<?xml version='1.0' encoding='utf-8'?>
<!-- This template is for creating an Internet Draft using xml2rfc,
    which is available here: http://xml.resource.org. -->
<!DOCTYPE rfc SYSTEM "rfc2629-xhtml.ent" [
  <!ENTITY rfc2119 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml">
  <!ENTITY rfc4360 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4360.xml">
  <!ENTITY rfc4684 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4684.xml">
  <!ENTITY rfc5213 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5213.xml">
  <!ENTITY rfc7333 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.7333.xml">
  <!ENTITY rfc7454 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.7454.xml">
  <!ENTITY rfc8402 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8402.xml">
  <!ENTITY rfc8754 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8754.xml">
  <!ENTITY rfc8885 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8885.xml">
  <!ENTITY rfc8986 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8986.xml">
  <!ENTITY rfc9433 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.9433.xml">
  <!ENTITY I-D.ietf-spring-segment-routing-policy SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-spring-segment-routing-policy.xml">
  <!ENTITY I-D.ietf-spring-sr-service-programming SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-spring-sr-service-programming.xml">
  <!ENTITY rfc8754 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8754.xml">
  <!ENTITY I-D.ietf-spring-segment-routing-central-epe SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-spring-segment-routing-central-epe.xml">
  <!ENTITY I-D.ietf-lsr-flex-algo SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-lsr-flex-algo.xml">
  <!ENTITY I-D.ietf-dmm-fpc-cpdp SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-dmm-fpc-cpdp.xml">
  <!ENTITY I-D.matsushima-spring-srv6-deployment-status SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.matsushima-spring-srv6-deployment-status.xml">
  <!ENTITY I-D.ietf-mpls-mna-fwk SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-mpls-mna-fwk.xml">
]>
<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
<!-- used by XSLT processors -->
<!-- For a complete list and description of processing instructions (PIs), 
    please see http://xml.resource.org/authoring/README.html. -->
<rfc
      xmlns:xi="http://www.w3.org/2001/XInclude"
      category="std"
      docName="draft-ietf-dmm-mup-architecture-02"
      ipr="trust200902"
      obsoletes=""
      updates=""
      submissionType="IETF"
      xml:lang="en"
      tocInclude="true"
      tocDepth="4"
      symRefs="true"
      sortRefs="true"
      version="3">
  <!-- xml2rfc v2v3 conversion 2.38.1 -->
  <!-- category values: std, bcp, info, exp, and historic
    ipr values: trust200902, noModificationTrust200902, noDerivativesTrust200902,
       or pre5378Trust200902
    you can add the attributes updates="NNNN" and obsoletes="NNNN" 
    they will automatically be output with "(if approved)" -->

 <!-- ***** FRONT MATTER ***** -->

 <front>
    <!-- The abbreviated title is used in the page header - it is only necessary if the 
        full title is longer than 39 characters -->

   <title abbrev="MUP Architecture">Mobile User Plane
   Architecture for Distributed Mobility Management</title>
    <seriesInfo name="Internet-Draft" value="draft-ietf-dmm-mup-architecture-02"/>
    <!-- add 'role="editor"' below for the editors if appropriate -->

   <!-- Another author who claims to be an editor -->

   <author fullname="Satoru Matsushima" initials="S." surname="Matsushima">
      <organization>SoftBank</organization>
      <address>
        <postal>
          <street/>
          <!-- Reorder these if your country does things differently -->

         <city></city>
          <region/>
          <code/>
          <country>Japan</country>
        </postal>
        <email>satoru.matsushima@g.softbank.co.jp</email>
        <!-- uri and facsimile elements may also be added -->
     </address>
    </author>
    <author fullname="Katsuhiro Horiba" initials="K." surname="Horiba">
      <organization></organization>
      <address>
        <postal>
          <street/>
          <city></city>
          <region/>
          <code/>
          <country>Japan</country>
        </postal>
        <email>katsuhiro.horiba@gmail.com</email>
      </address>
    </author>
   <author fullname="Yuya Kawakami" initials="Y." surname="Kawakami">
      <organization>SoftBank</organization>
      <address>
        <postal>
          <street/>
          <!-- Reorder these if your country does things differently -->

         <city></city>
          <region/>
          <code/>
          <country>Japan</country>
        </postal>
        <email>yuya.kawakami01@g.softbank.co.jp</email>
        <!-- uri and facsimile elements may also be added -->
     </address>
    </author>
    <author fullname="Tetsuya Murakami" initials="T." surname="Murakami">
      <organization abbrev="Arrcus, Inc">
            Arrcus, Inc.</organization>
      <address>
        <postal>
          <street/>
          <city/>
          <region/>
          <code/>
          <country>USA</country>
        </postal>
        <email>tetsuya@arrcus.com</email>
      </address>
    </author> 
    <author fullname="Keyur Patel" initials="K." surname="Patel">
      <organization abbrev="Arrcus, Inc">
            Arrcus, Inc.</organization>
      <address>
        <postal>
          <street/>
          <city/>
          <region/>
          <code/>
          <country>USA</country>
        </postal>
        <email>keyur@arrcus.com</email>
      </address>
    </author> 
    <author fullname="Jakub Horn" initials="J." surname="Horn">
      <organization abbrev="Cisco Systems, Inc.">
            Cisco Systems, Inc.</organization>
      <address>
        <postal>
          <street/>
          <city/>
          <region/>
          <code/>
          <country>Czech Republic</country>
        </postal>
        <email>jakuhorn@cisco.com</email>
      </address>
    </author> 

    <date year="2026"/>
    <!-- If the month and year are both specified and are the current ones, xml2rfc will fill 
        in the current day for you. If only the current year is specified, xml2rfc will fill 
   in the current day and month for you. If the year is not the current one, it is 
   necessary to specify at least a month (xml2rfc assumes day="1" if not specified for the 
   purpose of calculating the expiry date).  With drafts it is normally sufficient to 
   specify just the year. -->

   <!-- Meta-data Declarations -->

   <area>General</area>
    <workgroup>Internet Engineering Task Force</workgroup>
    <!-- WG name at the upperleft corner of the doc,
        IETF is fine for individual submissions.  
   If this element is not present, the default is "Network Working Group",
        which is used by the RFC Editor as a nod to the history of the IETF. -->

   <keyword>mobile userplane, segment routing, SRv6</keyword>
    <!-- Keywords will be incorporated into HTML output
        files in a meta tag but they have no effect on text or nroff
        output. If you submit your draft to the RFC Editor, the
        keywords will be used for the search engine. -->

    <abstract>
        <t>This document defines the Mobile User Plane 
        (MUP) architecture for Distributed Mobility Management. 
        The requirements for Distributed Mobility Management described in 
        <xref target="RFC7333"/> can be satisfied by routing fashion.</t>

        <t>In MUP Architecture, session information 
        between the entities of the mobile user plane 
        is turned to routing information so that 
        mobile user plane can be integrated into dataplane.</t>
        
        <t>MUP architecture is designed to be pluggable
        user plane part of existing mobile service architectures,
        enabled by auto-discovery for the use plane. 
        Segment Routing provides network programmability
        for a scalable option with it.</t>

        <t>While MUP architecture itself is independent 
        from a specific dataplane protocol, 
        several dataplane options are available for 
        the architecture. 
        This document describes IPv6 dataplane in 
        Segment Routing case (SRv6 MUP) due to the DMM requirement, 
        and is suitable for mobile services which require 
        a large IP address space.</t>



    </abstract>
  </front>
  <middle>
    <section anchor="intro" numbered="true" toc="default">
      <name>Introduction</name>
      <t>Mobile service systems require IP connectivity for communication 
      between the entities defined by mobile service architectures for the mobile service systems.
      <xref target="RFC5213"/><xref target="TS.23501"/>. </t>
      <!--
      To provide the IP connectivity, Segment Routing (SR) 
      <xref target="RFC8402"/>can be a candidate solution.</t>
      -->

      <t>In PMIPv6 <xref target="RFC5213"/>, IP connectivity 
      is required between 
      LMA (Local Mobility Anchor) and MAG (Mobility Access Gateway), as well as LMA and 
      Internet. In 3GPP 5G <xref target="TS.23501"/>,
      IP connectivity for N3 interface between gNodeB(es) 
      and UPFs (User Plane Function) is required, as well as for N6 interface 
      between UPFs and DNs (Data Network).</t>

      <t>These IP connectivities may be covered by multiple 
      dataplane networks, such as IPv4, IPv6, MPLS, or bunch of
      dataplane protocols.
      When just one dataplane protocol network is adopted 
      for simplicity, it is expected that 
      the address space of the dataplane network should be large enough 
      to cover a vast number of nodes, such as millions of base stations. 
      For this reason, use of IPv6 dataplane looks 
      sufficiently suitable.</t>
      
      <!-- 
       while MPLS Label space in an SR network 
      is limited even though MPLS dataplane 
      may also be available for IPv6 networks.</t>
      -->

      <t>IPv6 dataplane has been able to instantiate 
      Segment Routing over IPv6 (SRv6) with 
      network programming capability described in 
      <xref target="RFC8986"/>.</t>

