AUTOCONF Working Group                                      C. Bernardos
Internet-Draft                                                      UC3M
Intended status: Informational                             R. in 't Velt
Expires: April 29, 2010                                              TNO
                                                        October 26, 2009


       Addressing Model for Router Interfaces in Ad Hoc Networks
              draft-bernardos-autoconf-addressing-model-01

Status of this Memo

   This Internet-Draft is submitted to IETF in full conformance with the
   provisions of BCP 78 and BCP 79.

   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.

   This Internet-Draft will expire on April 29, 2010.

Copyright Notice

   Copyright (c) 2009 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents in effect on the date of
   publication of this document (http://trustee.ietf.org/license-info).
   Please review these documents carefully, as they describe your rights
   and restrictions with respect to this document.

Abstract

   This document describes a practical IP addressing model for
   interfaces that take part in router-to-router communications in ad



Bernardos & in 't Velt   Expires April 29, 2010                 [Page 1]


Internet-Draft          autoconf addressing model           October 2009


   hoc networks.

Requirements Language

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in RFC 2119 [RFC2119].


Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . 3
   2.  Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 3
   3.  Addressing model  . . . . . . . . . . . . . . . . . . . . . . . 4
     3.1.  Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
     3.2.  IPv4/IPv6 practical addressing model  . . . . . . . . . . . 5
     3.3.  DAD considerations  . . . . . . . . . . . . . . . . . . . . 7
   4.  IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 7
   5.  Security Considerations . . . . . . . . . . . . . . . . . . . . 7
   6.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . 7
   7.  References  . . . . . . . . . . . . . . . . . . . . . . . . . . 8
     7.1.  Normative References  . . . . . . . . . . . . . . . . . . . 8
     7.2.  Informative References  . . . . . . . . . . . . . . . . . . 8
   Authors' Addresses  . . . . . . . . . . . . . . . . . . . . . . . . 9



























Bernardos & in 't Velt   Expires April 29, 2010                 [Page 2]


Internet-Draft          autoconf addressing model           October 2009


1.  Introduction

   In order to communicate among themselves, ad hoc routers [RFC2501]
   need to configure their network interface(s) with addresses that are
   valid within an ad hoc network.  Ad hoc routers may also need to
   configure globally routable addresses, in order to communicate with
   devices on the Internet.  From the IP layer perspective, an ad hoc
   network presents itself as a L3 multi-hop network formed over a
   collection of links.

   This document describes a practical addressing model for ad hoc
   networks.  It is required that a such model does not cause problems
   for ad hoc-unaware parts of the system, such as standard applications
   running on an ad hoc router or regular Internet nodes attached to the
   ad hoc routers.


2.  Terminology

   Readers are expected to be familiar with all the terms defined in the
   RFC 2501 [RFC2501].  In addition the document makes use of the
   following definitions:

   Wireless Link

      According to [I-D.iab-ip-model-evolution], a "link" in the IP
      service model refers to the topological area within which a packet
      with an IPv4 TTL or IPv6 Hop Limit of 1 can be delivered.  That
      is, where no IP-layer forwarding (which entails a TTL/Hop Limit
      decrement) occurs between two nodes.  A "wireless link" can be
      defined similarly, with the topological area in this case given by
      the radio-range coverage of the wireless technology used.  Due to
      the nature of the wireless medium, links are intermittent, and
      potentially short-lived.  Node movement exacerbates these
      characteristics.

   MANET interface

      Any interface over which a MANET protocol is run.

   MANET domain

      A MANET domain is delimited by a set of MANET routers that run a
      common MANET routing protocol and corresponds to its routing
      domain.

   Attached MANET (domain)




Bernardos & in 't Velt   Expires April 29, 2010                 [Page 3]


Internet-Draft          autoconf addressing model           October 2009


      A MANET domain attached to an infrastructure based network (e.g.,
      the Internet).  The MANET interfaces of routers of an attached
      MANET should be configured with unique global IP addresses, if
      these addresses are somehow exposed beyond the MANET domain.  By
      infrastructure network, we refer to any existing network which
      presents a certain hierarchical organisation (e.g., different
      subnets) and that is delegated a certain set of IP addresses/
      prefixes.

