Internet Draft                                             C. Janneteau
Document: draft-janneteau-nemo-multicast-mldproxy-00.txt        E. Riou
Expires: October 2004                                       A. Petrescu
                                                           A. Olivereau
                                                             H.-Y. Lach
                                                          Motorola Labs
                                                             April 2004


          IPv6 Multicast for Mobile Networks with MLD-Proxy
           <draft-janneteau-nemo-multicast-mldproxy-00.txt>


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 addresses the problem of delivering IPv6 multicast
   traffic to nodes inside a mobile network. An approach based on the
   deployment of "MLD-based Multicast Forwarding (MLD-proxying)" [1]
   in the mobile network, combined with the use of MLD [2][3] between
   the mobile router (MR) and the visited network, is proposed. Such a
   configuration allows multicast support for mobile networks, without
   the need to run a multicast routing protocol. In addition of being
   simple to implement and deploy, this approach provides global
   mobility in the Internet, and optimal multicast routing with nested
   mobile networks, while optimizing network and radio resources. This
   document does not specify new protocols, nor extensions to existing
   protocols.










Janneteau et al.          Expires October 2004                 [Page i]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

Table of Contents

   Status of this Memo................................................i
   Abstract...........................................................i
   Conventions Used in this Document..................................1
   1. Introduction....................................................1
   2. Multicast Configuration in the Mobile Network...................3
   3. Example.........................................................4
   4. Discussion......................................................6
   4.1 Simplicity.....................................................6
   4.2 Global Mobility in the Internet................................6
   4.3 Optimal Routing with Nested Mobile Networks....................7
   4.4 Optimization of Network and Radio Resources....................7
   4.5 Seamless Mobility..............................................8
   4.6 Fault Tolerance................................................8
   4.7 Operation in Disconnected Mode................................10
   4.8 Independence from the NEMO Basic Support Protocol.............10
   4.9 Support of Multicast Sources inside the Mobile Network........10
   5. Security Considerations........................................11
   6. Conclusion.....................................................11
   Acknowledgements..................................................11
   Copyright Notice..................................................13


Conventions used in this document

   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 [4].

   Some terms from the NEMO terminology [5] are used in this document;
   as well as the following definitions from [1]:

   Upstream Interface: A proxy device's interface in the direction of
                       the root of the tree. Also called the "Host
                       interface".

   Downstream Interface: Each of a proxy device's interfaces that is
                         not in the direction of the root of the
                         tree. Also called the "Router interfaces".

1. Introduction

   The NEMO Basic Support protocol [6] specifies extensions to Mobile
   IPv6 [7] in order to provide reachability at a permanent address
   and unicast session continuity for nodes in a mobile network, while
   the mobile router (MR) moves between IPv6 subnets.  The protocol
   relies on the establishment of a bi-directional tunnel between the
   mobile router (MR) and its home agent (HA).  MR forwards outgoing
   IPv6 packets through the tunnel, towards its home network.
   Similarly, any packet arriving on the MR's home link are
   intercepted by the HA and forwarded through the tunnel, if the
   packet's destination address matches the mobile network prefix



Janneteau et al.          Expires October 2004                 [Page 1]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

   (MNP).  This is achieved by extending the MR's HA to support
   redirection for the whole MNP, in addition to the MR's home
   address.

   While mobility support for IPv6 unicast sessions is fully addressed
   by the above mentioned mechanism, [6] (up to now) does not discuss
   the case of IPv6 multicast traffic.  The issue here is to enable
   session continuity both for multicast receivers and multicast
   sources located inside a mobile network, while MR is moving.

   Since multicast routing highly differs from unicast routing, the HA
   forwarding mechanism defined by [6] for unicast packets does not
   serve the forwarding of multicast packets.  The key reason is that
   multicast packets are sent to a multicast address that is unrelated
   to the prefixes of the networks where potential multicast receivers
   are located (e.g. the mobile network prefix).

   A possible solution is the use of the MR-HA bi-directional tunnel
   for the forwarding of multicast traffic too, in both directions,
   between the MR and its home link.  By co-locating a multicast
   routing function on both the MR and its HA, multicast routing
   protocol messages can be exchanged over the MR-HA tunnel.  This
   allows the creation of multicast branches from the Internet to
   receivers in the mobile network, as well as from sources in the
   mobile network to receivers in the Internet.  This solution is
   somehow similar to running a dynamic unicast routing protocol over
   the MR-HA tunnel as described in section 8 of [6].