      <t>SRv6 network programmability enhances IPv6 dataplane 
      to be integrated with mobile user plane 
      <xref target="RFC9433"/>. 
      It will make an entire IPv6 network support the user plane 
      in a very efficient distributed routing fashion.</t>

      <t>On the other hand, the requirements for 
      Distributed Mobility Management (DMM) described in 
      <xref target="RFC7333"/> can be satisfied by session management 
      based solutions. <xref target="RFC8885"/> defines 
      protocol extension to PMIPv6 for the DMM requirements. 
      3GPP 5G defines an architecture in which 
      multiple session anchors can be added to one 
      PDU session by the session management.</t>

      <t>As a reminder, the user plane related requirements 
      in <xref target="RFC7333"/> are reproduced here:</t>

      <dl newline="true" spacing="normal" indent="8">

        <dt>REQ1: Distributed mobility management</dt>
        <dd>IP mobility, network access solutions, and forwarding 
        solutions provided by DMM MUST enable traffic to avoid 
        traversing a single mobility anchor 
        far from the optimal route. It is noted that the 
        requirement on distribution 
        applies to the data plane only.</dd>

        <dt>REQ3: IPv6 deployment</dt>
        <dd>DMM solutions SHOULD target IPv6 as the primary deployment 
        environment and SHOULD NOT be tailored specifically to support 
        IPv4, particularly in situations 
        where private IPv4 addresses and/or NATs are used.</dd>

        <dt>REQ4: Existing mobility protocols</dt>
        <dd>A DMM solution MUST first consider reusing and extending 
        IETF standard protocols before specifying new protocols.</dd>

        <dt>REQ5: Coexistence with deployed networks/hosts and 
        operability across different networks</dt>
        <dd>A DMM solution may require loose, tight, or no integration 
        into existing mobility protocols and host IP stacks. Regardless 
        of the integration level, DMM implementations MUST be able to 
        coexist with existing network deployments, end hosts, and routers 
        that may or may not implement existing mobility protocols. 
        Furthermore, a DMM solution SHOULD work across different networks, 
        possibly operated as separate administrative domains, 
        when the needed mobility management signaling, forwarding, 
        and network access are allowed by the trust relationship 
        between them.</dd>
      </dl>

      <t>This document defines the  
      Mobile User Plane (MUP) architecture
      for Distributed Mobility Management. 
      MUP is not a mobility management system itself, 
      but an architecture enables the dataplanes to integrate 
      mobile user plane into it for the IP networks.</t>
            
      <t>Although MUP architecture is independent 
      from a specific dataplane protocol, 
      this document describes IPv6 dataplane in 
      Segment Routing case (SRv6 MUP) due to the DMM requirement, 
      and as a suitable solution for scalable mobile service
      deployments. Other dataplane options is out of scope of
      this document, and may be described in the future.</t>

      <t>In this routing paradigm, a session information from 
      a mobile control plane of a mobile service system will be transformed to 
      routing information. It means that any MUP dataplane nodes
      become functional to the session instead of the mobile user plane 
      specific nodes for the role of anchor or intermediate points.
      The user plane anchor and intermediate functions 
      can be supported by MUP enabled networks (REQ1), 
      not to mention that MUP will naturally be 
      deployed over IPv6 networks (REQ3).</t>

      <t>MUP architecture is independent from the mobile 
      service system. For the requirements (REQ4, 5), 
      MUP architecture is designed to be pluggable user plane 
      part of existing mobile service architectures. 
      Those existing architectures are for example defined in 
      <xref target="RFC5213"/>, <xref target="TS.23501"/>, 
      or if any.</t>

      <t>The level of MUP integration for mobile service systems 
      running based on the existing mobile service architecture will be varied 
      and depending on the level of MUP awareness of the control 
      and user plane entities.</t>

      <t>Specifying how to modify the existing architecture to 
      integrate MUP is out of scope of this document. 
      What this document provides for the existing architecture 
      is an interface for MUP which the existing or future 
      architectures can easily integrate.</t>


      <section numbered="true" toc="default">
        <name>Requirements Language</name>
        <t>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 <xref target="RFC2119" format="default">RFC 2119</xref>.</t>
      </section>
    </section>
    <section anchor="term" numbered="true" toc="default">
      <name>Terminology</name>
      <dl newline="false" spacing="normal" indent="8">
        <dt>MUP:</dt><dd>Mobile User Plane</dd>
        <dt>MUP Segment:</dt><dd>Representation of mobile user plane segment</dd>
        <dt>MUP PE:</dt><dd>MUP aware Provider Edge node</dd>
        <dt>MUP Controller:</dt><dd>Controller node for MUP networks</dd>
        <dt>UE:</dt><dd>User Equipment, as per <xref target="TS.23501"/></dd>
        <dt>MN:</dt><dd>Mobile Node, as per <xref target="RFC5213"/></dd>
        <dt>Serving operator:</dt><dd>Operator of the visited network
        which serves roaming UEs, as per <xref target="TS.23501"/></dd>
        <dt>Home operator:</dt><dd>Operator of the home network
        to which roaming UEs subscribe, as per <xref target="TS.23501"/></dd>
        <dt>HR roaming:</dt><dd>Home Routed roaming, a roaming scenario
        in which the traffic of a roaming UE is routed to the home
        operator network, as per <xref target="TS.23501"/></dd>
      </dl>
    </section>
    <section anchor="principle" numbered="true" toc="default">
      <name>Architecture Principles</name>
        <t>This section describes the principles for MUP architecture that guide
        its design and operation.</t>

        <t>The first key principle is the abstraction of the mobile 
        user plane. A network segment consisting of a mobile service is 
        abstracted and represented as a MUP segment. 
        It is noted that MUP segment described in this document is 
        NOT Segment Routing<xref target="RFC8402"/> specific, 
        as "segment" is widely common terminology through 
        many networking technologies, and is defined in each technology
        context.</t>

        <t>A MUP PE may accommodate MUP segment(s), such as
        an Interwork Segment and/or a Direct Segment described in
        <xref target="mup_seg"/>. 
        <xref target="fig_overview"/> depicts the overview. 
</t>

        <figure anchor="fig_overview" title="Overview of MUP Architecture">
          <artwork align="center" type="" alt=""><![CDATA[


                        *   Mobility   *
                         * Management *
                          *  System  *
                           *........*
                                |
                       Session Information
                                |
                  ______________v______________
   _______       /           |MUP-C|           \       _______
  /       \     /            +-----+            \     /       \
 /Interwork\__  |                               |  __/ Direct  \
 \ Segment /  \ |-------+                +------| /  \ Segment /
  \_______/    \| MUP PE|       MUP      |MUP PE|/    \_______/
   _______     /|-------+     Network    +------|\     _______
  /       \   / |                               | \   /       \
 / Direct  \_/   \                              /  \_/Interwork\
 \ Segment /      \____________________________/     \ Segment /
  \_______/                                           \_______/

          ]]></artwork>
        </figure>

        <t>The second principle is auto-discovery for MUP segment.
        A MUP PE should be able to discover a MUP segment 
        in a remote MUP PE. In MUP architecture, the remote MUP PE 
        should advertise an auto-discovery route for a hosted MUP 
        segment. The MUP PE can discover the MUP segment 
        in the remote MUP PE when the MUP PE finds  the MUP segment 
        information in the received auto-discovery route 
        from the remote MUP PE.</t>

        <t><xref target="A-D_route"/> in this document defines 
        auto-discovery route for each type of MUP segment discovery.</t>

        <t>It is noted that the auto-discovery route must be 
        independent from a specific dataplane. 
        But with an attribute for the specific dataplane, 
        the auto-discovery route indicates specific dataplane 
        behavior required for the MUP segment.</t>

        <t>The third principle is that transforming 
        session information to routing information.
        Assuming session information for a UE or a MN 
        includes a pair of endpoints between the entities 
        of the mobile user plane,        
        a MUP Controller (MUP-C) advertises MUP PE(s) 
        routing information for the UE or the MN, transformed 
        from the input of corresponding session information 
        in mobile service systems. In MUP architecture,
        it is called session-transformed route.</t>

        <t><xref target="ST_route"/> in this document defines 
        each type of session-transformed route.
        The session-transformed route must also be dataplane
        independent.</t>

        <t>A MUP PE should resolve reachability for a received 
        session-transformed route (ST route for short). 
        When the MUP PE succeeds to resolve the ST route
        reachability with a MUP segment in local, or in remote 
        via an auto-discovery route, and an appropriate dataplane
        behavior indicated in it for the ST route, the MUP PE can 
        perform the dataplane action to the corresponding packet 
        received from a local MUP segment.</t>
        
        <t>The MUP PE sends the packet toward the egress MUP segment
        after the dataplane action applied to the packet. 
        The egress MUP segment may exist in local, or in a
        remote MUP PE. In latter case, the remote MUP PE applies the 
        dataplane action indicated by the received packet, 
        and sends it out to the egress MUP segment.</t>

        <t>The illustrations are described
        in <xref target="illustration"/>.</t>
        <t> To carry these new routing information, this architecture 
        requires extending the existing routing protocols. 
        Any routing protocol can be used to carry this information 
        but this document recommends using BGP. 
        Thus, this document describes extensions on BGP as an example.</t>