   Non-overlapping prefix

      Two IP prefixes p::/l_p and q::/l_q are non-overlapping if and
      only if there is no IP address p::a/l_p configured from p::/l_p
      that also belongs to q::/l_q, and the other way around.  For
      example, 2001:DB8:1:1::/64 and 2001:DB8:1:2::/64 are non-
      overlapping prefixes, while 2001:DB8:1::/48 and 2001:DB8:1:2::/64
      are not.


3.  Addressing model

   This section describes a practical IPv4/IPv6 addressing model for ad
   hoc networks.  We first define the scope of the addressing model,
   then propose how to practically configure IP on MANET interfaces.
   Finally, we provide some considerations on address uniqueness.

3.1.  Scope

   This document describes an addressing model for MANET interfaces.
   Regular (non-MANET) interfaces are not in the scope of the present
   document, as they are expected to be configured using standard
   mechanisms (such as SLAAC [RFC4862] or DHCP [RFC2131], [RFC3315]).
   Note, that MANET routers may need to acquire IP address prefixes to
   facilitate the configuration of IP addresses on nodes reachable via
   non-MANET interfaces.  How to do this is a topic that is also outside
   the scope of this document.

   This document does not place restrictions on the use of IP addresses
   configured on MANET interfaces.  We assume that these IP addresses
   are used by MANET routing protocols.  We also assume that, once MANET
   routing protocols have started to populate the Forwarding Information
   Bases (FIB) of routers with routing entries, these IP addresses will
   play a role in the forwarding of user data packets.  In particular,
   it is assumed that these addresses will be found as next-hop
   addresses in the routing tables of MANET routers.  The forwarding of
   user data in many cases includes the resolution of the link-layer
   address of the interface to which the next-hop IP address is bound.
   Furthermore, it cannot be ruled out that the IP addresses configured



Bernardos & in 't Velt   Expires April 29, 2010                 [Page 4]


Internet-Draft          autoconf addressing model           October 2009


   on MANET interfaces will be used as source or destination addresses
   by end-user applications in cases where such applications reside on
   MANET routers.  An architecture in which applications are separated
   by one hop from MANET interfaces is conceptually elegant, but may not
   always be practical.

   This document considers MANET domains for the purposes of IP
   configuration.  Therefore, when we use the term "MANET" throughout
   this document, we are referring to a MANET domain.  For example,
   MANET local uniqueness refer to uniqueness within the MANET domain.

   Globally unique IP addresses MUST be provided for routers of attached
   MANETs for those cases where these addresses are visible outside the
   MANET domain, while only uniqueness within the MANET domain is
   required for non-attached MANETs.

   This document does not rule out that IP addresses might be configured
   by non-autoconf mechanisms (e.g., manually) on MANET interfaces.

3.2.  IPv4/IPv6 practical addressing model

   This section describes the basic principles for IP addressing for
   MANET interfaces, in as much an IP version agnostic manner as
   possible.

   MANET interfaces of attached MANETs SHOULD be configured with global
   IPv6 addresses if these addresses are somehow exposed outside the
   MANET domain.  For non-attached MANETs, ULAs or global addresses
   SHOULD be used.

   Since the topology of a mobile ad hoc network is expected to be
   frequently changing, MANET interfaces MUST be configured with unique/
   non-overlapping prefixes.  This principle does not assume any prefix
   length.  The use of /32 (in the IPv4 case) or /128 (in the IPv6 case)
   prefix lengths can be an effective way to ensure that prefixes are
   non-overlapping.  However, it would be needlessly restrictive to
   mandate the use of only these prefix lengths.  Due to its larger
   address space, it is much easier to generate addresses for IPv6 that
   are unique than is the case for IPv4.  This is equally true for
   prefixes with non-maximum lengths.

