{"title":"IP/NDN:多层次的转换和迁移机制","authors":"Sheng Luo, Shangru Zhong, Kai Lei","doi":"10.1109/NOMS.2018.8406213","DOIUrl":null,"url":null,"abstract":"Different from TCP/IP architecture, Named Data Networking (NDN) adopts a data-centric and in-network caching approach to achieve improved network efficiency and reduced traffic redundancy. As NDN begins to incrementally deploy on the real-world, there will be a hybrid network of TCP/IP and NDN during the transitional period. However, redeveloping TCP/IP- based applications for NDN is a daunting and time-consuming task. Thus, a more sensible and economical approach is to design an effective strategies for leveraging NDN to deliver IP datagrams and porting TCP/IP-based applications on NDN without changing its original code. Towards this end, this paper introduces three different migration methods at Internet layer, TCP layer and application layer, respectively to translate TCP/IP-based packets into NDN packets and run TCP/IP-based applications on NDN. By implementing these methods on a real NDN test-bed, we demonstrate the multi-level IP/NDN translation and migration mechanism is feasible. The analysis of testing results shows that each of the proposed methods is valid with varying benefits and costs: translating at TCP layer with more time overhead can have a less packets overhead than Internet layer. And when translated at application layer, the network load is reduced by half because of the in-network cache.","PeriodicalId":19331,"journal":{"name":"NOMS 2018 - 2018 IEEE/IFIP Network Operations and Management Symposium","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2018-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"8","resultStr":"{\"title\":\"IP/NDN: A multi-level translation and migration mechanism\",\"authors\":\"Sheng Luo, Shangru Zhong, Kai Lei\",\"doi\":\"10.1109/NOMS.2018.8406213\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Different from TCP/IP architecture, Named Data Networking (NDN) adopts a data-centric and in-network caching approach to achieve improved network efficiency and reduced traffic redundancy. As NDN begins to incrementally deploy on the real-world, there will be a hybrid network of TCP/IP and NDN during the transitional period. However, redeveloping TCP/IP- based applications for NDN is a daunting and time-consuming task. Thus, a more sensible and economical approach is to design an effective strategies for leveraging NDN to deliver IP datagrams and porting TCP/IP-based applications on NDN without changing its original code. Towards this end, this paper introduces three different migration methods at Internet layer, TCP layer and application layer, respectively to translate TCP/IP-based packets into NDN packets and run TCP/IP-based applications on NDN. By implementing these methods on a real NDN test-bed, we demonstrate the multi-level IP/NDN translation and migration mechanism is feasible. The analysis of testing results shows that each of the proposed methods is valid with varying benefits and costs: translating at TCP layer with more time overhead can have a less packets overhead than Internet layer. And when translated at application layer, the network load is reduced by half because of the in-network cache.\",\"PeriodicalId\":19331,\"journal\":{\"name\":\"NOMS 2018 - 2018 IEEE/IFIP Network Operations and Management Symposium\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-04-23\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"8\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"NOMS 2018 - 2018 IEEE/IFIP Network Operations and Management Symposium\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/NOMS.2018.8406213\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"NOMS 2018 - 2018 IEEE/IFIP Network Operations and Management Symposium","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/NOMS.2018.8406213","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 8
摘要
与TCP/IP架构不同,NDN (Named Data Networking)采用以数据为中心的网内缓存方式,提高网络效率,减少流量冗余。随着NDN在现实世界中的逐步部署,过渡时期将出现TCP/IP和NDN的混合网络。然而,为NDN重新开发基于TCP/IP的应用程序是一项艰巨而耗时的任务。因此,一个更明智和经济的方法是设计一个有效的策略,利用NDN来传递IP数据报,并在NDN上移植基于TCP/IP的应用程序,而不改变其原始代码。为此,本文分别在互联网层、TCP层和应用层介绍了三种不同的迁移方法,将基于TCP/ ip的数据包转换为NDN数据包,并在NDN上运行基于TCP/ ip的应用程序。通过在实际的NDN测试平台上实现这些方法,我们证明了多级IP/NDN转换和迁移机制是可行的。对测试结果的分析表明,每种方法都是有效的,并且具有不同的收益和成本:在时间开销较大的TCP层进行转换可以比Internet层进行转换获得更少的数据包开销。当在应用层进行转换时,由于网络内缓存,网络负载减少了一半。
IP/NDN: A multi-level translation and migration mechanism
Different from TCP/IP architecture, Named Data Networking (NDN) adopts a data-centric and in-network caching approach to achieve improved network efficiency and reduced traffic redundancy. As NDN begins to incrementally deploy on the real-world, there will be a hybrid network of TCP/IP and NDN during the transitional period. However, redeveloping TCP/IP- based applications for NDN is a daunting and time-consuming task. Thus, a more sensible and economical approach is to design an effective strategies for leveraging NDN to deliver IP datagrams and porting TCP/IP-based applications on NDN without changing its original code. Towards this end, this paper introduces three different migration methods at Internet layer, TCP layer and application layer, respectively to translate TCP/IP-based packets into NDN packets and run TCP/IP-based applications on NDN. By implementing these methods on a real NDN test-bed, we demonstrate the multi-level IP/NDN translation and migration mechanism is feasible. The analysis of testing results shows that each of the proposed methods is valid with varying benefits and costs: translating at TCP layer with more time overhead can have a less packets overhead than Internet layer. And when translated at application layer, the network load is reduced by half because of the in-network cache.