    </section>
    <section anchor="mup_seg" numbered="true" toc="default">
      <name>Mobile User Plane Segment</name>
      <t>This document defines two types of Mobile User Plane (MUP) 
      segment. A MUP segment represents a network segment 
      consisting of a mobile service. The MUP segment can be created by 
      a MUP PE which provides connectivity for the mobile user plane.</t>

      <t>Direct Segment is a type of MUP segment that provides 
      connectivity between MUP segments through the MUP networks. 
      Interwork Segment is another type of MUP segment. It
      provides connectivity between a user plane protocol of existing
      or future mobile service architecture and other MUP segments
      through the MUP networks.
      For example in 3GPP 5G case, an Interwork Segment may
      accommodate an N3 network for RANs, or an N9 network
      interconnecting with a home operator network in a home routed
      roaming case.</t>

      <t>A MUP PE may be instantiated as a physical node or a virtual node.
      The MUP PE may also be instantiated on a device which accommodates 
      a mobile user plane node of a mobile service system.</t>

    <section numbered="true" toc="default">
      <!-- <name>IPv6 Dataplane</name> -->
      <name>Dataplane Considerations</name>
      <t>As in <xref target="intro"/>, this document 
      describes IPv6 dataplane in Segment Routing case due to the 
      DMM requirement, and as a suitable solution for scalable 
      mobile service deployments.</t>

      <t>When SRv6 is adopted as the dataplane, an SRv6 SID 
      (Segment Identifier) can represent a MUP segment.
      The SID can be any behavior defined in <xref target="RFC8986"/>, 
      <xref target="RFC9433"/>, 
      or any other extensions for further use cases.
      The behavior of the MUP segment will be chosen by the role 
      of the representing MUP segment. </t>

      <t>For example, in case of a MUP PE interfaces to 5G user plane 
      on the access side defined as "N3" in <xref target="TS.23501"/>, 
      the MUP PE accommodates the N3 network as Interwork Segment in a routing
      instance and then the behavior of created segment SID by the MUP PE
      will be "End.M.GTP4.E", or "End.M.GTP6.E".
      In this case, the MUP PE may associate the SID to
      the routing instance for the N3 access network (N3RAN).</t>

      <t>Another example here is that a MUP PE interfaces to 5G DN on the 
      core side defined as "N6" in <xref target="TS.23501"/>, the MUP PE
      accommodates the N6 network in a routing instance as Direct Segment
      and then the behavior of the created segment SID by the MUP PE will be
      "End.DT4", "End.DT6", or "End.DT2".
      In this case, the MUP PE may associate the SID to
      the routing instance for the N6 data network (N6DN).</t>
    </section>
    <!--
    <section numbered="true" toc="default">
      <name>MPLS Dataplane</name>
      <t>An MPLS Label can represent a MUP segment as a SID. 
      Since it is an MPLS Label, non-SR capable MUP node can use 
      the label to send packets to the segment.</t>

      <t>For example when an MPLS dataplane PE interfaces to 5G user plane 
      on the access side defined as "N3" in <xref target="TS.23501"/>, 
      an Interwork Segment prefix is a gNB host address. The PE allocates
      an MPLS Label to the address and then set the instruction for the 
      MPLS Label to resolve the gNB address to reproduce the 5G user plane
      protocol header. Specifying that instruction is out of scope of this 
      document and it will be specified in another document.</t>

      <t>It is noted that the Interwork Segment prefix length restriction 
      for the MPLS dataplane could be mitigated or removed if 
      an <xref target="I-D.ietf-mpls-mna-fwk">
      MPLS Network Action (MNA)</xref> capable datplane is available and
      it provides similar functionalities with 
      <xref target="RFC9433"/> in the future.</t>

      <t>When an MPLS dataplane PE interfaces to 5G DN on the 
      core side defined as "N6" in <xref target="TS.23501"/>, the PE
      accommodates the N6 network in a routing instance as Direct Segment
      as same as the IPv6 dataplane case, but allocate an MPLS Label to it.
      The instruction of the MPLS Label depends on label allocation mode
      for the instance whether it is table lookup or cross-connect.
      In this case, the PE may associate the MPLS Label to lookup the table of
      the routing instance for the N6 data network (N6DN).</t>
      
    </section>
    -->
    </section>

   <section anchor="A-D_route" numbered="true" toc="default">
      <name>Distribution of Mobile User Plane Segment Information</name>
        <t>Distribution of MUP segment information can be done by 
        advertising routing information with the MUP segment for 
        mobile service. A MUP PE distributes MUP segment information when
        a MUP segment is connected to the MUP PE.</t>

        <t>A MUP Segment Discovery route is routing information
        that associates the MUP segment with network reachability, 
        and assistant metadata if applicable. 
        This document defines the basic discovery route types, 
        Direct Segment Discovery (DSD for short) route, 
        and Interwork Segment Discovery (ISD for short) route.</t>

        <t>Other types of segment discovery route may be mobile service
        architecture specific. 
        Defining the architecture specific network reachability is 
        out of scope of this document and it will be specified 
        in another document.</t>

        <t>To carry the assistant metadata for the MUP Segment Discovery
        route, this document defines MUP extended community. 
        When a metadata applicable for a set of MUP Segment, 
        a MUP extended community carries the metadata in the corresponding
        MUP Segment Discovery routes. The MUP extended community
        must be structured to indicate types of MUP segment specific metadata.
        <xref target="DSD"/> describes Direct Segment type
        MUP extended community, and <xref target="ISD"/> describes
        Interwork Segment type MUP extended community.
        They are illustrated in
        <xref target="illustration"/>.</t>

        <t>MUP Segment Discovery route MUST be used 
        only for resolving reachability for the ST routes. 
        The connectivity among the routing instances for 
        MUP Segments may be advertised as VPN routes. This is to
        avoid forwarding entries to the prefixes of the MUP Segment 
        mingled in the other type of routing instance. </t>

        <t>A MUP PE may discard the received MUP Segment Discovery route 
        if the Route Target extended communities of the route 
        does not meet the MUP PE's import policy.</t>

      <section anchor="DSD" numbered="true" toc="default">
        <name>Direct Segment Discovery (DSD) Route</name>
        <t>When a MUP PE accommodates a network, a service, 
        or an any type of resource through an interface
        or a routing instance as a Direct Segment,
        the MUP PE advertises the corresponding Direct Segment Discovery (DSD)
        route for the interface or the routing instance to the SR domain. 
        The DSD route includes an address to indicate the Direct Segment 
        in the MUP PE in the network layer reachability 
        information (NLRI) with an extended community indicating the 
        corresponding Direct Segment. Dataplane specific attribute 
        should be attached to the DSD route, as a auto-discovery route, 
        which itself must be dataplane independent defined in
         <xref target="principle"/>.</t>

        <t>For example in 3GPP 5G specific case with SRv6 dataplane, 
        an MUP PE may connect to N6 interface on a DN side, and
        a DSD route for the DN will be advertised with an address of
        the MUP PE in NLRI, an corresponding SRv6 SID in the SRv6 
        specific attribute, and a Direct Segment type MUP extended community
        to the routing instance for the DN from the MUP PE.</t>

        <t>When a MUP PE receives a DSD route from other PEs,
        the MUP PE keeps the received DSD route in the RIB.
        The MUP PE uses the received DSD route to resolve Type 2
        Session Transformed (ST2 for short) routes, described in
        <xref target="ST2"/>.</t>

        <t>The ST2 route is resolved with the DSD route which has the
        matching Direct Segment extended community, and the MUP PE
        updates the FIB entry for the ST2 route with the SID of the
        matched DSD route.
        When multiple DSD routes have the matching Direct Segment
        extended community, the DSD route to resolve the ST2 route
        is selected by the BGP best path selection.</t>

        <t>The DSD routes MUST NOT be used to resolve the ST2 routes
        which do not have a Direct Segment extended community:
        such ST2 routes are resolved with the ISD routes as
        described in <xref target="ST2"/>.</t>


      </section>
      <section anchor="ISD" numbered="true" toc="default">
        <name>Interwork Segment Discovery (ISD) Route</name>
        <t>When a PE accommodates a network through an interface
        or a routing instance for the user plane protocol of the mobile
        service architecture as an Interwork Segment,
        the PE advertises the corresponding Interwork Segment Discovery (ISD) 
        route with the prefixes of the Interwork Segment. 
        Dataplane specific attribute should be attached to the ISD route, 
        as a auto-discovery route, which itself must be dataplane independent 
        defined in <xref target="principle"/>. </t>

        <!--
        <t>For MPLS dataplane, the prefix length 
        of the Interwork Segment MUST be full length of IPv4 (32) or 
        IPv6 (128), due to exact match forwarding paradigm of the 
        MPLS dataplane. </t>
        -->

        <t>For example in 3GPP 5G specific case with SRv6 dataplane, 
        an MUP PE may connect to N3 network accommodating a RAN, and
        a ISD route will be advertised with IP prefix(es) of the N3 RAN
        in NLRI, and an corresponding SRv6 SID in the SRv6 specific
        attribute.</t>

        <t>When the MUP networks accommodate multiple Interwork Segments
        whose routing and addressing policies need to be separated
        from each other, such as N9 networks interconnecting with
        different home operators in a home routed roaming case
        described in <xref target="hr_roaming"/>,
        the MUP PE advertises the ISD route with an Interwork Segment
        type MUP extended community which identifies the Interwork
        Segment that the ISD route represents.</t>

