对移动IPv6支持的综合高效切换程序 外文资料翻译

上传人:仙*** 文档编号:28023347 上传时间:2021-08-22 格式:DOC 页数:28 大小:372.50KB
返回 下载 相关 举报
对移动IPv6支持的综合高效切换程序 外文资料翻译_第1页
第1页 / 共28页
对移动IPv6支持的综合高效切换程序 外文资料翻译_第2页
第2页 / 共28页
对移动IPv6支持的综合高效切换程序 外文资料翻译_第3页
第3页 / 共28页
点击查看更多>>
资源描述
本科毕业设计(论文)外文资料翻译 外文资料题目 A Comprehensive and Efficient Handoff Procedure for IPv6 Mobility Support对移动IPv6支持的综合高效切换程序摘要移动 IPv6 路由优化的切换性能取决于优化的 IP 层自动机制,以及安排和并行化及其信号的移动节点的灵活性。本文提供了标准的 ipv6 协议套件和移动 IPv6切换性能综合分析延迟的几个来源。虽然一些拖延已是众所周知的优化和广泛适用的切换方法是尚未被发现。本文因此继续讨论现有和新的优化建议,其中一些目前正在内 IETF、标准化,并阐述如何组合的那些可以 显著的提高切换的经验。1 .介绍互联网服务渗透的日常生活越来越多,用户越来越希望他们能够在任何地点、任何时间使用互联网。同时,实时通信的重要性正在增加,如音频和视频流媒体、 IP 电话、 视频会议等实时通信都是高度敏感数据延迟,传播延迟和切换延迟。高效地移动支持是下一代互联网建设过程中的主要目标,设计和路由优化的一种模式被纳入移动 IPv6移动性协议。路由优化允许直接路径通过。这是补充路由通过固定的代理服务器,其代理移动节点修建的经典的方法。当路由优化减少传播延迟的同时,切换延迟仍然相当有效地排除了有意义的实时支持。事实上,标准的 IPv6 部署中的切换延迟大约是几秒。这不只是因为移动 IPv6,但也会影响标准 IPv6 和运动检测机制。非常幸运的是,最近把大量的优化技术提出以简化个人切换相关的活动。测量数据证实任何特殊技术化都是有利的。但是,研究如何优化集成到目前为止已经很大程度上被忽视的。本文从较高的角度来看流动性检查的挑战 它解释标准的 IPv6 部署 IP 层视野中的整体切换过程和分析方面的期望。由于结果强烈建议优化,文件将继续探索有前途的现有和新的建议,最近获得了互联网工程任务组 (IETF) 和学术研究界的势头。对他们的相互作用也计算优化。本文建议改进的切换性能的集成的解决方案。2 .标准切换过程 A 移动节点经历 IP 层切换,或者只需切换,当它更改 IP 连接。这开头链路层附件中的变化,还链路层切换,后面跟着发现新路由器、地址、运动检测和 移动 IPv6 登记。图 1 说明了这些切换的步骤,分别讨论下一步了。2.1 路由器发现 A 移动节点在路由器发现的过程中学会有关本地路由器和链路上的前缀。这一过程被透过松散定期上的链路本地节点的路由器多播路由器广告邮件。IPv6 邻居发现 RFC 指出未经请求的路由器广告邮件发送 3 和 4 秒之间的随机间隔至少和 1350年 1800年秒之间最。由于这些保守的限制旨在平稳的节点,而不能有意义地支持移动,移动 IPv6 RFC 减少下限,一个信标每 30 至 70 毫秒为单位)。因此,移动节点可以期待收到 25 毫秒之后的第一次切换的广告,这减少了50 毫秒连续广告之间的平均时间。另一方面,高频率的多址广播广告可能是在低带宽、 广域网络,其中许多用户可能经常离开受同一 IP 子网的地理区域的问题。图 1 显示的广告相关的切换过程中 ;掩盖了其他的广告。2.2 地址配置移动节点配置新的全球 IP 地址收到的未知的前缀路由器广告邮件。这种情况通常发生符合无状态地址自动: 移动节点或者随机选择的接口标识符,或基于接口的 MAC 地址,及在此将添加前获得的前缀。然后,它会发送多播侦听程序报告消息要订阅请求节点多播组相对应的新地址。如果路由器通告消息是多播的传输,通常是这种情况,多播侦听程序报告消息被延误第二解脱与相邻的节点,可能对相同的广告作出的反应。移动节点,然后运行重复地址检测协议,以验证是否是唯一的新地址: 它传输地址的邻居请求消息,并且,如果在 1 秒的时间内收到没有响应,为标志的接口的地址。如果是一个唯一的地址,因此总 期范围 1 和 2 秒之间。可能已在另一个节点使用 IPv6 地址的可能性很小,可以使其可以忽略不计。即使该链接本地地址保持其前缀在切换过程中,移动节点必须重新仍验证此地址的唯一性时 IP 连接更改,因为新的链路上的节点可能已经使用相同的链接本地地址。这是通过另一个运行的重复地址检测。因为只有运动检测可以建立 IP 连接是否已更改,re-核查的链接本地地址通常运动检测后开始。这不是显示在图像,但是,鉴于链接本地地址的可用性并不是影响其他切换活动的日程安排。2.3 运动检测移动节点执行运动检测,以识别 IP 连接的更改。这种变化意味着移动节点选择新的默认路由器、 无效陈旧的全局地址,其链路本地地址,再核实唯一性和启动移动 IPv6 登记。运动检测依赖于分析广告路由器广告邮件中的链接上前缀和可能还探讨了路由器考虑关闭链接的可到达性。当移动节点使用的前缀不再被视为刊登公告,但新的前缀 显示相反时,移动节点通常决定它已移动到不同的网络。另一方面,收到的前缀 也可能表示 IP 连接未改变链接层切换,尽管中例如,当移动节点交换机连接到同一子网的接入点。运动检测被复杂的路由器通告消息可能包含不完整集前缀 的这一事实。接待处的单个因此通常是广告的不足,以决定是否已更改 IP 连接。它也是通常不可能确定当广告应已收到,但没有出现,以保证的广告时间间隔的缺乏。移动 IPv6 RFC 帮助,在这方面,它引入了路由器广告邮件广告时间间隔选项。路由器使用此选项可以指示对其信标期间的上限。这是低至 70 毫秒 (参见第 2.1 节),为了计算调度中移动节点和路由器的粒度添加额外 20 毫秒。然后,移动节点期望路由器通告消息到达的最 90 毫秒的时间间隔。不过,没有一种单一的预期广告仍然并不意味着变化可能给数据包丢失的 IP 连接。三个丢失的广告更可靠地表明运动。然后决定最 270 毫秒后从旧的默认路由器接收到最后的广告。实际的链路层切换稍后就会出现最多 70 毫秒,以便运动检测可以采取任何 200 和 270 毫秒之间的时间。平均而言,从旧的默认路由器的最后一个广告的接待和链路层切换期间是 25 毫秒,收益率平均运动检测 245 毫秒的延迟。2.4 移动IPV6注册移动 IPv6 注册后地址 和运动检测移动节点选择其新的全球地址注册转交地址作为其家乡代理和对应的节点之一。这样就建立关怀的地址与移动节点的家乡地址,已从家乡代理的网络的前缀和跨运动保持稳定之间的绑定。