Internet DRAFT - draft-sfigueiredo-multimob-use-case-dmm

draft-sfigueiredo-multimob-use-case-dmm



MULTIMOB Group                                            S. Figueiredo
Internet Draft                                   Universidade de Aveiro
Intended status: Informational                                  S. Jeon
Expires: April 22, 2013                   Instituto de Telecomunicacoes
                                                            R. L. Aguiar
                                                 Universidade de Aveiro
                                                       October 22, 2012



       IP Multicast Use Cases and Analysis over Distributed Mobility
                                Management


               draft-sfigueiredo-multimob-use-case-dmm-03.txt


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 22, 2013.

Copyright Notice

   Copyright (c) 2012 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



Figueiredo, et al.         Expires April 22                    [Page 1]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document. Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document. Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Abstract

   Mobile networks are changing towards distributed mobility management
   (DMM), tackling inefficiencies of existing mobility protocols
   regarding network management and packet routing. Identifying IP
   multicast use cases applicable to DMM is a logical step before
   exploring solution spaces. This document describes use cases where IP
   multicast is applied in DMM environments, considering two main
   deployment options: multicast router or MLD-Proxy deployment at a
   Mobility Access Router (MAR). Due to the lack of standard terminology,
   we refer to MAR as the entity embedding mobility-related functions,
   e.g. providing network access and flow anchoring capabilities. Each
   deployment option is thoroughly analyzed regarding its advantages and
   disadvantages, and both multicast listener and source mobility
   support are considered.

Table of Contents

   1. Introduction...................................................3
   2. Conventions and Terminology....................................3
   3. Use Cases Description..........................................4
      3.1. Assumptions...............................................4
      3.2. Mobile Multicast listener.................................5
         3.2.1. MLD Proxy deployment at MAR..........................5
            3.2.1.1. Operation.......................................5
            3.2.1.2. Detailed analysis...............................6
         3.2.2. Multicast Router deployment at MAR...................8
            3.2.2.1. Operation.......................................8
            3.2.2.2. Detailed analysis...............................8
      3.3. Multicast sender support..................................9
         3.3.1. MLD Proxy deployment at MAR..........................9
            3.3.1.1. Operation.......................................9
            3.3.1.2. Detailed analysis..............................10
         3.3.2. Multicast Router deployment at MAR..................14
            3.3.2.1. Operation......................................14
            3.3.2.2. Detailed analysis..............................14
      3.4. Summary of analysis......................................15
   4. IANA Considerations...........................................16
   5. Security Considerations.......................................16


Figueiredo, et al.      Expires April 22, 2013                 [Page 2]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   6. References....................................................17
      6.1. Normative References.....................................17
      6.2. Informative References...................................17



1. Introduction

   Centralized mobility management approach brings several drawbacks
   such as non-optimal routing or severe overloading on the anchor point.
   Such problems are expected to be more severe as mobile devices data
   consumption (and generation) increases.

   In order to tackle these problems, the concept of distributed
   mobility management (DMM) has emerged. It couples per flow and
   distributed anchoring, bringing the mobility anchor closer to the MN.

   IP multicast, as one of the enablers for efficient distribution of
   multimedia content, is composed of two main functions: multicast
   routing and membership subscription. When those are coupled with IP
   mobility, it is very critical to know possible use cases and
   respective issues, since those techniques were mainly designed for
   fixed networks.

   This document presents possible use cases of IP multicast in a DMM
   environment, by aligning to DMM Requirements [DMMREQ]. The different
   use cases result from the different functionalities provided at the
   MAR - MLD Proxy or MR -, and both mobile listener and sender support
   are analyzed. The goal was to identify the advantages (e.g. ease of
   deployment, resource-lightness) and constrains (e.g. lack of resource
   efficiency, non-optimal routing) of each deployment option,
   considering its impact for the support of mobile sender or mobile
   listeners.

2. Conventions and Terminology

   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].

   This document uses the terminology defined in [RFC5213], [RFC6275],
   and [RFC3810], and [RFC4601]. Also, new entities are defined relying
   on the PMIPv6 entities specified in [RFC5213]:

     - Mobility Access Router (MAR): A router with the capability of
     acting both as a mobility anchor and as an access router, in a per
     flow basis. It can act as either a P-MAR, a N-MAR, or both.