        <t>An Interwork Segment type extended community value identifies
        a single Interwork Segment which consists of one or more
        networks under a consistent routing and addressing policy:
        the ISD routes for the networks of the Interwork Segment
        share the community value, and the value MUST NOT identify
        more than one Interwork Segment within the MUP networks.</t>

        <t>An ISD route without an Interwork Segment type MUP extended
        community is implicitly treated as an ISD route for the
        default Interwork Segment.
        The Route Target based import policy SHOULD be used to keep
        only the ISD routes of the intended networks available for
        the default Interwork Segment, so that the ISD routes of the
        policy-separated Interwork Segments do not fall into the
        default Interwork Segment when the community is missing by
        a misconfiguration.</t>

        <t>When a MUP PE receives a ISD route, the MUP PE keeps the 
        received ISD routes in the RIB. 
        The MUP PE uses the received ISD routes to resolve the reachability 
        for remote endpoint of Type 1 Session Transformed (ST1 for short) 
        routes and for the destination endpoint of ST2 routes,
        according to the resolution rules described in
        <xref target="ST1"/> and <xref target="ST2"/> respectively.
        If the ISD route resolves the reachability for the ST routes,
        the MUP PE updates the FIB entries for the prefixes of the
        ST routes with the SID of the matched ISD route.</t>


      </section>
    </section>
    <section anchor="ST_route" numbered="true" toc="default">
      <name>Distribution of Session Transformed Route</name>
      <t>MUP architecture defines two types of session 
      transformed route.</t>

      <section anchor="ST1" numbered="true" toc="default">
        <name>Type 1 Session Transformed (ST1) Route</name>
        <t>First type route, called Type 1 Session Transformed (ST1) route,
        encodes IP prefix(es) for a UE or MN in a BGP
        MP-NLRI attribute with associated session information of the
        tunnel endpoint identifier on the access side. 
        The MUP-C advertises the ST1 
        route with the Route Target extended communities for the UE or MN 
        to the MUP networks.</t>

        <t>A MUP PE may receive the ST1 routes from the MUP-C
        in the MUP networks. The MUP PE may keep the received ST1
        routes advertised from the MUP-C. The receiving MUP PE will perform the
        importing of the received ST1 routes in
        the configured routing instances based on the Route Target
        extended communities. A MUP PE may discard the received ST1
        route if the MUP PE fails
        to import the route based on the Route Target extended communities.</t>

        <t>The MUP PE resolves the reachability for the remote
        endpoint in the session information of the ST1 route with
        the received ISD routes.
        The Interwork Segment extended community is not required
        for the ST1 route resolution.</t>

        <t>When the MUP networks accommodate the policy-separated
        Interwork Segments whose prefixes may overlap each other,
        the Route Target based import policy for the ISD routes
        SHOULD be used to keep only the ISD routes of the intended
        Interwork Segments available for the ST1 route resolution.</t>

        <t>The matched packets are carried to the remote endpoint by
        the dataplane behavior indicated in the ISD route which
        resolves the ST1 route, together with the session information
        of the ST1 route.</t>

        <t>For example in a 3GPP 5G case with SRv6 dataplane, the
        remote endpoint will be the F-TEID of the gNB on the N3
        interface, and the downlink packets toward the UE are
        encapsulated into GTP-U toward the gNB according to the SID
        behaviors of the MUP PEs, as illustrated in
        <xref target="deploy2"/>.</t>

        <t>When a MUP PE fails to resolve the reachability for an
        imported ST1 route, the MUP PE does not install the
        forwarding entry for the ST1 route, and the packets matched
        to the prefix of the ST1 route follow the normal IP routing
        of the routing instance.</t>
      </section>

      <section anchor="ST2" numbered="true" toc="default">
        <name>Type 2 Session Transformed (ST2) Route</name>
        <t>Second type route, called Type 2 Session Transformed (ST2) route,
        encodes the tunnel endpoint identifier of
        the session in a BGP MP-NLRI attribute
        with the nature of tunnel endpoint base packet handling.
        The encoded tunnel endpoint identifier is the key to match
        the received packets to the ST2 route, and is called the
        matching endpoint of the ST2 route in this document.
        Longest match algorithm for the prefix in this type of session
        transformed route should be applicable to aggregate the routes
        for scale.</t>

        <t>The MUP-C may attach a MUP extended community to the ST2
        route, in addition to the Route
        Target extended communities, to indicate the corresponding
        MUP segment and the packet handling for the segment.</t>

        <t>When the MUP-C advertises the ST2 route with a Direct
        Segment extended community, it indicates the corresponding
        Direct Segment and tunnel decapsulation. The packets matched
        to the ST2 route are decapsulated and looked up in the
        corresponding Direct Segment.
        Such an ST2 route is resolved with the DSD routes as
        described in <xref target="DSD"/>.</t>

        <t>When the MUP-C advertises the ST2 route with an Interwork
        Segment extended community, it indicates the destination
        Interwork Segment for the packets matched to the ST2 route.
        The ST2 route without a MUP extended community indicates the
        default Interwork Segment as the destination Interwork
        Segment.</t>

        <t>The tunnel endpoint to which the matched packets are sent on
        the destination Interwork Segment is called the destination
        endpoint of the ST2 route in this document.
        The destination endpoint is determined as follows:</t>

        <ul>
          <li>When the ST2 route includes the session information of
          a tunnel endpoint identifier on the destination Interwork
          Segment, the endpoint in the session information is the
          destination endpoint. The MUP-C attaches the session
          information when the matched packets need to be translated
          toward a tunnel endpoint different from the matching
          endpoint.</li>
          <li>Otherwise, the matching endpoint itself is the
          destination endpoint.</li>
        </ul>

        <t>A MUP PE may receive the ST2 routes
        from the MUP-C in the MUP networks.
        The MUP PE may keep the received ST2 routes advertised from
        the MUP-C. The receiving MUP PE will perform the
        importing of the received ST2 routes in
        the configured routing instances based on the Route Target
        extended communities. A MUP PE may
        discard the received ST2 route if the MUP PE fails
        to import the route based on the Route Target extended communities.</t>

        <t>When the ST2 route indicates a destination Interwork
        Segment, the MUP PE resolves the reachability for the
        destination endpoint of the ST2 route with the received ISD
        routes.
        The ST2 route MUST be resolved only with the ISD routes of
        the destination Interwork Segment: the ISD routes with the
        matching Interwork Segment extended community, or the ISD
        routes for the default Interwork Segment when the ST2 route
        has no MUP extended community.
        This allows the MUP PE to resolve the ST2 route with the
        intended Interwork Segment under the separated routing and
        addressing policies, even when the prefixes of the Interwork
        Segments overlap among them.</t>

        <t>The matched packets are carried to the destination
        endpoint by the dataplane behavior indicated in the ISD route
        which resolves the ST2 route, together with the attached
        session information if present.</t>

        <t>For example in a 3GPP 5G home routed roaming case with
        SRv6 dataplane, the destination endpoint will be the F-TEID
        of the UPF in the home operator network on the N9 interface,
        and the packets are carried over the SRv6 dataplane according
        to the SID behaviors of the MUP PEs, as illustrated in
        <xref target="hr_roaming"/>.</t>

        <t>When a MUP PE fails to resolve the reachability for an
        imported ST2 route, the MUP PE does not install the
        forwarding entry for the ST2 route, and the packets matched
        to the prefix of the ST2 route follow the normal IP routing
        of the routing instance.</t>
      </section>


      <section anchor="MUP-C" numbered="true" toc="default">
        <name>MUP Controller</name>
        <t>A MUP Controller (MUP-C) retrieves or receives session information
        for a UE or a MN from mobile service system.
        The MUP-C transforms 
        the session information to routing information and 
        will advertise the session transformed routes with the corresponding
        extended communities to the MUP networks.</t>

        <t>The session information is expected to include 
        the UE or MN IP prefix(es), tunnel endpoint identifiers for 
        both ends, and any other attributes for the mobile networks. 
        For example in a 3GPP 5G specific case, 
        the tunnel endpoint identifier will be a pair of the F-TEIDs on 
        both the N3 access side (RAN) and core side (UPF).</t>

        <t>As the MUP architecture is independent 
        from specific mobile service systems,
        the MUP-C may leverage any available interfaces or APIs 
        on the mobile service systems for the session information 
        to be received. Each mobile service system specific interface/API
        of MUP-C is out of scope of this document and 
        other documents may describe that specific interface/API.</t>
      </section>
    </section>


    <section anchor="illustration" numbered="true" toc="default">
      <name>Illustration</name>

      <t>This section illustrates possible MUP deployments with SRv6 dataplane.
      3GPP 5G is an example mobile service for the deployment cases 
      in this section. </t>

      <section anchor="deploy1" numbered="true" toc="default">
        <name>SR Network Accommodating Existing Mobile Network Services</name>

        <t><xref target="fig_deploy1"/> shows how SR networks can accommodate 
        existing mobile network service before enabling MUP.
        The PEs S1, S2, and S3 compose an SR network. 
        A routing instance is configured to each network of the mobile 
        service. N6DN in S1 and S2 are providing connectivity 
        to edge servers and the Internet respectively.</t>