   MANET interfaces MUST also be configured with IPv6 Link-local
   addresses (as required by RFC 4861 [RFC4861] and RFC 4291 [RFC4291]).
   Two main concerns may arise when considering the use of IPv6 Link-
   local addresses:

   o  Address uniqueness: the event of having two duplicate addresses in
      the same link has proved to be very low (EUI64 derived interface



Bernardos & in 't Velt   Expires April 29, 2010                 [Page 5]


Internet-Draft          autoconf addressing model           October 2009


      identifiers very rarely collide, since MAC addresses are expected
      to be globally unique), and even some mechanisms have been
      proposed to reduce the collision probability [paper.DAD].
      Therefore, in most scenarios it is safe to assume that the
      probability of having two or more duplicated link-local addresses
      in a MANET is negiglible.  For those scenarios, in which this
      cannot be safely assumed, we refer to the DAD considerations of
      Section 3.3.

   o  Reachability: connectivity among neighbours in wireless links may
      be intermittent and/or short-lived.  Therefore, the use of link-
      local addresses may lead to reachability issues, since two nodes
      that were in direct coverage range at one moment, might not be
      anymore shortly after.  These problems might also arise in wired
      networks (nodes going up/down), but it is not the common case.

   Designers of MANET routing protocols (and other protocols) should be
   aware of these concerns and assess their impact, in order to make an
   informed decision whether to make use of link-local addresses or not.

   Fluctuating reachability as discussed above is also of concern to the
   data forwarding process in ad hoc networks.  This is especially true,
   if existing mechanisms for neighbour discovery and address resolution
   are to be applied.  In order to mitigate these problems, several
   solutions may be used, such as (but not limited to): decrease some of
   the ND default timer values (specified in RFC 4861 [RFC4861]), such
   as REACHABLE_TIME, RETRANS_TIMER, DELAY_FIRST_PROBE_TIME,
   MIN_RANDOM_FACTOR, MAX_RANDOM_FACTOR; implement a stronger
   interaction between the MANET routing protocols and the ND process,
   so the MANET routing protocol helps to keep updated the ND tables.
   Finally, if none of these solutions (or alternative ones) may be
   implemented, processes running on the MANET routers that need to be
   isolated from this problem can decide not to use link-local addresses
   for their local communications.  Since IPv4 lacks any standardised
   unreachability detection mechanism, these considerations about
   reachability only concern IPv6.

   Configuration and use of IPv4 link-locals on MANET interfaces are not
   forbidden.  However, while in IPv6, an interface may be
   simultaneously configured with a link-local address and with unicast
   (global or local) addresses, this is not recommended in IPv4
   [RFC3927].

   When forwarding user data packets from one MANET router to the next,
   along the path from source to destination, standard mechanisms for
   layer-2 address resolution of next-hop IP addresses, such as ND or
   ARP, may be used.  In this context, it should be noted that the
   presence of IPv6 link-local addresses on MANET interfaces may lead to



Bernardos & in 't Velt   Expires April 29, 2010                 [Page 6]


Internet-Draft          autoconf addressing model           October 2009


   their use, e.g. as source address in a neigbor solicitation.  As
   mentioned before, the exchange of MANET routing protocols packets is
   a potential alternative source of link-layer address information.

3.3.  DAD considerations

   This document assumes that DAD is disabled by default for the IP
   addresses configured on MANET interfaces (this is allowed in RFC 4862
   [RFC4862].  For the case of link-local addresses, we assume the
   collision probability is negiglible, and that it therefore is safe to
   avoid the overhead of an active DAD process (which would need to be
   modified to be run in a MANET domain wide fashion).  For the case of
   the non-overlapping prefixes, we do not specify how their uniqueness
   is ensured (this is out-of-scope of this document and falls in the
   solution space).

   However, this document does not forbid the use of any DAD mechanism,
   if it is required in some certain scenarios.  From the point of view
   of MANETs, it seems appropriate to consider as well the use of
   passive DAD approaches (such as [paper.PACMAN],
   [paper.PACMAN_assessment]).