住宅地址 IP 之上的堆栈层用作终点 鉴定的一部分。与同行交流的移动节点的数据包有 IP 报头中的护理的地址和在电线上的 IPv6 扩展标头中的家庭住址。同时结束节点交换地址时遍历的数据包的 IP 层这样的运输协议和应用程序可以访问的住址,像往常一样。图 1 说明了家乡代理和记者的单个节点的移动 IPv6 的注册过程。首页注册包含绑定更新消息的通告新转交地址和绑定确认的一条消息,指示成功或失败的家乡代理。必须注意防止非法绑定,哪些恶意的节点可以尝试模拟或重定向的目的建立基于洪水。移动节点和家乡代理通常根据相同的管理和共享凭据,以引导 IPsec 安全关联。首页注册可以因此被身份验证和加密。记者注册允许路由优化。它包括传达对应的节点,并响应的绑定的确认新转交地址的绑定更新消息。这些无法一般保护通过 IPsec,不过,因为移动节点均可能要共享的身份验证凭据,他们可能在某一时刻的所有相应节点进行沟通,也不是全局公钥基础结构,可任意对的节点,才能投入存在任何时间很快 。记者注册而是通过身份验证和授权通过返回路径能力的程序,基于非加密的家庭和照顾的地址在移动节点的可访问性的核查。在这两个地址可达凭移动节点,以启动地址之间的绑定。家庭地址测试中,移动节点隧道主测试初始化消息到家乡代理,将转发到通信节点的消息。通信节点返回到首页测试邮件内的家庭住址的不可预知的家庭关键一代令牌。家乡代理截获此消息和隧道它到移动节点。护理的地址测试是移动节点和通信节点之间的直接交流。它包括照顾的测试初始化消息和照顾的测试消息具有不可预知的护理的关键一代令牌。国内外护理的关键一代标记的知识证明分别接收数据包的家庭住址和照顾的地址,在移动节点的能力。移动节点通过使用来自这两个标记密钥验证通信节点的绑定更新消息演示了这方面的知识。通信节点使用相同的密钥进行身份验证的最后绑定的确认消息。移动 IPv6 RFC 叶片对调度信号和数据包中移动节点的自由。图 1 显示了一个保守的移动节点,等待来自家乡代理的绑定的确认消息之前它启动的返回路径能力的过程。与此相反,乐观的移动节点并行执行首页注册和回报击溃能力的过程。乐观的移动节点而且开始发送数据包到通信节点等通信节点的绑定更新消息已在路上,而保守的移动节点接收的确认后,才使用的新的护理的地址。在任一情况下,直到它接收到的绑定更新消息,不知道有新的转交地址的通信节点。rst 数据包发送到新的护理的地址将因此会送交移动大约随绑定的确认消息,假设这一要求移动节点的节点。首页注册失败的情况下,保守的移动节点避免无用的返回路径能力的过程。他们亦不可能不久后丢失或被拒绝的绑定更新消息发送的数据包的丢失。对应的节点将丢弃这些安全措施绑定不匹配的面孔的数据包。这是额外的切换延迟为代价的当两次注册成功。传出路由优化的包,这是移动节点和家乡代理加移动节点和通信节点之间的往返时间之间的往返时间。传入的数据包,额外的切换延迟是移动节点和家乡代理之间的往返时间。乐观的移动节点性能更好,一般情况下。但他们可能白费尝试返回路径能力过程或有包损失应家庭或代理注册失败。3 .基于现有条件的解决方法现有和改善的切换性能的标准切换过程可以显著延迟损害质量的实时应用程序,即使路由优化设计意图改善建议方法对这些应用程序支持。研究社会一直在努力减少一些时候的切换延迟,并取得了多项建议。特别是有前途的是下列方法。3.1 路由器发现更复杂的调度路由器中的时间间隔可以提高路由器发现对带宽消耗和限度地优化。加快融入允许移动节点请求立即的广告。当移动节点的链路层可以指示在网络连接中的更改时,这非常有用。基于链路上的路由器的链接本地地址和请求的来源地址,每个路由器自主计算动态的排名,指示路由器应响应立即和可能的其它路由器应该不久发送更多的广告。快速路由器发现建议接入点重播高速缓存的路由器广告邮件,一旦相关联的节点。这种网络一侧的链路层支持消除了在移动节点的链路层触发器的要求。3.2 地址配置为了避免手关闭造成的延误标准重复地址检测,取得了地址 不同的提案。IPv6 工作组内的 IETF 正在乐观重复地址检测 ,它允许有限使用可能重复的 IP 地址。移动节点暂时更改的规则,他们做了 IPv6 邻居发现,以免污染可能是非法的地址解析信息与其他节点的邻居高速缓存的信号。先进的重复地址检测,路由器生成,他们再分配给移动节点的唯一地址的池。重复地址检测执行地址上提前这样,移动节点可以配置他们立即无需验证自己的唯一性。3.3 运动检测 DNA 内 IETF 工作组处理慢运动检测的两种互补的方法的问题。前缀的完整列表协议适用于路由器。移动节点维护学术上链接前缀,可能获得多个路由器通告消息接收的列表。列表中已经成熟了一会儿后,移动节点可以假定更改 IP 连接高概率时新收到的广告仅包含列表中没有的前缀。这种预测基于可能不完整的信息,所以移动节点可能会在没有实际发生时,甚至断言运动。DNA 协议使用加快融入及时传输的请求路由器广告邮件。路由器选择某些前缀作为链接标识符,并为此所显示的所有传播广告中。这允许移动节点,可靠地检测中基于单个广告的 IP 连接的更改。另外,移动节点可以显式检查路由器征集广告交换的一部分用于前的链路层切换,因此称为一个里程碑,网络 prex 是否仍有效的可能是新的链接。DNA 协议作为链接与旧式的路由器的回退机制集成了 前缀的完整列表。3.4 移动 IPv6 优化许多移动 IPv6 优化减少路由优化的切换延迟,通过修改程序的返回路径的能力。早期绑定更新和基于信用的授权的组合达到这一点,对纯粹的端到端的基础,具有以下四个组成优化: 1。 主动家庭地址测试: 移动节点主动的家庭地址测试期间获取未来的切换回家的注册机里的标记。这关键的切换期间保存通过家乡代理可能长的往返行程。移动节点可以调用只是时间的基础上积极家庭地址测试,如果其链路层提供指示即将切换,触发器或定期每当最近取得的最大主关键一代令牌将过期。2.同时照顾的地址测试: 数据包可以已交换,在有限的程度上,通过新的护理的地址,而在该转交地址移动节点的可达性正在北朝鲜。3.试绑定: 移动节点注册其家乡地址与地址未经核实的护理的暂定绑定通过交换与通信节点的早期绑定更新和早期绑定的确认消息。仅以家庭的关键一代令牌索取最近的主动家庭地址测试,从而促进后续的并行的护理的地址测试情况下,消息进行身份验证。一旦移动已执行并行的护理的地址测试,它对标准的绑定更新消息进行身份验证,并与通信节点注册北朝鲜转交地址。4.家庭和记者注册的并行 移动 IPv6规格不允许绑定更新消息发送到对应的节点之前确认收到来自家乡代理, 移动节点。如果结合的主动家庭地址测试和并发护理的地址测试隐藏程序的返回路径能力滞后时间,这就会成为性能问题。移动 IPv6 的规则因此放宽,以允许移动节点,早期绑定更新的讯息,首页注册时仍然挂起。著名安全指引禁止向的可达性尚未北朝鲜转交地址发送数据包的通信节点。