        <t>VRF (Virtual Routing Forwarding) is the routing instance 
        to accommodate MUP segments in this section.
        All example cases in this section follow the typical routing policy 
        control using the BGP extended community described 
        in <xref target="RFC4360"/> and <xref target="RFC4684"/></t>

        <figure anchor="fig_deploy1" title="">
          <artwork align="center" type="" alt=""><![CDATA[
   __ N3   /-----------+-----+------------\
  /  \RAN /            |MUP-C|             \
 / V/v\_  |            +-----+             | N6   __
 \    / \ |----+                      +----| DN  /  \
  \__/   \| S1 |                      | S2 |----/W/w \
   __    /|----+                      +----|    \    /
  /  \__/ |             +----+             |     \__/
 / E/e\N6 \             | S3 |             /
 \    /DN  \------------+----+------------/
  \__/             N3UPF  /\ N6UPF
                     X/x /  \ Y/y
                       +-----+
                       | UPF |
                       +-----+
            ]]></artwork>
        </figure>

        <t>The following routing instances are configured:</t>

        <ul>
          <li>N3RAN in S1</li>
          <li>
              <ul>
                <li>export route V/v with route-target (RT) community C1</li>
                <li>import routes which have route-target (RT) community C1 and C2</li>
              </ul>
          </li>
        </ul>

        <ul>
          <li>N6DN in S1</li>
          <li>
            <ul>
              <li>export route E/e with RT C4</li>
              <li>import routes which have RT C3 and C4</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N6DN in S2</li>
          <li>
            <ul>
              <li>export route W/w with RT C4</li>
              <li>import routes which have RT C3 and C4</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N3UPF in S3</li>
          <li>
            <ul>
              <li>export route X/x with RT C2</li>
              <li>import routes which have RT C1</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N6UPF in S3</li>
          <li>
            <ul>
              <li>export route Y/y with RT C3</li>
              <li>import routes which have RT C4</li>
            </ul>
          </li>
        </ul>


        <dl newline="false" spacing="normal" indent="6">
          <dt>Note: </dt>
          <dd>The above configurations are just to provide typical IP 
          connectivity for 3GPP 5G.
          When the above configurations have been done, 
          each endpoint in V/v and X/x can communicate through S1 
          and S3, but they can not communicate with nodes in 
          E/e, W/w and Y/y.</dd>
        </dl>

      </section>

      <section anchor="deploy2" numbered="true" toc="default">
        <name>MUP PE Deployment at All SR Domain Edges</name>

        <t>Here, the PEs S1 and S2 are configured to enable MUP as follows:</t>

        <ul>
          <li>S1</li>
          <li>
            <ul>
              <li>advertises Interwork type discovery route: V/v with SID S1::</li>
              <li>set S1:: behavior End.M.GTP4.E or End.M.GTP6.E</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S1</li>
          <li>
            <ul>
              <li>advertise Direct type discovery route: MUP Direct Segment community D1 and SID S1:1::</li>
              <li>set S1:1:: behavior End.DT4 or End.DT6 for the N6DN in S1</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S2</li>
          <li>
            <ul>
              <li>advertise Direct type route: MUP Direct Segment community D1 and SID S2::</li>
              <li>set S2:: behavior End.DT4 or End.DT6 for the N6DN in S2</li>
            </ul>
          </li>
        </ul>

        <t>S3 does not enable MUP and serves the conventional user
        plane path as an L3VPN PE.</t>

        <t>S1 adopts the local N6DN to prioritize the
        closer segment for the same Direct Segment. Another PE 
        may adopt D1 from S2, if the PE has no local N6DN for D1 and closer
        to S2 than S1.</t>

        <figure anchor="fig_deploy2" title="">
          <artwork align="center" type="" alt=""><![CDATA[
                         U1
                          |
U/u                       v
  \__ N3   /-----------+-----+------------\
  /  \RAN /            |MUP-C|             \
 / V/v\_  |            +-----+             | N6   __
 \    / \ |----+                      +----| DN  /  \
  \__/   \| S1 |                      | S2 |----/W/w \
   __    /|----+                      +----|    \    /
  /  \__/ |             +----+             |     \__/
 / E/e\N6 \             | S3 |             /
 \    /DN  \------------+----+------------/
  \__/             N3UPF  /\ N6UPF
                     X/x /  \ Y/y
                       +-----+
                       | UPF |
                       +-----+
            ]]></artwork>
        </figure>

        <t>Now, session information U1 is put to a MUP Controller,
        MUP-C, and MUP-C is configured to transform U1
        to the routes as follows:</t>

        <ul>
          <li>MUP-C</li>
          <li>
            <ul>
              <li>attach the MUP Direct Segment ID D1 and RT C3 to the DN in U1</li>
              <li>transforms UE's prefix U/u, the F-TEID on access side (gNB) and QFI in U1 to the ST1 route 
              for the prefix U/u with the F-TEID, the QFI, and RT C3</li>
              <li>transforms F-TEID on core side (UPF) X in U1 
              to the ST2 route for X with MUP segment-ID D1 and RT C2</li>
            </ul>
          </li>
        </ul>

        <t>Then N3RAN and N6DN import route X and U/u respectively.
        S1 and S2 resolves U/u's remote endpoint with V/v and then install SID S1:: 
        for U/u in FIB. S1:: will not appear in the packet from E/e to 
        U/u over the wire.</t>

        <t>As S1 adopts local N6DN for D1, N3RAN in S1 decapsulates GTP-U 
        packets from V/v to X and then lookup the inner packets from U/u in N6DN
        after the decapsulation.
        </t>

        <dl newline="false" spacing="normal" indent="6">
          <dt>Note: </dt>
          <dd>When the above configurations have been done, 
          MUP is applied only to the packets from/to U/u. 
          Each endpoint in U/u, W/w and E/e can communicate through S1 and S2.
          The rest of traffic from/to other UEs go through the usual 3GPP
          5G user plane path using UPF via S3.</dd>
        </dl>

      </section>

      <section anchor="deploy3" numbered="true" toc="default">
        <name>Adding Direct Segment with New MUP PE </name>

        <t>Another case shown in <xref target="fig_deploy3"/> is that 
        S4 joins the SR network and accommodates edge servers in 
        the N6DN in S4.</t>

        <figure anchor="fig_deploy3" title="">
          <artwork align="center" type="" alt=""><![CDATA[
                         U1
                          |
U/u                       v                       __
  \__ N3   /-----------+-----+------------\      /  \
  /  \RAN /            |MUP-C|             \  __/W/w \
 / V/v\_  |            +-----+        +----|_/N6\    /
 \    / \ |----+                      | S2 |  DN \__/
  \__/   \| S1 |                      +----|      __
   __    /|----+                      +----|_    /  \
  /  \__/ |             +----+        | S4 | \__/E/e \
 /    \N6 \             | S3 |        +----/  N6\    /
 \    /DN  \------------+----+------------/   DN \__/
  \__/             N3UPF  /\ N6UPF
                     X/x /  \ Y/y
                       +-----+
                       | UPF |
                       +-----+
            ]]></artwork>
        </figure>

        <t>The following routing instances are configured:</t>

        <ul>
          <li>N3RAN in S1 (same with the previous case)</li>
          <li>
              <ul>
                <li>export route V/v with RT C1</li>
                <li>import routes which have RT C1 and C2</li>
              </ul>
          </li>
        </ul>

        <ul>
          <li>N6DN in S1</li>
          <li>
            <ul>
              <li>export no route</li>
              <li>import routes which have RT C4</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N6DN in S2 (same with the previous case)</li>
          <li>
            <ul>
              <li>export route W/w with RT C4</li>
              <li>import routes which have RT C3 and C4</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N3UPF in S3 (same with the previous case)</li>
          <li>
            <ul>
              <li>export route X/x with RT C2</li>
              <li>import routes which have RT C1</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N6UPF in S3 (same with the previous case)</li>
          <li>
            <ul>
              <li>export route Y/y with RT C3</li>
              <li>import routes which have RT C4</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N6DN in S4</li>
          <li>
            <ul>
              <li>export route E/e with RT C4</li>
              <li>import routes which have RT C3 and C4</li>
            </ul>
          </li>
        </ul>

        <t>Here, the PEs are configured to enable MUP as following:</t>

        <ul>
          <li>S1 (same with the previous case)</li>
          <li>
            <ul>
              <li>advertises Interwork type route: V/v with SID S1::</li>
              <li>set S1:: behavior End.M.GTP4.E or End.M.GTP6.E</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S1</li>
          <li>
            <ul>
              <li>advertise Direct type route: MUP Direct Segment community D1 for the local N6DN</li>
              <li>set S1:1:: behavior End.DT4 or End.DT6 for the N6DN in S1</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S2 (same with the previous case)</li>
          <li>
            <ul>
              <li>advertise Direct type route: MUP Direct Segment community D1 and SID S2::</li>
              <li>set S2:: behavior End.DT4 or End.DT6 for the N6DN in S2</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S4</li>
          <li>
            <ul>
              <li>advertise Direct type route: MUP Direct Segment community D2 and SID S4::</li>
              <li>set S4:: behavior End.DT4 or End.DT6 for the N6DN in S4</li>
            </ul>
          </li>
        </ul>