Figueiredo, et al.      Expires April 22, 2013                 [Page 3]

Internet-Draft       Use Cases for Multicast DMM           October 2012


     - Previous Mobility Access Router (P-MAR): The MAR where the MN was
     attached to previously to the IP mobility event, and which is
     acting as an anchor for one or multiple flows.

     - New Mobility Access Router (N-MAR): The MAR to which the MN is
     currently attached. It provides the network access and thus
     delivers all the flows destined to the MN's HNPs - including those
     anchored to previously visited P-MARs.

     - Multicast Listener Discovery Proxy (MLD-P): An entity providing
     MLD based forwarding following the operation defined in [RFC4605].
     In the current document, only MLDv2-based signaling is considered,
     targeting IPv6 networks (REQ3 from [DMMREQ]).

3. Use Cases Description

   We identify different use cases that result from the application of
   existing standards for IP multicast support over mobile environments.
   We first consider mobile multicast listeners and then mobile
   multicast senders, and for each case we analyze the impact of
   deploying distinct functionality at the MAR: either MLD Proxy or
   Multicast Router (namely PIM-SM capability).

3.1. Assumptions

   A1: This draft refers to the requirements in [DMMREQ] as the base DMM
   framework.

   A2: Network access and data anchor functionalities are based in
   [RFC5213], and are assumed to be provided by a Mobility Access Router
   (MAR).

   A3: It is assumed that when the IP mobility happens, at least one
   multicast flow is being received (listener) / sent (sender). and a
   mobility tunnel will be created between the P-MAR and the N-MAR.

   A4: Using MLD Proxy, and when a user without mobility session starts
   a multicast session at a MAR, the upstream interface of MLD Proxy is
   configured towards an upstream multicast router in the multicast
   infrastructure. But when a user moves to another MAR (N-MAR), under
   assumption A3, the upstream interface of MLD Proxy will be configured
   towards the anchor that enables unicast IP address continuity to be
   kept (P-MAR, analogous to LMA function).

   A5: Mobility occurs within a single PIM-SM multicast routing domain.




Figueiredo, et al.      Expires April 22, 2013                 [Page 4]

Internet-Draft       Use Cases for Multicast DMM           October 2012




3.2. Mobile Multicast listener

3.2.1. MLD Proxy deployment at MAR

3.2.1.1. Operation

   In this use case, MLD Proxy is considered as being deployed at all
   MARs, and operating in an analogous way to the Base Multicast support
   solution for PMIPv6 [RFC4605]. As such, after mobility, the upstream
   interface MUST be configured towards the mobility tunnel endpoint,
   i.e., P-MAR.

   When a MN initially attaches to the P-MAR (as shown in Figure 1), it
   receives a HNP address which will be associated with communications
   started at that MAR. As the P-MAR detects the new logical link, it
   transmits a General MLD Query message to which the MN will not yet
   reply, as it is not yet running any multicast session. The P-MAR then
   adds the downstream logical link to the MLD Proxy instance [RFC4605].
   In this case, i.e. when users subscribe to multicast content only
   after associating with the MAR, the MLD Proxy will set its uplink to
   the multicast infrastructure. When the MN intends to start receiving
   the multicast session, it will send an unsolicited MLD Report,
   triggered by its application. On receiving the latter message, the
   MLD Proxy tries to join the multicast channel(s) by sending an
   aggregated MLD Report through the MLD Proxy upstream interface. Note
   that the same MLD Proxy instance will be assigned to all MNs which
   initiated their multicast subscriptions in the current MAR (i.e. the
   MNs having no multicast mobility session). When the joining procedure
   ends, multicast data is transmitted through the same interfaces,
   until reaching the MN.

















Figueiredo, et al.      Expires April 22, 2013                 [Page 5]

Internet-Draft       Use Cases for Multicast DMM           October 2012


                  +----------------+
                  |   Multicast    |
                  | Infrastructure |
                  +----------------+
                          *
                          * (S,G)
                          *
                     +----------+               +----------+
                     |  P-MAR   |---------------|  N-MAR   |
                     |          |***************|          |
                     | (MLD-P)  |---------------| (MLD-P)  |
                     +----------+               +----------+
                          *                           *
                          *                           *
                       +------+                    +------+
                       |  MN  |       ----->       |  MN  |
                       +------+                    +------+


       Figure 1 Multicasting architecture using distributed mobility
                                management

   When the MN moves from P-MAR, the N-MAR is required to establish a
   tunnel for IP session continuity of the flows being sent towards and
   from the MN's HNP assigned by the P-MAR. This implies that N-MAR has
   an appropriate method to know the P-MAR. Multiple ideas are supposed
   to be made at the solution stage of DMM WG, therefore it is out of
   scope of this document. Following the operation of the MLD Proxy
   [RFC4605], after the bidirectional tunnel establishment, the
   following process takes place. First, the N-MAR sends a General MLD
   Query, and verifies whether the MN is admissible for multicast
   sessions. Then, the MLD Proxy at the N-MAR adds the downstream
   interface corresponding to the MN, and configures the upstream
   interface towards the MN's P-MAR, i.e. the tunnel endpoint.

