指导有效的流量工程管理和资源部署从而均衡网络负载,提高服务可靠性,减少网络运营成本;从理论上来看,获取全面准确的自治域关系可以分析互联网的演进模型,指导设计新的路由协议和互联网架构.虽然互联网路由注册中心会记录自治域之间的商业关系,但是由于商业敏感以及注册中心缺乏维护更新,公开的自治域商业关系并不全面准确,因而通过推测算法获得更加全面准确的自治域商业关系成为了学术界与产业界关注的热点.文中系统地综述了现有的互联网自治域商业关系推测算法及其存在的问题,并对该领域有待进一步研究的问题和可能的解决方案给出了一些建议.

With the Internet’s entry to the commercialization stage,Autonomous Systems （ASs）operated by different administrative domains collaborate on routing traffic through the businessrelationship.Accurate and complete AS relationships,from engineering perspective,can helpguiding effective traffic engineering and resource placement,thus optimizing network load,increasing service reliability and reducing network operating costs.From theory perspective,itcan be used to analyze Internet evolvement model,guide design of new routing protocol and Internetarchitecture.Although Internet Routing Registry （IRR）records the business relationshipsbetween ASs,due to the commercial sensitivity and the lack of maintenance and updates in IRR,public business relationships between ASs are not accurate and complete,thus obtaining ASbusiness relationships through inference algorithms has been a hot issue of academic and industrialcircles.This paper systematically surveys the existing inference algorithms of AS business rela-tionships and their shortcomings.Furthermore,this paper addresses the research trends,andpresents the potential solutions.