这是防范恶意,否则可能诱供水请求数据包的第三方对应的节点的节点。这种基于重定向的洪水攻击的吸引力是工艺扩大的潜力。例如,攻击者可以完成初始的 TCP 握手,自己地址 (或家庭住址,这件事),通过大量大 128gb 下载,然后将排放重定向到其受害者的地址。攻击者可能,并会,以基于它在初始握手期间学到的序列号的受害者的名义欺骗确认。但确认将小比作通信节点生成的数据段。基于信用的授权防止基于重定向的放大的洪水攻击,但可以通过为核实转交地址的双向通信。通信节点维护一个字节计数器的移动节点,也称为移动节点的信贷,从移动节点接收的数据量的增加而减少发送到移动节点的转交地址是未核实的数据量。指数老化保证现有信贷表示只最近收到移动节点的数据。当通信节点的移动节点的数据包时,发送至转交地址如果地址是北朝鲜,或者该地址是未核实,但数据包的大小不超过当前可用的信贷。否则为通信节点可能丢弃数据包、 缓冲它直到转交地址变已查清,或将其发送给家乡的地址。其他路线优化增强功能需要某种形式的预: 端节点共享密钥或安全关联,更多的效率,引导的凭据,并加密的身份验证可以取代家庭地址测试。这两项建议目前正在讨论中,IETF,预配置与共享、 秘密身份验证密钥的移动节点和对应的节点。使用 IPsec 和互联网密钥交换协议。这些技术患可扩展性问题,但是,鉴于端节点必须设置使用成对的凭据。此外,技术都提供了核查的移动节点的可访问性,所以都不能从技术上讲没有照顾的地址测试。端节点对等方的可访问性的信任并可进一步忽略照顾的地址测试,但这种信任是在很多重要的商业模式中不可用。例如,移动电话运营商可能能够用秘密身份验证密钥,配置订阅服务器,但可能无法偿还所有订阅服务器使用这些注册表项,以可靠的方式。移动 IPv6 优化的另一家基于移动节点的访问网络中的路由器支持。在快速切换的移动 IPv6正在部署,移动节点可以请求其当前的默认路由器建立双向隧道到一个新的护理的地址。这同时允许暂时沟通通过其旧的转交地址后切换,和注册新的照顾其家乡代理地址的移动节点和对应的节点。代理路由器发现和辅助的地址的帮助,移动节点可能会要求隧道之前切换,只要它可以预测变动。附加功能优化的意外的链接中断的情况下无功的切换管理。与此相反,媒体独立预身份验证使用旧的转交地址和一个新的默认路由器之间的双向隧道。移动节点分配新转交地址从远程和影响家庭和记者注册之前,它将更改链接。如果相邻单元格之间的重叠是充分的大允许切换准备工作及时完成,化的这种做法是类似的快速切换。然而,细胞重叠在哪里小相对于节点的速度,推迟全球后切换到一个阶段信号是有利,因为无线信号质量则通常更高和更持久。媒体独立预身份验证的力量是进行预身份验证切换之前,新的网络移动节点的能力。分层移动 IPv6可以绑定到一个更稳定的区域护理的地址的其当前的链路上照顾的地址从移动锚点网络的移动节点位于其他位置访问域中。移动节点发送和接收数据包通过双向隧道本身和移动锚点之间的区域护理的地址。它对应的节点,家乡代理注册的区域转交地址,并更新移动锚点,每当运动后的变化及其对链接的地址。运动可以使隐藏从家乡代理和对应的节点,只要移动节点在同一移动锚点领域内的山巅。5. 结论高效端到端切换需要优化,不仅为移动性协议,路由器发现、地址和运动检测。在今天的 IPv6 协议标准,探讨了自己的缺点和各种现有的优化,审查这些交互,以及如何可以将它们合并到一个完整和高效的移动解决方案。它是必须认识到明天的流动性支持的基础是今天提交。这尤其适用于路径的优化,并需要从两个同行的支持,因此取决于相应的节点中实现了坚实的基础。因此应包括路由优化功能早期新兴 IPv6 堆栈。增强功能还必须访问路由器,其响应能力至关重要的高效IPv6 和运动检测到他们的方式。越早的必要优化的一组被广泛接受这套将最终的可能性会无处不支持。A Comprehensive and Efficient Handoff Procedure for IPv6 Mobility SupportAbstractHandoff performance with Mobile IPv6 Route Optimization strongly depends on the efciency of IP-layer autoconguration mechanisms as well as the exibility of mobile nodes to schedule and parallelize their signaling. This paper provides a comprehensive analysis of the handoff performance with the standard IPv6 protocol suite and Mobile IPv6, and it identies several sources for delay. While some of the delays are already well known, an optimized and widely applicable handoff approach is yet to be found. The paper hence proceeds to discuss existing and new optimization proposals, some of which are currently under standardization within the IETF, and elaborates how a combination of those can signicantly improve handoff experience.1.IntroductionAs Internet-based services pervade daily life more and more, users increasingly desire them to be accessible at any place and any time. At the same time grows the importance of real-time communications 19 such as audio and video streaming, IP telephony, or video conferencing. Realtime communications are highly delay-sensitive and exhibit a susceptibility to long propagation latencies and handoff delays. Efcient mobility support was hence amongst the