3.2.1.2. Detailed analysis

   This scheme is simple and directly applicable to a network-based
   multicast DMM approach, as no extensions for multicast-related
   operation are required. Besides, the MLDv2 Proxy was devised with a
   less resource-demanding nature compared to MRs, aimed for scenarios
   where multicast routing is not beneficial or required. The same
   applies in DMM scenarios. However, this approach leads to a couple of
   relevant issues. Traffic duplication is the result of the tunnel
   convergence problem, occurring e.g. in [RFC6224]. As shown in Figure
   2, MN1 and MN3, which moved from MAR1 and MAR3, respectively, are
   currently located at the MAR2. Through their respective tunnels, they


Figueiredo, et al.      Expires April 22, 2013                 [Page 6]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   receive multicast packets of the same channel through different
   anchoring MARs. This causes duplicated traffic to converge to the
   MAR2. Concretely, the magnitude of replicated traffic is expected to
   be much higher than that of PMIPv6, considering the consequences of
   having a single mobility entity coupling anchoring and network access
   provision: we assume that the number of MARs in future DMM domains
   will be much larger than that of LMAs at core level within a PMIPv6
   domain.

          +----------------+
          | Multicast Tree | *
          +----------------+     *
                  *        *         *
                  *          *           *
                  *            *             *
            (S,G) *       (S,G)  *               *  (S,G)
                  *                *                 *
             +----------+  (-->)  +----------+ (<--)   +----------+
             |   MAR1   |---------|   MAR2   |---------|   MAR3   |
             |          |*********|          |*********|          |
             | (MLD-P)  |---------| (MLD-P)  |---------|  (MLD-P) |
             +----------+  Tun.1  +----------+  Tun.2  +----------+
                                    *   *   *
                                   *    *    *
                                  *     *     *
              +---+  move      +---+  +---+ +---+   move   +---+
              |MN1|  --->      |MN1|  |MN2| |MN3|   <---   |MN3|
              +---+            +---+  +---+ +---+          +---+

                (<--/-->) : direction of the multicast packet flow


                         Figure 2 Data replication

   Another issue is non-optimal routing (Figure 3). If we consider a
   significantly large domain, multicast packets may traverse a long
   distance, depending on the setup direction of the upstream interface
   of MLD Proxy instances. The issue is more noticeable if we assume all
   MARs are connected to the multicast infrastructure.

   When MLD Proxy is used in mobile multicast, low performance handover
   will result from the need of going through the following process: 1)
   MLD Proxy sets up the new MLD interface, 2) MLD Proxy sends the
   General MLD Query, 3) MN sends the MLD Report. Only then MN will
   receive the content, which for a significant number of IP multicast-
   based applications, will represent a noticeable service disruption.



Figueiredo, et al.      Expires April 22, 2013                 [Page 7]

Internet-Draft       Use Cases for Multicast DMM           October 2012




            +----------------+
            |   Multicast    |
            | Infrastructure |
            +----------------+
                     *
                     * (S,G)
                     *
                +----------+                 +----------+
                |  P-MAR   |------     ------|  N-MAR   |
                |          |****** ... ******|          |
                |(MLD-P)   |------     ------| (MLD-P)  |
                +----------+                 +----------+
                     *                             *
                     *                             *
                  +------+                     +------+
                  |  MN  |        ----->       |  MN  |
                  +------+                     +------+


                   Figure 3 Non-optimal routing problem



3.2.2. Multicast Router deployment at MAR

3.2.2.1. Operation

   In this use case, PIM-SM is deployed at all MARs. After the MN
   attaches to a MAR, its PIM instance will act as the Designated Router
   (DR) for the MN (and all other attached nodes). The procedure for
   multicast subscription will be the same as in a fixed network, i.e.
   the MN sends the Explicit MLD Report, and the MR at the MAR will
   process that message and join any multicast channel if needed.