   This document addresses the problem of delivering IPv6 multicast
   traffic to nodes inside mobile networks.  It proposes an attractive
   alternative, in situations where the above mentioned approach is
   either not possible (e.g. due to MR or HA capabilities), or the
   network and radio overhead (i.e. sub-optimal routing, and
   encapsulating header) introduced by the MR-HA tunnel is not
   acceptable.

   Up to now, two main approaches have been identified by the IETF
   community enabling delivery of multicast traffic to mobile
   multicast hosts [8][9].  The first one, sometimes called
   Bi-directional Tunnelling (BT), relies on the Mobile IPv6 tunnel to
   forward the multicast traffic to the mobile host.  This is the same
   principle as what has been discussed above.  The second approach,
   sometimes called Remote Subscription (RS), proposes that the mobile
   host joins the multicast group via a local multicast router on the
   visited link.  For this purpose, the mobile host behaves like any
   other fixed multicast receiver in the visited network, e.g. sending
   MLD Report messages [2][3] to the local multicast router using an
   IPv6 link-local source address.

   Our approach proposes to apply the Remote Subscription principle to
   the case of a mobile router, serving a mobile network.  The MR uses
   MLD messages to re-subscribe to the ongoing multicast sessions
   through the local multicast router in the visited network.  In the
   case of the mobile network however, the two following requirements
   have to be considered:

Janneteau et al.          Expires October 2004                 [Page 2]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

     o First, MR must be able to discover which multicast groups nodes
       in the mobile network are interested in, in order to subscribe
       to them and forward the traffic to all interested receivers
       inside the mobile network.

     o Second, the multicast routing inside the mobile network should
       be optimal.  That is, routing multicast packets only in parts
       of the mobile network where receivers are located; and thus
       avoiding flooding.

   Our approach bases on the deployment of "MLD-based Multicast
   Forwarding (MLD-proxying)" [1] inside the mobile network, in order
   to solve the above mentioned issues.  Such a configuration,
   combined with the use of Remote Subscription at MR, allows
   multicast support for mobile networks, without the need to run a
   multicast routing protocol.

   In addition of being simple to implement and deploy, this approach
   provides global mobility in the Internet and optimal multicast
   routing with nested mobile networks.  In addition, it optimizes the
   usage of network and radio resources.

   In the following sections, we will first review the details of the
   multicast configuration inside the mobile network, as proposed in
   our approach.  Then, we will present some of the advantages
   foreseen.

2. Multicast Configuration in the Mobile Network

   The multicast configuration in the mobile network bases on the
   deployment of the "MLD-based Multicast Forwarding" function (or
   MLD-proxy) both on internal fixed routers and on the MR, in such a
   way that group membership information will propagate router-by-
   router (being aggregated at each router) from the routers serving
   multicast receivers towards the MR.

   This approach allows to route multicast traffic inside the mobile
   network without the need to run a multicast routing protocol. It is
   sufficient for each intermediate router to learn and proxy group
   membership information and simply forward multicast traffic based
   upon that information.

   As opposed to multicast routing protocols, the MLD-based Multicast
   Forwarding mechanism does not define a tree buidling algorithm, as
   a way to avoid forwarding loops.  The consequence is that the MLD-
   based Multicast Forwarding approach can be used only in tree
   topologies.  Hence, routers inside the mobile network must be
   manually configured to form a tree rooted at the MR, for the
   purpose of multicast forwarding within the mobile network.  This
   configuration process consist in (1) selecting which routers should
   run the MLD-proxy function, and (2) designating the unique upstream
   interface and the downstream interface(s) on each of these proxy
   devices.  A proxy device performs the Host part of the MLD protocol
   on the upstream interface, and the Router part of MLD on each of


Janneteau et al.          Expires October 2004                 [Page 3]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

   its downstream interfaces.  Especially, ingress interfaces of the
   MR should be configured as downstream interfaces, while the egress
   interface must be configured as the upstream interface.

   The reader should refer to [1] for more details on the behaviors of
   MLD-proxy devices.