primary objectives during the design of the next-generation Internet, and a mode for Route Optimization was incorporated into the Mobile IPv6 9 mobility protocol. Route Optimization allows peers to communicate via a direct path.This complements the classic approach of routing a mobile nodes trafc through a stationary proxy, its home agent.While Route Optimization mitigates the problem withpropagation latencies, handoff delays are still substantial enough to effectively preclude meaningful real-time support 2, 12, 13. In fact, handoff delays in a standard IPv6 deployment are in the order of seconds. This is not only due to Mobile IPv6, but also affects standard IPv6 conguration and movement-detection mechanisms 15, 24.Very fortunately, a multitude of optimization techniques3, 5, 10, 14, 18 have recently been put forth to streamline individual handoff-related activities. Measurement data is typically available to corroborate the benets of any specic technique. But a study of how well the optimizations integrate has so far been largely neglected 1.This paper examines the challenges with mobility from a higher perspective: It explains the overall handoff procedure in a standard IPv6 deployment from an IP layers perspective and analyzes inhowfar it falls short of expectations. Since the results strongly advise optimization, the paper proceeds to explore promising existing and new proposals that have recently gained momentum in both in the Internet Engineering Task Force (IETF) and the academic research community. The optimizations are also evaluated with respect to their interactions. The paper nally proposes an integrated solution for improved handoff performance.2. Standard Handoff ProcedureA mobile node undergoes an IP-layer handoff, or simply a handoff, when it changes IP connectivity. This begins with a change in link-layer attachment, also referred to as a link-layer handoff, and is followed by the discovery of newrouters, address conguration, movement detection, and nally Mobile IPv6 registrations. Figure 1 illustrates these handoff steps, which are separately discussed next.2.1 Router DiscoveryA mobile node learns about local routers and on-link prexes during router discovery. This process is facilitated through Router Advertisement messages, which routers multicast to link-local nodes on a loosely periodic basis.The IPv6 Neighbor Discovery RFC 16 states that unsolicited Router Advertisement messages are to be sent in random intervals of between 3 and 4 seconds at least and between 1350 and 1800 seconds at most. Since these conservative limits are tailored towards stationary nodes and fail to meaningfully support mobility, the Mobile IPv6 RFC decreases the lower bound to one beacon every 30 to 70 milliseconds. This reduces the mean time between successiveadvertisements to 50 milliseconds so that a mobile node can expect to receive