One document matched: draft-bryan-p2psip-app-scenarios-00.xml


<?xml version="1.0" encoding="US-ASCII"?>
<!-- you can convert this to html with xsltproc and rfc2629.slt 
      see  http://xml.resource.org/ and rfc 2629 -->
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
<!ENTITY rfc3261 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3261.xml">
<!ENTITY rfc3263 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3263.xml">
]>

<?xml-stylesheet type="text/xsl" href="rfc2629.xslt"?>
<?rfc toc="yes" ?>
<?rfc symrefs="yes"?>
<?rfc iprnotified="no" ?>
<?rfc strict="yes" ?>
<?rfc compact="yes" ?>
<?rfc sortrefs="yes"?>
<?rfc comments="yes"?>
<rfc category="info" docName="draft-bryan-p2psip-app-scenarios-00" ipr="full3978">
  <front>
    <title abbrev="P2PSIP Application Scenarios">Application Scenarios for Peer-to-Peer Session
    Initiation Protocol (P2PSIP)</title>

    <author fullname="David A. Bryan" initials="D. A." surname="Bryan">
      <organization>SIPeerior Technologies, Inc.</organization>

      <address>
        <postal>
          <street>3000 Easter Circle</street>

          <city>Williamsburg</city>

          <region>VA</region>

          <code>23188</code>

          <country>USA</country>
        </postal>

        <phone>+1 757 565 0101</phone>

        <email>dbryan@sipeerior.com</email>
      </address>
    </author>

    <author fullname="Eunsoo Shim" initials="E." surname="Shim">
      <organization abbrev="Locus Telecom">Locus Telecommunications,
      Inc.</organization>

      <address>
        <postal>
          <street>2200 Fletcher Ave. 6th FL</street>

          <street></street>

          <city>Fort Lee</city>

          <region>NJ</region>

          <code>07024</code>

          <country>USA</country>
        </postal>

        <email>eunsoo@locus.net</email>
      </address>
    </author>

    <author fullname="Bruce B. Lowekamp" initials="B. B." surname="Lowekamp">
      <organization>SIPeerior; William & Mary</organization>

      <address>
        <postal>
          <street>3000 Easter Circle</street>

          <city>Williamsburg</city>

          <region>VA</region>

          <code>23188</code>

          <country>USA</country>
        </postal>

        <phone>+1 757 565 0101</phone>

        <email>lowekamp@sipeerior.com</email>
      </address>
    </author>

  <author initials="S." surname="Dawkins" fullname="Spencer Dawkins" role="editor">
    <organization abbrev="Huawei (USA)">Huawei Technologies (USA)</organization>
    <address>
      <postal>
	  <street> 1547 Rivercrest Blvd.</street>

	  <city> Allen</city> <region>TX </region>

	  <code>75002 </code> <country>USA</country> </postal>

      <phone>+1 214 755 3870</phone>

      <email>spencer@mcsr-labs.org </email>
    </address>
  </author>  

    <date month="November" year="2007" />

    <area>RAI</area>

    <workgroup>P2PSIP</workgroup>

    <keyword>I-D</keyword>

    <keyword>Internet-Draft</keyword>

    <keyword>Mime</keyword>

    <abstract>
      <t>This document attempts to identify and classify application scenarios of P2P
      based SIP. It does not attempt to exhaustively enumerate these scenarios,
      and is focused exclusively on scenarios related to real-time IP
      communication.</t>
    </abstract>
  </front>

  <middle>
    <section anchor="sectIntro" title="Introduction">
      <t>This document attempts to identify and classify application scenarios for
      Peer-to-Peer (P2P) based Session Initiation Protocol (SIP) <xref
      target="RFC3261"></xref>. Identifying application scenarios will help to understand
      and clarify requirements of P2PSIP. In particular, these application scenarios will
      assist the P2PSIP community in identifying commonalities and differences between requirements
      for different application scenarios, which in turn will help define the
      near-term scope of specifications and provide a perspective on future
      specifications.</t>

      <t>Only application scenarios related to real-time IP communications, such as VoIP,
      Instant Messaging (IM), and presence are considered in this document.
      Application scenarios of other kinds, even if interesting and possibly useful
      applications of P2PSIP, are out of scope for this document. Thus, application
      scenarios described herein are application scenarios of P2P IP real-time communications,
      and P2PSIP is a protocol choice rather than a constraining factor for
      most of them. In describing application scenarios, no deliberation on implementation
      is provided. Some of the application scenarios presented may already be implemented
      or deployed, possibly using proprietary technology.</t>

	<t> The list of application scenarios compiled here is by no means a
      complete list of uses cases of P2PSIP, and further cases would be
      limited only by the imagination. P2PSIP participants who expect to use P2PSIP technology
	for application scenarios that don't match any of the combinations of attributes included in this
	document are invited to contribute descriptions of additional application scenarios to the 
	P2PSIP working group mailing list.</t>

	<t>We tried to capture the deployment characteristics of the application scenarios such as 
	whether the nodes will span over multiple physical network administrative domains or whether the 
	ID must be controlled by a central authority. The characteristics are presented as scenario-attribute tables. 
	The values in the tables are what we think are most likely and we understand there may be similar scenarios
	with different choices for some attribute values.</t>

      <t>Some of these application scenarios, while difficult to implement using a
      traditional client server SIP (CS SIP) architecture may not require P2P
      and could be implemented in other ways. While these have often been
      presented as scenarios calling for P2P communication, the authors
      recognize that other technologies may also be applicable to these application 
      scenarios.</t>

      <t>Since the original iteration of this document, the P2PSIP WG has been
      formed and numerous documents have been submitted that include some
      number of application scenarios. We will not try to enumerate them here. This draft
      draws from these documents, as well as discussions at the P2PSIP ad-hoc
      and WG meetings and numerous mailing list and personal conversations of
      the authors.</t>