3. Example

   This section illustrates, through an example, the multicast
   operations in a mobile network supporting the above mentioned
   approach.

   Let's assume the mobile network of Figure 1, with a quite complex
   internal topology.  The mobile network is made of 5 different IPv6
   subnets (link1 to link5), interconnected by fixed routers FR1 to
   FR4.  MR is attached to Foreign Link A, served by the access router
   AR1.  Both AR1 and AR2 are multicast routers.  No assumption is
   made whether AR1 and AR2 run the same multicast routing protocol or
   not.  It may be the case if they are located in the same multicast
   domain.  If they are located in different domains, it is assumed
   that an appropriate inter-domain multicast routing protocol is
   deployed in the multicast inter-network.

                  CN
                   |
          "===================================="
          "                                    "
          "      Multicast Inter-network       "
          "                                    "
          "===================================="
               |                 |
              AR1               AR2
   Foreign     |                 |        Foreign
   Link A   -----------       ----------- Link B
                    |i1
                   MR                                 ---
                    |i2                                |
          ------------------------------------- link1  |
               |i1              |i1       |i1          |
              FR1              FR2       FR3           | Mobile
            i2|  |i3            |i2       |i2          | Network
    link2 -----  ----- link3   -------------- link4    |
                   |                 |i1               |
                   `--------------- FR4                |
                                 i2  |i3               |
                                   ------------- link5 |
                                            |          |
                                           MNN         |
                                                      ---

       Figure 1: Mobile Network with Complex Internal Topology




Janneteau et al.          Expires October 2004                 [Page 4]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

   In order to support delivery of multicast traffic to MNN with the
   MLD-based Multicast Forwarding scheme, the tree of MLD-proxies must
   be configured in the mobile network.  The configuration of Table 1
   can be used.

       ==============================================================
       | MLD-proxy?  | Upstream Interface | Downstream Interface(s) |
       |             |  (MLD Host Part)   |   (MLD Router Part)     |
   ==================================================================
   | MR |    YES     |         i1         |           i2            |
   |FR1 |    YES     |         i1         |         i2, i3          |
   |FR2 |    YES     |         i1         |           i2            |
   |FR3 |    NO      |        none        |          none           |
   |FR4 |    YES     |         i1         |           i3            |
   ==================================================================
                   Table 1: Example #1 of MLD-based
                  Multicast Forwarding Configuration

   In this configuration, it is worth noting that FR3 is operating the
   MLD protocol (either Host or Router part) on none of its
   interfaces.  Similarly, FR4 does not operate the MLD protocol on
   its interface i2.

   From the multicast routing perspective, the topology in the mobile
   network is then a tree, as depicted in Figure 2.  Of course, this
   has no impact on the unicast routing inside the mobile network.
   For instance, unicast packets may still be routed through FR3.

                  CN
                   |
          "===================================="
          "                                    "
          "      Multicast Inter-network       "
          "                                    "
          "===================================="
               |                 |
              AR1               AR2
   Foreign     |                 |        Foreign
   Link A   -----------       ----------- Link B
                    |i1
                   MR                                 ---
                    |i2                                |
          ------------------------------------- link1  |
               |i1              |i1                    |
              FR1              FR2                     | Mobile
            i2|  |i3            |i2                    | Network
    link2 -----  ----- link3   -------------- link4    |
                                     |i1               |
                                    FR4                |
                                     |i3               |
                                   ------------- link5 |
                                            |          |
                                           MNN         |
                                                      ---
   Figure 2: Tree Multicast Routing Topology in the Mobile Network

Janneteau et al.          Expires October 2004                 [Page 5]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

   Once the above configuration is in place, multicast traffic can be
   delivered to receivers in the mobile network.  Let's consider CN as
   the source of a multicast group G, served by a multicast tree in
   the multicast inter-network.  Let's consider that none of the nodes
   in the mobile network have yet subscribed to group G.