the rst post-handoff advertisement after 25 milliseconds. On the other hand, high frequencies for multicast advertisements may be an issue in low-bandwidth,wide-area networks, where many users may not frequently leave the geographic area covered by the same IP subnet.2.2 Address CongurationA mobile node congures a new global IP address upon receipt of a Router Advertisement message with an unknown prex. This typically happens in compliance with Stateless Address Autoconguration 22: The mobile node chooses an interface identier, either randomly or based on the interfaces MAC address, and prepends to this the obtained prex. It then sends aMulticast Listener Report message 4 to subscribe to the solicited-node multicast group corresponding to the new address. If the Router Advertisement message was a multicast transmission, which usually is the case, the Multicast Listener Report message is de-layed by up to 1 second to desynchronize with neighboring nodes that may be reacting to the same advertisement. The mobile node then runs the Duplicate Address Detection protocol to verify whether the new address is unique: It transmits a Neighbor Solicitation message for the address and, if no responses are received within a period of 1 second, assigns the address to the interface. The total conguration period hence ranges between 1 and 2 seconds if the address is unique. The probability for an IPv6 address to already be in use by another node is small enough to make it negligible. Even though the link-local address keeps its prex during handoff, the mobile node must still re-verify uniqueness of this address when IP connectivity changes, because a node on the new link may already be using the same linklocal address. This is done through another run of Duplicate Address Detection.Since only movement detection can establish whether IP connectivity has changed, re-verication of the link-local address typically begins after movement detection. This is not shown in gure 1, however, given that the availability of the link-local address does not inuencethe schedule for other handoff-related activities.2.3 Movement Detection Mobile nodes implement movement detection to recog-nize changes in IP connectivity. Such a change implies that a mobile node chooses a new default router, invalidates stale global addresses, re-veries uniqueness of its link-local address, and initiates Mobile IPv6 registrations. Movement detection relies on analyzing the on-link prexes advertised in Router Advertisement messages and possibly also probing reachability of routers considered off-link. When the prexes in use by the mobile node are no longer seen to be advertised, but new prexes show up instead, the mobile node typically decides that it has moved to a different network. On the other hand, received prexes may also indicate that IP connectivity did not change in spite of a