4.  IANA Considerations

   This document makes no request of IANA.


5.  Security Considerations

   This document does currently not describe any security
   considerations.


6.  Acknowledgements

   Some of the ideas included in this draft have been proposed in the
   AUTOCONF ML by several people.  Thanks for all the AUTOCONF WG
   participants for the fruitful discussions over these years.

   The authors would like to thank Thomas Clausen and Teco Boot for
   their comments and discussion on this document.

   The research of Carlos J. Bernardos leading to these results has
   received funding from the European Community's Seventh Framework
   Programme (FP7/2007-2013) under grant agreement n. 214994 (CARMEN
   project) and also from the Ministry of Science and Innovation of
   Spain, under the QUARTET project (TIN2009-13992-C02-01).



Bernardos & in 't Velt   Expires April 29, 2010                 [Page 7]


Internet-Draft          autoconf addressing model           October 2009


7.  References

7.1.  Normative References

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC2131]  Droms, R., "Dynamic Host Configuration Protocol",
              RFC 2131, March 1997.

   [RFC3315]  Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C.,
              and M. Carney, "Dynamic Host Configuration Protocol for
              IPv6 (DHCPv6)", RFC 3315, July 2003.

   [RFC3927]  Cheshire, S., Aboba, B., and E. Guttman, "Dynamic
              Configuration of IPv4 Link-Local Addresses", RFC 3927,
              May 2005.

   [RFC4291]  Hinden, R. and S. Deering, "IP Version 6 Addressing
              Architecture", RFC 4291, February 2006.

   [RFC4861]  Narten, T., Nordmark, E., Simpson, W., and H. Soliman,
              "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861,
              September 2007.

   [RFC4862]  Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless
              Address Autoconfiguration", RFC 4862, September 2007.

7.2.  Informative References

   [I-D.iab-ip-model-evolution]
              Thaler, D., "Evolution of the IP Model",
              draft-iab-ip-model-evolution-01 (work in progress),
              November 2008.

   [RFC2501]  Corson, M. and J. Macker, "Mobile Ad hoc Networking
              (MANET): Routing Protocol Performance Issues and
              Evaluation Considerations", RFC 2501, January 1999.

   [paper.DAD]
              Bagnulo, M., Soto, I., Garcia-Martinez, A., and A.
              Azcorra, "Avoiding DAD for Improving Real-Time
              Communication in MIPv6 Environments", Joint International
              Workshop on Interactive Distributed Multimedia Systems/
              Protocols for Multimedia Systems IDMS-PROMS 2002, Coimbra
              (Portugal). Lecture Notes in Computer Science 2515, pps
              73-79, Ed. Springer-Verlag, 2002. , November 2002.




Bernardos & in 't Velt   Expires April 29, 2010                 [Page 8]


Internet-Draft          autoconf addressing model           October 2009


   [paper.PACMAN]
              Weniger, K., "PACMAN: passive autoconfiguration for mobile
              ad hoc networks",  IEEE Journal on Selected Areas in
              Communications 23 (3) , 2005.

   [paper.PACMAN_assessment]
              Bernardos, C., Calderon, M., Soto, I., Solana, A., and K.
              Weniger, "Building an IP-based Community Wireless Mesh
              Network: Assessment of PACMAN as an IP Address
              Autoconfiguration Protocol", Computer Networks, accepted
              for publication , 2009.


Authors' Addresses

   Carlos J. Bernardos
   Universidad Carlos III de Madrid
   Av. Universidad, 30
   Leganes, Madrid  28911
   Spain

   Phone: +34 91624 6236
   Email: cjbc@it.uc3m.es
   URI:   http://www.it.uc3m.es/cjbc/


   Ronald in 't Velt
   TNO Information and Communication Technology
   Brassersplein 2
   Delft  2600 GB
   The Netherlands

   Phone: +31 15 2857306
   Email: Ronald.intVelt@tno.nl

















Bernardos & in 't Velt   Expires April 29, 2010                 [Page 9]