   When MNN subscribes to group G, it sends an MLD Report message on
   link5.  FR4, which is the proxy device for this link, intercepts
   this message, determines that this group is not yet received at
   FR4, and thus sends its own MLD Report message for group G through
   its upstream interface i1.  The membership subscription for group G
   propagates router-by-router (FR4 -> FR2 -> MR) within the mobile
   network up to MR.  Similarly, MR determines that this group is not
   yet received, and thus sends its own MLD Report message for group G
   through its upstream/egress interface towards its access router
   (AR1) on the Foreign Link A.  This will trigger establishment of a
   new delivery branch for multicast group G towards AR1 along which
   multicast packets will be forwarded.  Upon reception of the
   multicast packets, MR will look at the group membership information
   associated with each of its downstream interfaces and forward the
   packets through the interfaces for which the group G is mentioned.
   Other routers in the mobile network will apply the same forwarding
   mechanism. Finally, the packet will be received by MNN.

   When MR moves to a new network, for instance to Foreign Link B, it
   must re-subscribe to all multicast groups it was receiving through
   its old access router (AR1).  This is the Remote Subscription.  For
   this purpose MR sends new MLD Report messages for all the groups
   (and related source lists) listed in its aggregated group list
   towards the Foreign Link B.  Upon reception of these messages, AR2
   will trigger establishment of new multicast branches so that
   multicast packets for these groups can be delivered to MR.  Then MR
   forwards these packets inside the mobile network as before.  Note
   that the mobility of the mobile network is transparent to the
   multicast receivers in the mobile network since MR handles the re-
   subscription on its own.

4. Discussion

4.1 Simplicity

   The approach proposed is quite simple to implement on any type of
   system thanks to the widespread availability of MLD.

   It is quite simple to deploy and operate too, because no multicast
   routing protocol is needed inside the mobile network.  In addition,
   the required configuration to form the tree of MLD-proxies seems
   affordable in the context of mobile networks, which are typically
   small-to-medium leaf networks.

4.2 Global Mobility in the Internet

   The approach bases on the use of the MLD protocol between the MR
   and the visited network, in order to maintain multicast session
   continuity.  This allows global mobility in the multicast inter-

Janneteau et al.          Expires October 2004                 [Page 6]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

   network since MLD is expected to be supported in any IPv6 subnet.
   With this approach the mobile network can roam in any visited
   network, irrespective of the type of multicast routing protocol
   locally deployed.

   If the mobile network moves to a foreign link where multicast is
   not supported, MR can still maintain ongoing multicast sessions by
   performing re-subscription to the group, with MLD Report messages,
   through the MR-HA bi-directional tunnel.  Several approaches can be
   envisaged for the multicast support on the HA.  It can be
   configured as a multicast router, or simply as an MLD-proxy.  In
   the later case, the HA would then issue its own MLD Reports towards
   the multicast-enabled Border Router (BR) on the MR's home link.

4.3 Optimal Routing with Nested Mobile Networks

   The approach provides optimal routing both in the infrastructure
   and inside the mobile network.

   Optimal routing is achieved in the infrastructure thanks to the
   Remote Subscription concept, by reconstructing a new multicast
   branch at the current location of MR.

   Optimal routing inside the mobile network is achieved by the "MLD-
   based Multicast Forwarding" mechanism which routes packets on the
   shortest-possible-path (according to the tree topology configured),
   and only in subnets where receivers are located.

   It is worth noting that optimality of the path is preserved even in
   scenarios of nested mobile networks.

   It is acknowledged that under certain circumstances the remote
   subscription process may involve overload in the network, for
   instance when the MR moves frequently into foreign networks which
   are topologically far from the existing multicast tree.  In the
   context of this document, however, it is considered that this
   inconvenient is largely overcome by the advantages of offering
   optimal routing paths outside the mobile network (no mandatory long
   path to Home Agent) and inside an aggregation of mobile networks.

4.4 Optimization of Network and Radio Resources

   The approach preserves native multicast routing of the traffic all
   along the path from the source to the receivers, even when located
   under several nested mobile routers.  Flooding is avoided both in
   the infrastucture thanks to the routing protocol (e.g. PIM-SM), and
   inside the mobile network thanks to MLD-based Multicast Forwarding,
   which optimizes network resources.

   In addition, radio resources are also optimized since the sending
   of native IPv6 multicast (i.e. not unicast-encapsulated) over the
   air interface, between the infrastructure and MR or even within the
   mobile network, avoids header overhead and allows to take full
   advantage of a potential multicast support available at the link
   level.