linklayer handoff, e.g., when the mobile node switches access points that connect to the same subnet. Movement detection is complicated by the fact that Router Advertisement messages may include incompletesets of prexes. Reception of a single advertisement is therefore usually insufcient to decide whether IP connectivity has changed. It is also generally impossible to determine when an advertisement should have been received,but did not appear, due to the lack of a guaranteed advertisement interval. The Mobile IPv6 RFC helps in this respect in that it introduces an Advertisement Interval option for Router Advertisement messages. Routers use this option to indicate an upper bound on their beaconing periods. Wherethis is as low as 70 milliseconds (cf. section 2.1), an extra 20 milliseconds are added in order to account for scheduling granularities in mobile nodes and routers. Mobile nodes then expect Router Advertisement messages to arrive in intervals of at most 90 milliseconds.Nevertheless, the absence of a single expected advertisement still does not imply a change in IP connectivity given the potential for packet loss. Three missing advertisements indicate movement more reliably. A decision can then be made at most 270 milliseconds after the last advertisement was received from the old default router. The actual linklayer handoff may occur up to 70 milliseconds later, so movement detection can take any time between 200 and 270 milliseconds. On average, the period between reception of the last advertisement from the old default router and the link-layer handoff is 25 milliseconds, yielding a mean movement-detection delay of 245 milliseconds.2.4 Mobile IPv6 RegistrationAfter address conguration and movement detection, the mobile node selects one of its new global addresses to be registered as a care-of address with its home agent and correspondent nodes. This establishes a binding between the care-of address and the mobile nodes home address, which has a prex from the home agents network and remains stable across movements. The home address is used at stack layers above IP as part of end-point identication. Data packets that a mobile node exchanges with a peer have the care-of address in the IP header and the home address in an IPv6 extension header while on the wire. Both end nodes swap the addresses when a packet traverses the IP layer so that transport protocols and applications can access the home address as usual.Figure 1 illustrates the Mobile IPv6 registration procedure for the home agent and a single correspondent node.The home registration consists of a Binding Update message which noties the home agent of the new care-of address, and a Binding Acknowledgment message indicating success or failure. Care must be taken to preclude illegitimate bindings 17, which malicious nodes could attempt to establish for the purpose of