Personal P. Roberts Compiler Internet Draft Title: draft-proberts-local-subnet-mobility- problem-01.txt Category: Informational May 2001 Expires : November 2001 Local Subnet Mobility Problem Statement Status of This Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at: http://www.ietf.org/ietf/1id-abstracts.txt The list of Internet-Draft Shadow Directories can be accessed at: http://www.ietf.org/shadow.html. Abstract This document specifies a problem statement for investigation within the IRTF. There are a couple of seemingly unrelated problems that have led investigators to converge on the desirability of implementing a routing protocol whose purpose is to allow a mobile node to retain connectivity via its current IP subnet while it moves within the scope of the micro mobility domain. In general this domain is expected to be contained within an autonomous system so that global aggregation of subnets is still preserved. An additional benefit of the protocols is that mobile nodes may more quickly react to failed links. The problem is to investigate the limits and issues with using (a) new protocol(s) to implement per node routes to facilitate better the movement of nodes and recovery of the network in the presence of failed links or routers. The document is a brief statement of the problem to be investigated with references to lots of other work that has been done already by many others in the area. 1 Internet Draft Local Subnet Mobility Problem Statement May 2001 1. Introduction The design of Internet routing has evolved largely based on the assumption that most end systems are stationary. As the idea of mobile devices came along, Mobile IP provided an easy migration path for supporting end system mobility without significant change or impact to the Internet infrastructure. This was achieved primarily by placing the support for routing to mobile devices on the end systems and at a limited number of centralized points within the infrastructure, effectively "hiding" the end system mobility from the infrastructure routing protocols. However, as we anticipate the desire to support real-time traffic flows to mobile devices and the possibility that mobile devices may become a significant portion of all Internet end nodes, investigation of alternative designs merit consideration. Many investigators in this area have converged on solutions that propose the use of local subnet mobility routing to support mobility within a limited domain, effectively exposing the mobility of end systems to the routers. This draft represents a problem statement to enable investigation of how a local subnet mobility routing protocol can be employed to enable mobility and fast network recovery. It is a problem statement based on a lot of work done in attempting to provide low latency handover in mobile networks and fast recovery in networks with fast recovery requirements. Please see the reference list for a large set of work that has been done in the past along these lines. A good summary of the problem as it pertains to network recovery is [MC]. The micromobility design team of the seamoby working group produced a problem statement also [SMM]. See the reference list for lots of previous work that has been done especially in the area of mobility with alternate proposals from Mobile IP. 2. Problem Statement This section proposes questions that have led to a perceived need for investigation of node routing within the IRTF. Why is there a perceived need for a non-tunnel-based routing solution for mobility? While Mobile IP provides transparency of the mobility to correspondent nodes communicating with hosts on the mobile network, Mobile IP does not provide transparency for any state contained within routers along the paths from correspondent nodes and the mobile node's home agent. That is, state in routers which is dependent on the stability of the source or destination address in the IP header will be negatively impacted by the mobile nodeÆs change of care of addresses. For example, when a mobile node changes its care of address, the mobile node's current reservation would require that the filter specs be updated with the new care of address. By instead treating the mobility as a topology change within the local area, these unpleasant second order effects can be avoided. Design Team 2 Internet Draft Local Subnet Mobility Problem Statement May 2001 2.1 Problem One There are applications that involve supporting real-time traffic flows to end devices that are mobile. These applications require that there be minimal (ideally no) interruption to the packet flows and that these flows need to preserve the existing QoS and security characteristics as the end station moves. In addition the movement of the end devices should not cause excessive signaling in the network as the devices move. Such applications often also run on networks in which link failures or router failures can cause a significant service disruption. Is it possible to use a similar protocol to allow for faster restoration of service during link failures. Could such mobility requirements and fast restoration requirements be met in a way that is more efficient and simpler than approaches based on Mobile IP and current intradomain routing protocols by using a routing protocol that implements local subnet mobility within a limited scope within the network? What would the limits of such a solution be in terms of scaling? Specifically how many nodes could be supported across what kind of network breadth and depth at what cost of complexity in the routers around the edge of such a network? Is it possible to meet traffic engineering requirements using such a protocol? Could it simplify management of QoS in the part of the network where mobility is most readily felt? 2.2 Problem Two One of the primary principles of the Internet has been end-to-end communication. Most (if not all) of the protocols running over IP have been designed with end-to-end signaling in mind. Currently, changes due to mobility induce unwanted end-to-end signaling. This may cause applications running over IP to fail due to delay and latency induced by unwanted signaling. Security, QoS and AAA signaling all suffer due to this. Could a local subnet mobility protocol aid in this respect? 2.3 Problem Three Why is there a need for an investigation of it in the IRTF rather than a direct move to work on a standard within a working group of the IETF? The area directors overseeing the activities of the seamoby working group and the mobile-ip working group have raised questions about the scale of local subnet mobility routing and the potential need to introduce both another routing protocol and another mobility protocol. A comparison with existing mobility management and routing protocols are involved in making such an assessment both in terms of relative scalability, performance, and complexity. 3. Acknowledgements The compiler would like to acknowledge all the contributors who have produced work relating to host routing for the various problems. . Design Team 3 Internet Draft Local Subnet Mobility Problem Statement May 2001 John Loughney, Vince Park, and Michael Thomas contributed valuable text for this problem statement. References [MC] M. Scott Corson, A. OÆNeill, G. Tsirtsis. IP Fast Restoration. Work in Progress. Draft-corson-fastrestore-00.txt, November 2000. [SMM] J. Loughney, et. al. ôSeaMoby Micromobility Problem Statement.ö Work in Progress. Draft-ietf-seamoby-mm-problem-01.txt, February 2001. [FHO] G. Tsirtsis, et. al. ôFast Handovers for Mobile Ipv6.ö Work in Progress. Draft-ietf-mobileip-fast-mipv6-00.txt, February 2001. [MT] M. Thomas. ôAnalysis of Mobile IP and RSVP Interactions.ö Work in Progress. Draft-thomas-seamoby-rsvp-analysis-00.txt, February 2001. [CI] Z. Shelby, et. al. ôCellular IP v6.ö Work in Progress. Draft-shelby-seamoby-cellularipv6-00.txt, November 2000. [HSR] A. OÆNeill. ôHost Specific Routing.ö Work in Progress. Draft-oneill-li-hst-00.txt, November 2000. [IMMP] A. Campbell and J. Gomez. ôIP Micro-mobility Protocols.ö ACM SIGMOBILE Mobile Computer and Communications Review (MC2R), 2001, available at http://www.comet.columbia.edu/micromobility. [CIP] Campbell, et. al. ôCellular IP.ö Work in Progress. Draft expired, but available at http://www.comet.columbia.edu/micromobility. [HAW] Ramjee, et. al. ôHawaii.ö Work in Progress. Draft expired, but available at http://www.comet.columbia.edu/micromobility. [EMA] M. Scott Corson, and Alan O'Neill. An Approach to Fixed/Mobile Converged Routing. University of Maryland, Institute for Systems Research, Technical Report, TR 2000-5. 2000. available at http://www.isr.umd.edu/TechReports/ISR/2000/TR_2000-5/TR_2000-5.phtml Addresses The compiler of this document is: Phil Roberts, Megisto Systems, Inc., proberts@megisto.com Design Team 4