   When the MN moves, the mobility tunnel will be setup between the two
   MARs. As soon as the MN sends a MLD Report to the N-MAR, N-MAR will
   join the multicast group / channel (if needed), the traffic will
   start flowing from the tunnel and a new downstream state will be
   configured.

3.2.2.2. Detailed analysis

   An advantage of using MRs within MARs, is that multicast routing is
   not affected due to RPF check. which leads to the selection of a
   single upstream interface per MAR. This selection is independent of


Figueiredo, et al.      Expires April 22, 2013                 [Page 8]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   the number of existing mobility tunnels. On the other hand, the usage
   of MLD Proxy might lead to the tunneling of the same multicast group
   to a common MAR, and might mean severe replication, similarly to
   [RFC6224].

   As referred, although the mobility tunnel is activated after the MN
   mobility, N-MAR will only subscribe the required multicast group /
   channel after receiving the explicit MLD Report This can translate
   into several lost packets, as the MN isn't aware of the mobility
   process, and the General MLD Queries sent by MRs are periodic with a
   varying frequency in the order of several seconds. Summarizing,
   smooth handover cannot be assured without extending existing
   mechanisms. When a MN moves to its N-MAR, there is the possibility
   that the multicast channel(s) is (are) already being distributed
   there. In such case, the new MAR will directly forward the multicast
   traffic to the MN without using the tunnel for subscription function.
   However, different MARs might be receiving common channels at
   distinct times, which from the point of view from the receiver, will
   result in frames replay (if receiving the same frames again) or miss
   (in case the new MAR transmission is more advanced in time).

   A possible non-efficiency is the unnecessary tunnel creation.
   Following assumption A3, the tunnel is created due to the existence
   of at least an ongoing multicast flow. The tunnel is created
   independently of N-MAR's awareness to multicast subscriptions,
   because the MLD Query occurs after its creation. Although, if the
   subscription(s) of interest is (are) already being subscribed at the
   N-MAR's MR, the tunnel will not be used at all for the multicast
   subscription transmission. Thus, large scale creation of unnecessary
   tunnels represents non-negligible wasted processing overhead.

3.3. Multicast sender support

3.3.1. MLD Proxy deployment at MAR

3.3.1.1. Operation

   Sender mobility support is known to lead to service disruption
   problems impacting all multicast tree, in particular if SPT is active.
   In [SENDER], the utilization of MLD Proxy is proposed, being the
   upstream interface always configured towards the fixed anchoring
   entity - the LMA. To allow the sender to transmit multicast content
   to the multicast tree in a DMM framework, the MLD Proxy should
   configure its upstream interface towards a multicast router [PM-HOME].
   Depending on the network topology, it may also be configured towards
   a MLD Proxy placed on a neighbor MAR. On the multicast source's
   mobility (Figure 4), an identical operation to the listener mobility


Figueiredo, et al.      Expires April 22, 2013                 [Page 9]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   case is expected from the MLD Proxy behavior. In this case, the
   source uploads multicast traffic through one of MLD Proxy's
   downstream interfaces, and the traffic is forwarded through the
   uplink interface towards the P-MAR.


                      +------+         +----------------+
                      |  RP  |---------|   Multicast    |
                      +------+         | Infrastructure |
                          *            +----------------+
                          * (S,G)               |
                          *                     |
                     +----------+          +----------+
                     |  P-MAR   |----------|  N-MAR   |
                     |          |**********|          |
                     | (MLD-P)  |----------| (MLD-P)  |
                     +----------+          +----------+
                          *                      *
                          *                      *
                       +------+               +------+
                       |   S  |   ---->       |  S   |
                       +------+               +------+


                    Figure 4 Multicast sender mobility

3.3.1.2. Detailed analysis

   The utilization of MLD Proxy carries the previously referred
   advantages, as ease of deployment and operation lightness.

   When a source moves to N-MAR from P-MAR, multicast data will be sent
   through the mobility tunnel between N-MAR and P-MAR (Figure 5). If a
   listener (L1) attaches to the same MAR (N-MAR), it will receive the
   multicast data through multicast infrastructure, following the
   configuration of MLD Proxy. Hence, the multicast data is routed non-
   optimally between the source and the listener, going from N-MAR to P-
   MAR, to the multicast routing tree, and then back to N-MAR again
   before reaching the listener.