        <t>As in the previous case, S3 does not enable MUP and serves
        the conventional user plane path as an L3VPN PE.</t>

        <t>As in the previous case, S1 adopts the local N6DN for D1 as long as S1 prioritizes
        the closer segment for the same MUP Direct Segment. The Direct type route from S4 for D2 
        with SID S4:: will be kept in S1.</t>

        <ul>
          <li>MUP-C (same with the previous case)</li>
          <li>
            <ul>
              <li>attach the MUP Direct Segment ID D1 and RT C3 to the DN in U1</li>
              <li>transforms UE's prefix U/u, the F-TEID on access side (gNB) and QFI in U1 to the ST1 route 
              for the prefix U/u with the F-TEID, the QFI, and RT C3</li>
              <li>transforms F-TEID on core side (UPF) X in U1 
              to the ST2 route for X with MUP Direct Segment community D1 and RT C2</li>
            </ul>
          </li>
        </ul>

        <t>Then N3RAN and N6DN import route X and U/u respectively.
        S2 and S4 resolve U/u's remote endpoint with V/v and then install SID S1:: 
        for U/u in FIB.</t>

        <t>As in the previous case, S1 adopts local N6DN for D1, 
        N3RAN in S1 decapsulates GTP-U packets from V/v to X and then 
        lookup the inner packets from U/u in N6DN after the decapsulation.
        </t>
        <t>For D2 on the other hand, no corresponding N6DN existed in S1. 
        However, E/e with RT C4 from S4 is imported into N6DN in S1 as a 
        VPN route, E/e is reachable from U/u via N6DN for D1 in S1.</t>

        <t>If a session U1' includes the DN corresponding to D2, MUP-C advertises
        ST2 route X' with MUP Direct Segment community D2, and then
        N3RAN in S1 instantiates H.M.GTP4.D or End.M.GTP6.D for X with S4::
        as the last SID in the received Direct type route from S4.
        </t>

        <dl newline="false" spacing="normal" indent="6">
          <dt>Note: </dt>
          <dd>When the above configurations have been done, 
          MUP is applied only to the packets from/to U/u. 
          Each endpoint in U/u, W/w and E/e can communicate through S1, S2 
          and S4. 
          The rest of traffic from/to other UEs go through the usual 3GPP
          5G user plane path using UPF via S3.</dd>
        </dl>

      </section>

      <section anchor="collupsed" numbered="true" toc="default">
        <name>Collapsed MUP PE Deployment</name>

        <t>In this case only S1 enables MUP in a collapsed fashion.
        S2 and S3 are L3VPN PEs without MUP capability. In this section,
        S2 and S3 are illustrated as SRv6 nodes. But they can be 
        non-SR nodes if S1 provides SR independent connectivity to
        S2 and S3.</t>

        <figure anchor="fig_deploy4" title="">
          <artwork align="center" type="" alt=""><![CDATA[
                         U1
                          |
U/u                       v
  \__ N3   /-----------+-----+------------\
  /  \RAN /            |MUP-C|             \
 / V/v\_  |            +-----+             | N6   __
 \    / \ |----+                      +----| DN  /  \
  \__/   \| S1 |                      | S2 |----/W/w \
   __    /|----+                      +----|    \    /
  /  \__/ |             +----+             |     \__/
 / E/e\N6 \             | S3 |             /
 \    /DN  \------------+----+------------/
  \__/             N3UPF  /\ N6UPF
                     X/x /  \ Y/y
                       +-----+
                       | UPF |
                       +-----+
            ]]></artwork>
        </figure>

        <t>The difference between the previous case in <xref target="deploy1"/> 
        for the routing instance configuration is following:</t>

        <ul>
          <li>N6DN in S1</li>
          <li>
            <ul>
              <li>export route E/e with RT C4</li>
              <li>import routes which have RT C3, C4 and C5</li>
            </ul>
          </li>
        </ul>

        <t>Here, S1 is configured to enable MUP and S2 as an L3VPN PE is configured as follows:</t>

        <ul>
          <li>S1</li>
          <li>
            <ul>
              <li>may not advertise Interwork type discovery route for V/v</li>
              <li>may not advertise Direct type discovery route with MUP Direct Segment community D1 and S1:1::</li>
              <li>set S1:1:: behavior End.DT4 or End.DT6 for the N6DN in S1</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S2</li>
          <li>
            <ul>
              <li>set S2:: behavior End.DT4 or End.DT6 for the N6DN in S2</li>
            </ul>
          </li>
        </ul>

        <t>Now, session information U1 is added to the MUP Controller, MUP-C, 
        and MUP-C and S1 is configured to transform U1 to the routes as follows:</t>

        <ul>
          <li>MUP-C</li>
          <li>
            <ul>
              <li>attach the MUP Direct Segment ID D1 and RT C5 to the DN in U1</li>
              <li>transforms UE's prefix U/u, the F-TEID on access side (gNB) and QFI in U1 to the ST1 route 
              for the prefix U/u with the F-TEID, the QFI, and RT C5</li>
              <li>transforms F-TEID on core side (UPF) X in U1 
              to the ST2 route for X with MUP Direct Segment community D1 and RT C2</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S1</li>
          <li>
            <ul>
              <li>advertises U/u as an L3VPN route with RT C4 and SID S1:1::, 
              when the ST1 route is imported into the N6DN</li>
            </ul>
          </li>
        </ul>

        <t>Then the N3RAN and N6DN import route X and U/u respectively.
        S1 resolves U/u's remote endpoint with V/v and then create the corresponding
        GTP encap entry for U/u into the N3RAN FIB.
        S2 will create a regular L3VPN routing entry for U/u with SID S1:1:: in the N6DN
        when S2 imports the L3VPN route with RT C4 for U/u advertised from S1.</t>

        <t>As S1 adopts local N6DN for D1, N3RAN in S1 decapsulates GTP-U 
        packets from V/v to X and then lookup the inner packets from U/u in N6DN
        after the decapsulation.
        </t>

        <dl newline="false" spacing="normal" indent="6">
          <dt>Note: </dt>
          <dd>When the above configurations have been done, 
          MUP is applied only to the packets from/to U/u. 
          Each endpoint in U/u, W/w and E/e can communicate through S1 and S2.
          The rest of traffic from/to other UEs go through the usual 3GPP
          5G user plane path using UPF via S3.</dd>
        </dl>

      </section>

      <section anchor="hr_roaming" numbered="true" toc="default">
        <name>Distributed Home Routed Roaming Interconnection with Multiple Home Operators</name>

        <t>This case shows a home routed (HR) roaming scenario.
        The serving operator enables MUP in the SR network.
        The SR network accommodates the N3 network for RANs as an
        Interwork Segment, and it also accommodates multiple N9
        networks interconnecting with home operators as Interwork
        Segments. The N9 interconnections are distributed over the
        MUP PEs: any MUP PE may accommodate an N9 network, including
        the MUP PE which accommodates the RANs.
        In this scenario, MUP PEs in the serving operator
        network perform the user plane handling for HR roaming
        sessions between the RAN and the home operator networks,
        instead of the user plane specific nodes in the serving
        operator network.</t>

        <figure anchor="fig_hr_roaming" title="">
          <artwork align="center" type="" alt=""><![CDATA[
                         U1
                          |
U/u                       v
  \__ N3   /-----------+-----+------------\
  /  \RAN /            |MUP-C|             \
 / V/v\_  |            +-----+             |          ___
 \    / \ |----+                      +----|  N9A    /   \
  \__/   \| S1 |                      | S2 |--------/HO-A \
   ___   /|----+                      +----|        \ A/a /
  /   \_/ |                                |         \___/
 /HO-B \  |             +----+             |
 \ B/b /N9B\            | S3 |            /
  \___/     \-----------+----+-----------/
                   N3UPF  /\ N9UPF
                     X/x /  \ Z/z
                       +-----+
                       | UPF |
                       +-----+
            ]]></artwork>
        </figure>

        <t><xref target="fig_hr_roaming"/> shows that S1 accommodates
        the N3 RAN and the N9 network N9B which interconnects with
        home operator B (HO-B), and S2 accommodates the N9 network
        N9A which interconnects with home operator A (HO-A).
        The routing and addressing
        of an N9 interconnection are managed under the policy of each
        home operator. N9A and N9B are accommodated in the individual
        routing instances to separate the policies of the home
        operators from each other. The prefixes of the home operator
        UPF N9 endpoints are A/a in HO-A and B/b in HO-B.</t>

        <t>S3 accommodates the UPF of the serving operator as with
        the previous cases. The N3 side of the UPF connects to N3UPF
        in S3, and the N9 side of the UPF connects to N9UPF in S3
        with the prefix Z/z. The HR roaming sessions which are not
        handled by MUP go through the UPF between the N3 network and
        the N9 networks.</t>

        <t>The following routing instances are configured:</t>

        <ul>
          <li>N3RAN in S1 (same with the previous cases)</li>
          <li>
              <ul>
                <li>export route V/v with RT C1</li>
                <li>import routes which have RT C1 and C2</li>
              </ul>
          </li>
        </ul>