Janneteau et al.          Expires October 2004                 [Page 7]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

4.5 Seamless Mobility

   The fact that MR aggregates group membership information (from
   nodes inside the mobile network) and subscribes to those groups
   using MLD, makes the MR behave like an individual mobile host with
   respect to the multicast routing in the infrastructure.  One
   advantage is that seamless multicast handover mechanisms (such as
   [10] and [11]) that have been defined for mobile hosts still work
   for mobile networks that implement MLD-proxying locally.

   For instance, in [10], seamless mobile multicast is achieved thanks
   to the forwarding of multicast traffic through a temporary tunnel
   established between the previous and the new multicast access
   routers, while the multicast branch at the new access router is
   under construction.

4.6 Fault Tolerance

   Although the multicast delivery mechanism inside the mobile
   network, based on "MLD-based Multicast Forwarding", requires an
   appropriate designation of upstream and downstream interfaces on
   internal routers in order to form a multicast tree, it features
   some form of robustness against failure.

   Indeed, this approach allows several routers on the same link to be
   configured as MLD-proxy (and thus as MLD Router).  In order to
   avoid redundant traffic on the link, coming from those different
   proxies, only one of them must be authorized to propagate the group
   membership information upstream and to forward the multicast
   traffic on the link.  This is achieved by electing a single
   forwarding MLD-proxy per link.  The solution proposed in [1] is to
   "piggy-back" this forwarder proxy election on the MLD Querier
   election.  In case of failure of the forwarder/querier router,
   another one is automatically elected.  This new one will then start
   to propagate upstream the group membership information it had
   previously collected on the link, and delivered the related
   multicast traffic on this link.  If later the former router
   recovers, it will automatically win the Querier election (smaller
   IP address), and thus take the role of forwarder on the link.

   Considering the mobile network of Figure 1, another possible
   multicast configuration is given in Table 2.  Parentheses "()"
   indicate the changes with respect to the configuration of Table 1.

       ==============================================================
       | MLD-proxy?  | Upstream Interface | Downstream Interface(s) |
       |             |  (MLD Host Part)   |   (MLD Router Part)     |
   ==================================================================
   | MR |    YES     |         i1         |           i2            |
   |FR1 |    YES     |         i1         |         i2, i3          |
   |FR2 |    YES     |         i1         |           i2            |
   |FR3 |   (YES)    |        (i1)        |          (i2)           |
   |FR4 |    YES     |         i1         |        (i2), i3         |
   ==================================================================
  Table 2: Example #2 of MLD-based Multicast Forwarding Configuration

Janneteau et al.          Expires October 2004                 [Page 8]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

   According to [2] and [3], if the IPv6 adress assigned to FR1's i3
   interface is smaller than the one of FR4's i2 interface, then FR1
   will be the Forwarder/Querier for link3.  Similarly, if the IPv6
   address assigned to FR2's i2 interface is smaller than the one of
   FR3's i2 interface, FR2 will be the Forwarder/Querier for link4.
   The multicast tree is then equivalent to the one depicted in Figure
   2. Now, if FR1 fails for any reason, FR4 will automatically becomes
   the Forwarder/Querier for link3.  Similarly, if FR2 fails, FR3 will
   become the Forwarder/Querier for link4.  This new multicast tree is
   depicted in Figure 3.  Ongoing multicast sessions are maintained.

                  CN
                   |
          "===================================="
          "                                    "
          "      Multicast Inter-network       "
          "                                    "
          "===================================="
               |                 |
              AR1               AR2
   Foreign     |                 |        Foreign
   Link A   -----------       ----------- Link B
                    |i1
                   MR                                 ---
                    |i2                                |
          ------------------------------------- link1  |
               |i1                        |i1          |
              FR1                        FR3           | Mobile
            i2|                           |i2          | Network
    link2 -----  ----- link3   -------------- link4    |
                   |                 |i1               |
                   `--------------- FR4                |
                                 i2  |i3               |
                                   ------------- link5 |
                                            |          |
                                           MNN         |
                                                      ---

        Figure 3: Another Multicast Tree in the Mobile Network

   When configuring several MLD-proxy devices on the same link, it
   should be ensured that the overall multicast topology inside the
   mobile network will always form a tree rooted at MR (i.e. no loop),
   irrespective of the outcome of the forwarder election process.  If
   this rule is met, then an appropriate support for failure recovery
   can be provided.