Figueiredo, et al.      Expires April 22, 2013                [Page 10]

Internet-Draft       Use Cases for Multicast DMM           October 2012


               +------+         +----------------+
               |  RP  |*********|   Multicast    |
               +------+         | Infrastructure |
                   *            +----------------+
                   * (S,G)             *
                   *                   *
              +----------+       +----------+
              |   P-MAR  |-------|   N-MAR  |
              |          |*******|          |
              | (MLD-P)  |-------| (MLD-P)  |
              +----------+       +----------+
                   *                *     *
                   *                *     *
                +------+  move +------+ +-----+
                |  S   |  ---> |   S  | |  L1 |
                +------+       +------+ +-----+


             Figure 5 Triangular routing after source mobility

   A similar problem occurs in the opposite process, i.e. if a multicast
   source starts transmitting multicast content at a MAR, and a listener
   moves to the same MAR while receiving the source's content (Figure
   6).

               +------+         +----------------+
               |  RP  |*********|   Multicast    |
               +------+         | Infrastructure |
                   *            +----------------+
                   * (S,G)             *
                   *                   *
              +----------+       +----------+
              |   N-MAR  |-------|   P-MAR  |
              |          |*******|          |
              | (MLD-P)  |-------| (MLD-P)  |
              +----------+       +----------+
                *       *              *
                *       *              *
            +------+  +----+  move  +----+
            |  S   |  | L1 |  <---  | L1 |
            +------+  +----+        +----+


   Figure 6 Triangular routing after listener mobility (MLD-Proxy case)





Figueiredo, et al.      Expires April 22, 2013                [Page 11]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   When the source and the listener are within a same MAR (MAR2), if
   both the source and listener try to send the session and receive it,
   respectively, the traffic will be optimally sent to the listener
   without going through native multicast infrastructure. As the traffic
   reaches the MLD Proxy via the downstream interface to which the
   source is attached, it will be sent through the downstream interface
   to which the listener sent the MLD Report. However, if the source and
   the listener move to different MARs, the traffic will traverse the
   following non-optimal path, even though they share a common anchor:

     Source -> MAR1 -> MAR2 -> Multicast Tree -> MAR2 -> MAR3

   This problem is depicted in Figure 7.

                             +----------------+
                             | Multicast Tree |
                             +----------------+
                                    *  *
                                    *  *
                                    *  *
                                    *  *
                                    *  *
           +----------+  (-->)  +----------+  (-->)  +----------+
           |   MAR1   |---------|   MAR2   |---------|   MAR3   |
           |          |*********|          |*********|          |
           | (MLD-P)  |---------| (MLD-P)  |---------|  (MLD-P) |
           +----------+  Tun.1  +----------+  Tun.2  +----------+
               *                  *       *              *
               *                 *         *             *
               *                *           *            *
             +---+    move    +---+       +---+  move  +---+
             | S |    <---    | S |       | L |  -->   | L |
             +---+            +---+       +---+        +---+

                (<--/-->) : direction of the multicast packet flow

   Figure 7 Multicast traffic non-optimal routing due to both mobile
   sender and listener

   REQ1 from [DMMREQ] refers that "DMM MUST enable a distributed
   deployment of mobility management of IP sessions so that the traffic
   can be routed in an optimal manner without traversing centrally
   deployed mobility anchors". When a MN subscribes to a new multicast
   session with existing multicast mobility session, the Aggregated MLD
   Report containing all the MN's multicast subscriptions will be sent



Figueiredo, et al.      Expires April 22, 2013                [Page 12]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   from the current MLD Proxy through the same uplink interface, i.e.
   towards a single multicast mobility anchor. This results in some of
   previously identified issues (e.g. non-optimality in the path that
   both the subscription and multicast traffic traverse). It can be
   stated that the MLD Proxy nature doesn't comply with the
   aforementioned requirement, leading to the subscription of any
   multicast flow using the same multicast mobility data path.

   This problem is depicted in Figure 8, where both multicast flow 1 and
   flow 2 reach MAR2 from MAR1, being flow 2's optimal routing path
   affected by the mobility status of the MN, and in particular by the
   order in which the multicast flows were subscribed. While this issue
   is not exclusively related to mobile multicast sources, it is better
   depicted and its' impact in the routing is more obvious when
   considering one.