        <ul>
          <li>N3UPF in S3 (same with the previous cases)</li>
          <li>
            <ul>
              <li>export route X/x with RT C2</li>
              <li>import routes which have RT C1</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N9UPF in S3 (for the UPF N9 side)</li>
          <li>
            <ul>
              <li>export route Z/z with RT C7</li>
              <li>import routes which have RT C9 and C10</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N9A in S2 (for the N9 network interconnecting with HO-A)</li>
          <li>
            <ul>
              <li>export route A/a with RT C9</li>
              <li>import routes which have RT C5 and C7</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>N9B in S1 (for the N9 network interconnecting with HO-B)</li>
          <li>
            <ul>
              <li>export route B/b with RT C10</li>
              <li>import routes which have RT C6 and C7</li>
            </ul>
          </li>
        </ul>

        <t>The routing instance N9UPF exchanges the VPN routes with
        N9A and N9B for the connectivity between the UPF N9 side and
        the N9 networks.
        The HR roaming traffic through the UPF is carried among
        S3, S2 and S1 as the L3VPN traffic.</t>

        <t>Here, the PEs are configured to enable MUP as following:</t>

        <ul>
          <li>S1</li>
          <li>
            <ul>
              <li>advertises Interwork type route: V/v with SID S1::
              without an Interwork Segment community: the N3RAN is
              the default Interwork Segment</li>
              <li>set S1:: behavior End.M.GTP4.E or End.M.GTP6.E</li>
              <li>advertises Interwork type route: B/b with SID S1:1::
              and MUP Interwork Segment community I2 for N9B</li>
              <li>set S1:1:: behavior End.M.GTP4.E or End.M.GTP6.E for the N9B in S1</li>
            </ul>
          </li>
        </ul>

        <ul>
          <li>S2</li>
          <li>
            <ul>
              <li>advertises Interwork type route: A/a with SID S2:1::
              and MUP Interwork Segment community I1 for N9A</li>
              <li>set S2:1:: behavior End.M.GTP4.E or End.M.GTP6.E for the N9A in S2</li>
            </ul>
          </li>
        </ul>

        <t>The ISD routes for N9A and N9B represent the Interwork
        Segments for
        the N9 networks identified by the Interwork Segment communities
        I1 and I2. The identified Interwork Segments keep the routing
        and addressing policies of the home operators separated in the
        MUP networks. It also allows the home operators to use
        overlapping IP address spaces for their N9 networks.
        S3 does not enable MUP and serves the conventional user plane
        path as an L3VPN PE.</t>

        <t>A session information U1 for a UE subscribing to HO-A is put
        to MUP-C. U1 is a HR roaming session which includes the F-TEID
        V1 on the access side (gNB) in V/v, the F-TEID Xa on the
        serving core side of N3 in X/x, the F-TEID A1 of the HO-A UPF
        on N9A in A/a, and the F-TEID Z1 on the serving side of N9A
        in Z/z.</t>

        <t>The prefix Z/z which covers Z1 is advertised to HO-A through
        the N9A routing, and the reachability to A/a of HO-A is
        available in the N9A routing instance. MUP-C is configured to
        transform U1 to the routes as follows:</t>

        <ul>
          <li>MUP-C</li>
          <li>
            <ul>
              <li>transforms the F-TEID Xa on the serving core side in U1
              to the ST2 route for Xa with MUP Interwork Segment community I1,
              the session information of the tunnel endpoint identifier A1
              on the destination Interwork Segment N9A, and RT C2</li>
              <li>transforms the F-TEID Z1 on the serving N9 side in U1
              to the ST2 route for Z1 without a MUP extended community,
              with the session information of the tunnel endpoint
              identifier V1 on the default Interwork Segment (N3RAN),
              and RT C5</li>
            </ul>
          </li>
        </ul>

        <t>Then N3RAN in S1 imports the ST2 route Xa. Since the ST2
        route Xa has the Interwork Segment community I1, S1 resolves
        the reachability for the destination endpoint A1 in the session
        information of the route only with the ISD routes which have I1, that is
        A/a with SID S2:1::. N3RAN in S1 instantiates H.M.GTP4.D or
        End.M.GTP6.D for Xa with S2:1:: as the last SID in the matched
        ISD route.</t>

        <t>For uplink, the gNB sends GTP-U packets of U1 to Xa.
        S1 decapsulates the GTP-U packets, and encapsulates the inner
        packets into SRv6 with the SID composed from S2:1:: and the
        arguments embedding the tunnel endpoint identifier A1 from
        the session information of the ST2 route.
        S2 performs End.M.GTP4.E or End.M.GTP6.E for the received
        packets to encapsulate them into GTP-U toward A1, and then
        the packets are forwarded to the HO-A UPF through N9A.</t>

        <t>In N3RAN, the ST2 route Xa is preferred to the VPN route X/x
        imported from N3UPF due to the longest match, so that only the
        packets of U1 are handled by MUP.</t>

        <t>N9A in S2 imports the ST2 route Z1. Since the ST2 route Z1
        has no MUP extended community, S2 resolves the
        reachability for the destination endpoint V1 in the session
        information of the route with the ISD routes for the default
        Interwork Segment, that is V/v
        with SID S1::. N9A in S2 instantiates H.M.GTP4.D or
        End.M.GTP6.D for Z1 with S1:: as the last SID in the matched
        ISD route.</t>

        <t>For downlink, the HO-A UPF sends GTP-U packets of U1 to Z1
        through N9A.
        S2 decapsulates the GTP-U packets, and encapsulates the inner
        packets into SRv6 with the SID composed from S1:: and the
        arguments embedding the tunnel endpoint identifier V1 from
        the session information of the ST2 route.
        S1 performs End.M.GTP4.E or End.M.GTP6.E for the received
        packets to encapsulate them into GTP-U toward V1, and then
        the packets are delivered to the gNB.</t>

        <t>In N9A, the ST2 route Z1 is preferred to the VPN route Z/z
        imported from N9UPF due to the longest match, so that only the
        packets of U1 are handled by MUP.</t>

        <t>For a HR roaming session of a UE subscribing to HO-B, MUP-C
        advertises the ST2 routes with the Interwork Segment community
        I2 and RT C6 in the same manner. The ST2 routes are resolved
        only with the ISD route B/b with SID S1:1:: which has I2, and
        the packets of the session are forwarded to N9B under the
        policy of HO-B, independently from the policy of HO-A.
        Since S1 accommodates both the N3 RAN and N9B, the packets
        of the session are handled locally within S1, without
        traversing another MUP PE.</t>

        <t>A home operator may also deploy MUP in its network.
        <xref target="fig_hob_mup"/> shows the case where HO-B
        deploys MUP: S1 accommodates the RAN (V/v) in the routing
        instance N3RAN, and the UE with the prefix U/u attaches to
        the RAN. The MUP PE S4 of HO-B
        connects to S1 through the routing instance N9B, and S4
        accommodates the N6DN (E/e) of HO-B.</t>

        <figure anchor="fig_hob_mup" title="">
          <artwork align="center" type="" alt=""><![CDATA[
U/u     ___                                     ___
  \    /   \  N3RAN +----+  N9B  +----+  N6DN  /   \
   \__/ V/v \-------| S1 |-------| S4 |-------/ E/e \
      \     /       +----+       +----+       \     /
       \___/                                   \___/
            ]]></artwork>
        </figure>

        <t>In this case, the HR roaming sessions of HO-B are handled
        directly between the MUP PEs S1 and S4 over the N9B
        interconnection, and the packets of the sessions reach the
        N6DN in S4 without traversing the user plane specific nodes
        of the operators. This enables a low-latency interconnection
        for HR roaming, for example toward the edge services in E/e
        of the home operator.</t>

        <dl newline="false" spacing="normal" indent="6">
          <dt>Note: </dt>
          <dd>In this scenario, the GTP-U tunnels of a HR roaming
          session on the N3 side and the N9 side are terminated at
          the MUP PEs, and the packets of the session are carried
          over the SRv6 dataplane between S1 and S2 for the HO-A
          sessions, or handled locally within S1 for the HO-B
          sessions, instead of the
          user plane specific nodes on the serving side.
          The Interwork Segment communities
          keep the resolution of the ST2 routes aligned with the
          routing and addressing policy of each home operator,
          even when the N9 networks use overlapping IP address space.
          The rest of the HR roaming traffic from/to other UEs go
          through the usual 3GPP 5G user plane path using the UPF
          via S3.</dd>
        </dl>

      </section>


    </section>


   <section anchor="IANA" numbered="true" toc="default">
      <name>IANA Considerations</name>
      <t>This memo includes no request to IANA.</t>
    </section>
    <section anchor="Security" numbered="true" toc="default">
      <name>Security Considerations</name>
      <t>The MUP architecture defines routing information that is
      transformed from session information of a mobile service system.
      Spoofing or tampering of the transformed routing information or
      the related control messages may redirect mobile user plane
      traffic to unintended segments. Therefore, the routing control
      plane of the MUP networks MUST be protected from such spoofing
      and tampering. The interface between the mobile service system
      and the MUP-C SHOULD also be protected, while each specific
      interface/API is out of scope of this document as described in
      <xref target="MUP-C"/>.</t>