   It is worth mentioning that, from the multicast delivery
   perspective, this forwarder election mechanism also allows to
   duplicate the mobile router serving the mobile network, and to
   automatically switch to the backup mobile router only when the
   master one fails.




Janneteau et al.          Expires October 2004                 [Page 9]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

4.7 Operation in Disconnected Mode

   Another capability that may be expected from a mobile network, is
   to allow establishment of new communications between nodes inside
   the mobile network, and maintain existing ones, in situations where
   MR is disconnected from the fixed infrastructure.  This is referred
   to as operating in disconnected (or autonomous) mode.

   The MLD-proxy approach for multicast routing inside the mobile
   network supports the disconnected mode, since the multicast packets
   from a local source can be delivered to local receivers (fixed, and
   mobile ones using remote subscription) without the need for the
   packets to be routed outside of the mobile network.  Indeed, [1]
   recommends that each MLD-proxy device, receiving a multicast packet
   on a downstream interface which is in the Forwarder/Querier mode,
   forwards this packet through its upstream interface in addition to
   through all its other downstream interfaces (in Forwarder/Querier
   mode) which have a subscription pertaining to this packet.  This
   rule allows multicast packets from a local source to propagate
   hop-by-hop to the MR, and be redistributed at each hop in the
   downstream direction, so that all local receivers can be reached.

   Again, it is worth noting that the proposed approach supports
   multicast communications in disconnected mode, even in cases of
   nested mobile networks.  Indeed, if the top-level Mobile Router of
   an aggregation of nested mobile networks is disconnected from the
   infrastructure, multicast communications are still possible between
   any sources and receivers inside this aggregation, irrespective of
   their location.  In addition, multicast routing will again be
   optimal.

4.8 Independence from the NEMO Basic Support Protocol

   This proposed mobile multicast scheme for mobile networks is
   completely independent of the protocol used to maintain ongoing
   unicast communications, be it the NEMO Basic Support [12] or
   another.  This independence allows flexibility.

4.9 Support of Multicast Sources inside the Mobile Network

  The MLD-proxy forwarding rule discussed in section 4.7 allows
  delivery of multicast traffic from a local source to all local
  receivers in a mobile network.

  In order to reach receivers outside of the mobile network, the MR
  intercepting the multicast packet from the local source should then
  forward this packet towards the multicast delivery tree in the
  multicast inter-network.  Because of the loop-avoidance mechanism of
  multicast routing protocols (i.e. RPF check), multicast packets must
  enter the multicast delivery tree through a point wherefrom the
  source is considered topologically correct and subsequent RPF checks
  of multicast routers in the delivery tree will complete.




Janneteau et al.          Expires October 2004                [Page 10]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

  A possible solution is that MR forwards outgoing multicast packets
  towards its home network, for instance by taking advantage of the
  MR-HA tunnel of the NEMO Basic Support protocol.

5. Security Considerations

   The proposed solution for delivery of multicast to mobile networks
   relies on the use of MLD-based Multicast Forwarding inside the
   mobile network, combined with the use of MLD at the MR.

   Security aspects of MLD and MLD-based multicast forwarding are
   discussed in [2][3], and [1] respectivelly.

6. Conclusion

   This document has discussed the support of IPv6 multicast for
   mobile networks.  The approach presented is also applicable in the
   IPv4 context, where IGMP is used instead of MLD.

   This approach, including MLD-based Multicast Forwarding inside the
   mobile network and remote subscription at MR, has been demonstrated
   during various public events including the HyWiN workshop [12], on
   December 2003.  The implementation was based on the LIVSIX [13]
   open source IPv6 stack.

Acknowledgements

   The authors would like to thank their colleagues in the OverDRiVE
   project, in the framework of which this work has been partly
   performed.  Thanks too to Daniel Negru and Greg Daley, for the
   discussion held on the NEMO mailing list that has encouraged the
   publication of this document.
























Janneteau et al.          Expires October 2004                [Page 11]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

References

   [1] Fenner, B., He, H., Haberman, B. and Sandick, H., "IGMP/MLD-
       based Multicast Forwarding ("IGMP/MLD Proxying")", work in
       progress, draft-ietf-magma-igmp-proxy-06.txt, April 2004.