impersonation or redirectionbased ooding. The mobile node and the home agent arebtypically under the same administration and pre-share credentials to bootstrap an IPsec security association. The home registration can so be authenticated and encrypted.The correspondent registration permits Route Optimization. It includes a Binding Update message that conveys the new care-of address to the correspondent node, and a responding Binding Acknowledgment message2. Thesecannot generally be protected through IPsec, however, because mobile nodes are neither likely to share authentication credentials with all correspondent nodes they may at some point communicate with, nor is a ”global” public-key infrastructure, available for arbitrary pairs of nodes, expected to come into existence any time soon 17. Correspondent registrations are instead authenticated and authorized through a return-routability procedure, based on non-cryptographic verication of a mobile nodes reachability at the home and care-of addresses. Reachability at both addresses entitles the mobile node to initiate a binding between the addresses.For the home-address test, the mobile node tunnels a Home Test Init message to the home agent, which forwards the message to the correspondent node. The correspondent node returns an unpredictable home keygen token to the home address within a Home Test message. The home agent intercepts this message and tunnels it to the mobile node. The care-of-address test is a direct exchange between the mobile node and the correspondent node. It consists of a Care-of Test Init message and a Care-of Test message with an unpredictable care-of keygen token. Knowledge of the home and care-of keygen tokens proves the mobile nodes ability to receive packets at the home address and care-ofaddress, respectively. The mobile node demonstrates thisknowledge by authenticating the Binding Update message for the correspondent node with a key derived from both tokens. The correspondent node uses the same key to authenticate the nal Binding Acknowledgment message.TheMobile IPv6 RFC leaves mobile nodes liberties withn respect to scheduling signaling and data packets. Figure1 shows a conservative mobile node, which waits for the Binding Acknowledgment message from the home agent before it initiates the return-routability procedure. In contrast, an optimistic mobile node executes the home registration and the return-routability procedure in parallel. An optimistic mobile node furthermore starts sending packets to the correspondent node as soon as the Binding Update message for the correspon
展开阅读全文
相关资源
正为您匹配相似的精品文档
相关搜索

最新文档


当前位置:首页 > 办公文档


copyright@ 2023-2025  zhuangpeitu.com 装配图网版权所有   联系电话:18123376007

备案号:ICP2024067431-1 川公网安备51140202000466号


本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。装配图网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知装配图网,我们立即给予删除!