      <t>When BGP is used to distribute the MUP Segment Discovery
      routes and the Session Transformed routes, the deployment SHOULD
      follow the operational and security practices in
      <xref target="RFC7454"/>. In particular, BGP sessions that
      distribute these routes MUST be established only with authorized
      peers. This is especially important when the routes are
      exchanged across AS or administrative domain boundaries.</t>

      <t>The import and export policies of the MUP networks MUST
      ensure that the MUP Segment Discovery routes and the Session
      Transformed routes are propagated only within the intended
      routing instances and only to the intended MUP peers.
      Leakage of these routes may result in unintended access to
      mobile user plane traffic or unintended packet steering.
      For example in the home routed roaming case described in
      <xref target="hr_roaming"/>, the routes for an Interwork
      Segment of a home operator MUST NOT be leaked to the routing
      instances of the other home operators, so that the routing and
      addressing policies of the home operators are kept separated.</t>

      <t>The forwarding states resolved from the Session Transformed
      routes depend on the MUP Segment Discovery routes.
      When a MUP Segment Discovery route which resolves Session
      Transformed routes is withdrawn or its MUP extended community
      changes, the MUP PE SHOULD re-evaluate the resolution of the
      affected Session Transformed routes and remove the forwarding
      entries which no longer have an admissible route, so that
      stale forwarding states do not remain.</t>

      <t>The re-evaluation above does not protect the case where the
      operator of the MUP networks reassigns an Interwork Segment
      extended community value to a different Interwork Segment:
      the Session Transformed routes still carrying the value would
      be resolved with the unintended Interwork Segment, and the
      mobile user plane traffic may be delivered to a network of an
      unintended operator.
      Therefore, the operator SHOULD NOT reassign a community value
      to a different Interwork Segment while Session Transformed
      routes carrying the value remain advertised, and SHOULD take
      the propagation of the withdrawals across the MUP networks
      into account when reassigning the value.</t>

      <t>When SRv6 is used as the dataplane, the MUP deployment MUST
      follow the security considerations of <xref target="RFC8754"/>
      and <xref target="RFC8986"/>.</t>
    </section>

    <!-- End section "Contributors" -->

  </middle>
  <!--  *****BACK MATTER ***** -->

 <back>
    <!-- References split into informative and normative -->

   <!-- There are 2 ways to insert reference entries from the citation libraries:
    1. define an ENTITY at the top, and use "ampersand character"RFC2629; here (as shown)
    2. simply use a PI "less than character"?rfc include="reference.RFC.2119.xml"?> here
       (for I-Ds: include="reference.I-D.narten-iana-considerations-rfc2434bis.xml")

    Both are cited textually in the same manner: by using xref elements.
    If you use the PI option, xml2rfc will, by default, try to find included files in the same
    directory as the including file. You can also define the XML_LIBRARY environment variable
    with a value containing a set of directories to search.  These can be either in the local
    filing system or remote ones accessed by http (http://domain/dir/... ).-->

   <references>
      <name>References</name>
      <references>
        <name>Normative References</name>
        <!--?rfc include="http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml"?-->
            &rfc7333;
            &rfc8402;
            &rfc8754;
            &rfc8986;
            <!-- &I-D.ietf-spring-segment-routing-policy; -->

     <reference anchor="RFC2119" target="https://www.rfc-editor.org/info/rfc2119" xml:base="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/reference.RFC.2119.xml">
          <front>
            <title>Key words for use in RFCs to Indicate Requirement Levels</title>
            <seriesInfo name="DOI" value="10.17487/RFC2119"/>
            <seriesInfo name="RFC" value="2119"/>
            <seriesInfo name="BCP" value="14"/>
            <author initials="S." surname="Bradner" fullname="S. Bradner">
              <organization/>
            </author>
            <date year="1997" month="March"/>
            <abstract>
              <t>In many standards track documents several words are used to signify the requirements in the specification.  These words are often capitalized. This document defines these words as they should be interpreted in IETF documents.  This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements.</t>
            </abstract>
          </front>
        </reference>
      </references>
      <references>
        <name>Informative References</name>
        <!-- Here we use entities that we defined at the beginning. -->
            <!-- &I-D.ietf-mpls-mna-fwk;-->
            &rfc4360;
            &rfc4684;
            &rfc5213;
            &rfc7454;
            &rfc8885;
            &rfc9433;


        <reference anchor="TS.23501" target="http://www.3gpp.org/ftp/Specs/html-info/23501.htm">
            <front>
                <title>System architecture for the 5G System (5GS)</title>
                <author>
                    <organization>3GPP</organization>
                </author>
                <date day="24" month="September" year="2021"/>
            </front>
            <seriesInfo name="3GPP TS" value="23.501 17.2.0" />
        </reference>
      </references>
    </references>

    <!-- Filled after when substantial contributions are made: -->

    <section anchor="Acknowledgements" numbered="false">
      <name>Acknowledgements</name>
      <t>The authors would like to thank Jeffrey Zhang,
      Keiichi Makizono, Hideyuki Sasaki, Teruyuki Oya, 
      Yoichi Funabiki, Daisuke Koshiro, Derek Yeung,
      Kalyani Rajaraman, Csaba Keszei, Daiki Kijima,
      Ryuma Seno, Keiji Hara, Mikio Nakajima, Masayuki Yamazaki,
      Takuto Shiozawa, Manabu Mikami, Takaya Watanabe,
      Hidenori Aita, Hiroyuki Fukumori, Leo Fujita, 
      Toshihiro Yokomine, Toshikazu Morioka, Yadav Rituraji,
      Junya Yoshino, Takayuki Koshikawa,
      Kazunari Nagai, Tadayuki Okuda, Takahiro Hara,
      Yuya Kusakabe, Takeru Hayasaka,
      Hideki Takase, Yutaka Kikuchi, Kazuma Nishiuchi,
      Mitsuhiro Osaki, Suresh Krishnan, Sri Gundavelli,
      Zafar Ali, Luay Jalil, Markus Amend, Marco Liebsch, 
      and Lionel Morand for their reviews and discussions with the authors.</t>
    </section>
    <!-- Possibly a 'Contributors' section ... -->

    <section numbered="false">
      <name>Contributors</name>
      <contact fullname="Clarence Filsfils">
        <organization>Cisco Systems, Inc.</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>Belgium</country>
          </postal>
          <email>cf@cisco.com</email>
        </address>
      </contact>
      <contact fullname="Ketan Talaulikar" initials="K." surname="Talaulikar">
        <organization>Cisco Systems</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>India</country>
          </postal>
          <email>ketant.ietf@gmail.com</email>
        </address>
      </contact>
    </section>

    <section numbered="false">
      <name>Additional Authors</name>

      <contact fullname="Ashiq Khan" initials="A." surname="Khan">
        <organization>SoftBank</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>Japan</country>
          </postal>
          <email>ashiq.khan@g.softbank.co.jp</email>
        </address>
      </contact>
      <contact fullname="Miya Kohno" initials="M." surname="Kohno">
        <organization abbrev="Cisco Systems, Inc.">
              Cisco Systems, Inc.</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>Japan</country>
          </postal>
          <email>mkohno@cisco.com</email>
        </address>
      </contact> 
      <contact fullname="Teppei Kamata" initials="T." surname="Kamata">
        <organization abbrev="Cisco Systems, Inc.">
              Cisco Systems, Inc.</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>Japan</country>
          </postal>
          <email>tkamata@cisco.com</email>
        </address>
      </contact> 
      <contact fullname="Pablo Camarillo Garvia" initials="P." surname="Camarillo">
        <organization abbrev="Cisco Systems, Inc.">
              Cisco Systems, Inc.</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>Spain</country>
          </postal>
          <email>pcamaril@cisco.com</email>
        </address>
      </contact> 
      <contact fullname="Daniel Voyer" initials="D." surname="Voyer">
        <organization abbrev="Bell Canada">Bell Canada</organization>
        <address>
          <postal>
            <street></street>
            <city></city>
            <region></region>
            <code></code>
            <country>Canada</country>
          </postal>
          <email>daniel.voyer@bell.ca</email>
        </address>
      </contact>
      <contact fullname="Shay Zadok" initials="S." surname="Zadok">
        <organization abbrev="Broadcom">
              Broadcom</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>Israel</country>
          </postal>
          <email>shay.zadok@broadcom.com</email>
        </address>
      </contact> 
      <contact fullname="Israel Meilik" initials="I." surname="Meilik">
        <organization abbrev="Broadcom">
              Broadcom</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>Israel</country>
          </postal>
          <email>israel.meilik@broadcom.com</email>
        </address>
      </contact> 
      <contact fullname="Ashutosh Agrawal" initials="A." surname="Agrawal">
        <organization abbrev="Intel">
              Intel</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>USA</country>
          </postal>
          <email>ashutosh.agrawal@intel.com</email>
        </address>
      </contact> 
      <contact fullname="Kumaresh Perumal" initials="K." surname="Perumal">
        <organization abbrev="Intel">
              Intel</organization>
        <address>
          <postal>
            <street/>
            <city/>
            <region/>
            <code/>
            <country>USA</country>
          </postal>
          <email>kumaresh.perumal@intel.com</email>
        </address>
      </contact> 
    </section>

    <!-- Change Log

v00 2021-10-08  Matsushima   Skelton version
    -->
 </back>
</rfc>