   [2] Deering, S., Fenner, W. and Haberman, B., "Multicast Listener
       Discovery (MLD) for IPv6", RFC 2710, October 1999.

   [3] Vida, R. and Costa, L., "Multicast Listener Discovery Version 2
      (MLDv2) for IPv6", work in progress, draft-vida-mld-v2-08.txt,
      December 2003.

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

   [5] Ernst, T. and Lach, H.-Y., "Network Mobility Support
       Terminology", draft-ietf-nemo-terminology-01.txt, work in
       progess, February 2004.

   [6] Devarapalli, V., Wakikawa, R., Petrescu, A. and Thubert, P.,
       "Network Mobility (NEMO) Basic Support Protocol", work in
       progress, draft-ietf-nemo-basic-support-02.txt, December 2003.

   [7] Johnson, D., Perkins, C. and Arkko, J, "Mobility Support in
       IPv6", work in progress, draft-ietf-mobileip-ipv6-24.txt, June
       2003.

   [8] Daley, G., Kurup, G., "Requirements for Mobile Multicast
       Clients", work in progress, draft-daley-magma-mobile-00.txt,
       June 2003, (Expired).

   [9] Janneteau, C., Tian, Y., Csaba, S., Lohmar, T., Lach, Y.-H.,
       Tafazolli, R., "Comparison of Three Approaches Towards Mobile
       Multicast", IST Mobile Summit 2003, Aveiro, Purtugal, 15-18
       June 2003, http://www.mobilesummit2003.org.

   [10] Simon, C., Vida, R., Kersch, P., Janneteau, C., Leijonhufvud,
        G., "Seamless IP Multicast Handovers in OverDRiVE", IST Mobile
        Summit 2004, Lyon, France, 27-30 June 2004, (accepted for
        publication), http://www.mobilesummit2004.org.

   [11] Suh, K., Kwon, D.-H., Suh, Y.-J. and Park, Y., "Fast Multicast
        Protocol for Mobile IPv6 in the fast handovers environments",
        work in progress, draft-suh-mipshop-fmcast-mip6-00.txt,
        January 2004.

   [12] Janneteau, C., Demonstration entitled "Multicast for Moving
        Networks", HyWiN 2003 workshop, Turin, Italy, 2 December 2003,
        http://www.ist-overdrive.org/HyWiN2003/.

   [13] LIVSIX IPv6 Stack, http://www.enrl.motlabs.com/livsix




Janneteau et al.          Expires October 2004                [Page 12]


INTERNET-DRAFT        Multicast for Mobile Networks          April 2004

Authors' Addresses

  Christophe Janneteau               Emmanuel Riou
  Motorola Labs                      Motorola Labs
  Parc les Algorithmes St Aubin      Parc les Algorithmes St Aubin
  Gif-sur-Yvette 91193               Gif-sur-Yvette 91193
  France                             France
  Phone:  +33 1 69352548             Phone:  +33 1 69354829
  Christophe.Janneteau@motorola.com  Emmanuel.Riou@motorola.com

  Alexandru Petrescu                 Alexis Olivereau
  Motorola Labs                      Motorola Labs
  Parc les Algorithmes St Aubin      Parc les Algorithmes St Aubin
  Gif-sur-Yvette 91193               Gif-sur-Yvette 91193
  France                             France
  Phone:  +33 1 69354827             Phone:  +33 1 69352516
  Alexandru.Petrescu@motorola.com    Alexis@motorola.com

  Hong-Yon Lach
  Motorola Labs
  Parc les Algorithmes St Aubin
  Gif-sur-Yvette 91193
  France
  Phone:  +33 1 69352536
  Hong-Yon.Lach@motorola.com

Copyright (C) The Internet Society (2004).  All Rights Reserved.

This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published and
distributed, in whole or in part, without restriction of any kind,
provided that the above copyright notice and this paragraph are
included on all such copies and derivative works.  However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of developing
Internet standards in which case the procedures for copyrights defined
in the Internet Standards process must be followed, or as required to
translate it into languages other than English.

The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.

This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT
NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN
WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Funding for the RFC editor function is currently provided by the
Internet Society.




Janneteau et al.          Expires October 2004                [Page 13]