                             +----------------+
                             | Multicast Tree |
                           * +----------------+#
                         *                       #
                       *                           #
                     *                               #
                   *                                   #
           +----------+  (-->)                     +----------+
           |   MAR1   |---------            -------|   MAR2   |
           |          |#*#*#*#*#............#*#*#*#|          |
           | (MLD-P)  |---------            -------|  (MLD-P) |
           +----------+  Tunnel (used after mob.)  +----------+
               *                                      *    #
               *                          downstream) #    # (upstream)
               *                                      *    #
               *                                      #    #
             +---+                 move            +---+  +---+
             | L |             ---------->         | L |  | S |
             +---+                                 +---+  +---+

             * : Multicast flow 1
             # : Multicast flow 2 (subscribed after some time in MAR2)


   Figure 8 Non-optimal routing due to single MLD Proxy uplink





Figueiredo, et al.      Expires April 22, 2013                [Page 13]

Internet-Draft       Use Cases for Multicast DMM           October 2012


3.3.2. Multicast Router deployment at MAR

3.3.2.1. Operation

   When a source starts transmitting multicast traffic, the content will
   be encapsulated in PIM-Register messages, and sent towards the RP
   (e.g .statically configured or discovered through a Bootstrapping
   Router (BSR)). In DMM, the RP can be a core MR or a MAR (including
   the initial MAR). The RP's SPT and each of the DR's SPTs may then be
   created. When the source moves, N-MAR's MR will create the state for
   the new multicast group, and the traffic will be forwarded using the
   tunnel to the P-MAR, until reaching the RP (unless it is the PIM
   instance at the P-MAR itself). It is then sent down the RPT. Again,
   the creation of the SPTs will typically be triggered following PIM-SM
   regular operation.

3.3.2.2. Detailed analysis

   In case the RP's Source Path Tree is built before the mobility
   process, it will be destroyed due to mobility, and the tree
   construction process will be reinitiated at the new MAR. Also, in
   case the Shortest Path Tree between the listener's DRs and the
   sender's DR is being used, mobility will reset the PIM process to the
   RPT stage. This means that each sender mobility event results in
   increased signaling overhead and delay as consequence of the
   multicast routing convergence (i.e. Phase 2 and Phase 3 from PIM-SM
   operation). Moreover, non-optimal routing occurs when the RPT is used.
   When a sender moves to a MAR where listeners are subscribing to the
   channel it is sending, the multicast traffic will always reach the N-
   MAR by going through the RP (as shown in Figure 9).

   Using PIM-SM in DMM scenarios there is a trade-off between the
   routing non-optimality of RPT and the non-efficient consequences of
   frequent SPT establishment. It is important to note that this impact
   is magnified the more listener's DRs are receiving the multicast
   channel(s).













Figueiredo, et al.      Expires April 22, 2013                [Page 14]

Internet-Draft       Use Cases for Multicast DMM           October 2012


               +------+         +----------------+
               |  RP  |*********|   Multicast    |
               +------+         | Infrastructure |
                   *            +----------------+
                   * (S,G)             *
                   *                   *
              +----------+       +----------+
              |   P-MAR  |-------|   N-MAR  |
              |          |*******|          |
              |   (MR)   |-------|   (MR)   |
              +----------+       +----------+
                   *                *     *
                   *                *     *
                +------+  move +------+ +-----+
                |  S   |  ---> |   S  | |  L1 |
                +------+       +------+ +-----+


        Figure 9 Triangular routing after source mobility (MR case)

3.4. Summary of analysis

   Table I summarizes the previous analysis, globally depicting the
   differences between MLD Proxy and MR over DMM. The relevant sections
   for each feature are appended below, and the meaning of each feature
   is extended afterwards.

   ===================================================================
   | Feature \ Function |    MLD Proxy        |   Multicast Router   |
   ===================================================================
   |    Lightweight     |        Yes          |         No           |
   -------------------------------------------------------------------
   |     Optimal        |         No          |   Yes (using SPT)    |
   |     routing        |(3.2.1.2)/ (3.3.1.2) |     (3.3.2.2)        |
   -------------------------------------------------------------------
   |     Efficient      |         No          |        Yes           |
   |    distribution    |(3.2.1.2)/ (3.3.1.2) | (3.2.2.2),(3.3.2.2)  |
   -------------------------------------------------------------------
   |    Distributed     |        No           |         Yes          |
   |    anchoring       |    (3.3.1.2)        |      (3.2.2.2)       |
   -------------------------------------------------------------------
   | Seamless mobility  |        No           |          No          |
   |    (listener)      |    (3.2.1.2)        |      (3.2.2.2)       |
   -------------------------------------------------------------------
   |    HO signaling    |        Low          | High (when using SPT)|
   |      overhead      |(3.2.1.2), (3.3.1.2) |      (3.3.2.2)       |
   -------------------------------------------------------------------


Figueiredo, et al.      Expires April 22, 2013                [Page 15]

Internet-Draft       Use Cases for Multicast DMM           October 2012


   |     Tunnel is      |    Doesn't apply    |          No          |
   |    always used     |                     |      (3.2.2.2)       |
   ===================================================================
          Table I. Comparison between MLD Proxy and MR deployment

   The meaning of each of the entries is as follows:

   Lightweight: this entry reflects whether the deployed multicast
   feature has a resources-wise lightweight operation.

   Optimal routing: This entry reflects whether optimal routing is
   assured.

   Efficient distribution: This entry reflects vulnerability to
   multicast traffic replication.

   Distributed anchoring: This entry assesses whether for a single MN,
   different multicast streams can be anchored at different mobility
   anchors or not.

   Seamless mobility (listener): This entry reflects whether IP mobility
   is seamless from the point of view of the mobile listener's
   application.

   HO signaling overhead: This entry assesses the amount of signaling
   introduced by the IP mobility of a MN represents. This signaling may
   be relative to the mobility protocol or the multicast routing
   operation.

   Tunnel is always used: This entry assesses whether the mobility
   tunnel utilization is assured or not.



4. IANA Considerations

   This document makes no request of IANA.



5. Security Considerations

   TBD


Figueiredo, et al.      Expires April 22, 2013                [Page 16]

Internet-Draft       Use Cases for Multicast DMM           October 2012




6. References

6.1. Normative References

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

   [RFC6275] Johnson, D., Perkins, C., and J. Arkko, "Mobility Support
             in IPv6", RFC 6275, July 2011.

   [RFC3810] R. Vida, and L. Costa, "Multicast Listener Discovery
             Version 2 (MLDv2) for IPv6," IETF RFC 3810, June 2004.

   [RFC5213] S. Gundavelli, K. Leung, V. Devarapalli, K. Chowdhury, and
             B. Patil, "Proxy Mobile IPv6", IETF RFC 5213, August 2008.

   [RFC4605] B. Fenner, H. He, B. Haberman, and H. Sandick, "Internet
             Group Management Protocol (IGMP) / Multicast Listener
             Discovery (MLD) Based Multicast Forwarding ("IGMP/MLD
             Proxying")", IETF RFC 4605, August 2006.

   [RFC4601] B. Fenner, M. Handley, H. Holbrook, and I. Kouvelas,
             "Protocol Independent Multicast - Sparse Mode (PIM-SM):
             Protocol Specification (Revised)", RFC 4601, August 2006.

6.2. Informative References

   [RFC6224] T. Schmidt, M. Waehlisch, S. Krishnan, "Base Deployment for
             Multicast Listener Support in PMIPv6 Domain", RFC 6224,
             April 2011.

   [DMMREQ] H. Chan, "Requirements of distributed mobility management",
             draft-ietf-dmm-requirements-02 (work in progress),
             September 2012.

   [SENDER] T C. Schmidt et al, "Mobile Multicast Sender Support in
             Proxy Mobile IPv6 (PMIPv6) Domains", draft-ietf-multimob-
             pmipv6-source-01 (work in progress), July 2012.

   [PM-HOME] S. Jeon, N. Kang, and Y. Kim, "Mobility Management based on
             Proxy Mobile IPv6 for Multicasting Services in Home
             Networks," IEEE Transactions on Consumer Electronics (TCE),
             vol. 55, no. 3, pp. 1227-1232, August 2009.




Figueiredo, et al.      Expires April 22, 2013                [Page 17]

Internet-Draft       Use Cases for Multicast DMM           October 2012


Authors' Addresses

   Sergio Figueiredo
   Universidade de Aveiro
   3810-193 Aveiro, Portugal

   E-mail: sfigueiredo@av.it.pt

   Seil Jeon
   Instituto de Telecomunicacoes
   Campus Universitario de Santiago
   3810-193 Aveiro, Portugal

   E-mail: seiljeon@av.it.pt

   Rui L. Aguiar
   Universidade de Aveiro
   3810-193 Aveiro, Portugal

   E-mail: ruilaa@ua.pt





























Figueiredo, et al.      Expires April 22, 2013                [Page 18]