    </section>

    <section anchor="sectTermi" title="Terminology">
      <t>We use terminology defined in <xref target="RFC3261">RFC 3261</xref>
      in this document without further definition.</t>
      <t>We use terminology defined in <xref
      target="I-D.willis-p2psip-concepts">Concepts and Terminology for Peer to
      Peer SIP</xref> draft in this document without further definition.</t>
	<t>We define the attributes used in the discussion of each application scenario
	in <xref target="attributes"></xref>.</t>
    </section>

<section anchor="attributes" title="Application Scenario Attributes">

    	<t> The attributes used in the application scenarios matrixes in subsequent sections are explained here.</t>
	<t>
		<list style='hanging' hangIndent='6'>
        		<t hangText="Application Scenario:">
				The name of an application scenario (previously called a "use case").</t>
        		<t hangText="Section in Draft:">
				A cross-reference to the section number in this draft where the application 
				scenario is described.</t>
			<t hangText="Number of Peers:">
				The number of peers that will be active in an overlay at any given point in time.</t>
        		<t hangText="Number of Users:">
				The number of users that will be served by an overlay at any given point in time.</t>
			<t>Note that if there are more users than peers, this implies that some client protocol is required, 
				whether "client protocol" is a P2PSIP client protocol or the SIP protocol 
				(if the P2PSIP overlay is also providing 
					<xref target="RFC3263">RFC 3263</xref>-style routing for unmodified SIP clients).</t>
        		<t hangText="Overlay spans administrative domains:">
				Whether the overlay spans across multiple physical network administrative domains.
				If "yes", 
				this makes IP multicast and centralized operations and management unlikely.</t>
        		<t hangText="Multicast Available:">
				Whether "application-level multicast", "IP multicast", or "link multicast" 
				may be available for a typical overlay.</t>
				<t>Note that these are ordered - link multicast implies IP multicast could be available, and 
				IP multicast implies application-level multicast could be available.</t>
        		<t hangText="P2P Client Support:">
				Whether the overlay need to support a P2P Client protocol, i.e., whether the overlay contains 
					P2P Clients as well as Peers.</t>
        		<t hangText="Interoperation with CS-SIP:">
				Whether the overlay must also interact with legacy SIP clients and SIP proxies.</t>
				<t>Note that one or more peers in the overlay may also act as PSTN gateways.</t>
			<t hangText="Non-stop Operation:">Whether this application scenarios allows the overlay to become 
				unavailable for periods of time (for example, could an overlay stop operating in order 
				to change DHT algorithms, or would the overlay have to support two DHT algorithms in 
				"ships in the night" mode?)</t>
        		<t hangText="Centralized Operations and Management:">
				Whether any centralized operations/management entity is responsible for successful operation of 
				the overlay.</t>
        		<t hangText="Centralized ID Control:">
				Whether ID assignment by central authority is required within an overlay (basically, 
				whether the overlay can be Sybil-attacked - the theory is that if IDs are controlled by
				a centralized entity, overlay operators simply remove misbehaving users from the 
				authorization registry).</t>
        		<t hangText="Supports Anonymous Users:">
				Whether this application scenario allows users to connect to an overlay without providing 
				any identity.</t>
        		<t hangText="Carrier-Class Robustness:">
				Whether the overlay must provide reliable storage and retrieval in the face of node failure.</t>
        		<t hangText="NATs within a single overlay">
				Whether the peer protocol must expect to perform NAT traversal as part of normal operation.</t>
			<t hangText="DNS available:">
				Whether DNS is available so that peers may perform DNS lookups as part of the overlay JOIN 
				operation.</t>
			<t hangText="End-to-end SIP Encryption:">
				Whether this application scenario requires SIP traffic between two peers to be 
				encrypted, so SIP requests and responses are not visible to intermediate peers (peers
				that forward traffic between two peers that aren't directly connected). 
				In these cases, hop-by-hop TLS encryption, although appropriate when traversing trusted
				SIP Proxies, is not appropriate when traversing untrusted P2PSIP Peers.</t>
   		</list>
	</t>

</section>

    <section anchor="appscen" title="Application Scenarios">
      <t>Application scenarios are grouped according to the characteristics of the network
      environment in which the end users or devices participating in the P2P
      overlay are communicating with each other.</t>

      <section title="Global Internet Environment">
        <t>The global Internet environment consists of a large number of
        autonomous networks with diverse characteristics. Thus, there is no
        central administration or network control of the physical network on a
        global scale. Communication paths between two remote devices may span
        multiple administrative domains and should be assumed to be insecure.
        Note that most well-known P2P file sharing overlay networks have
        operated in this environment.</t>
														
        <section title="Public P2P VoIP Service Providers">
          <t>Skype is an outstanding example of a public VoIP service provider
          using P2P technology among end user devices, although Skype uses a
          proprietary protocol. Recent research has shown <xref
          target="skypestudy"></xref> that Skype uses a central login server,
          responsible for management of registered user names. End users are
          authenticated via a certificate signed by a central server. End user
          devices are distributed across the global Internet. The number of
          participating end user devices is very large. A major motivation of
          using P2P between end user devices for a commercial VoIP service is
          a reduction in infrastructure and operational costs.</t>

        <t><xref target="table_4.1.1"> </xref> provides a high-level overview of this category.</t>

    <texttable anchor='table_4.1.1'>								
<ttcol align='center'>	Application Scenario	</ttcol>	<ttcol align='center'>	Public P2P VoIP Service Providers	</ttcol>	<ttcol align='center'>	Open Global P2P VoIP Network	</ttcol>
<c>	Section in Draft	</c>	<c>	4.1.1	</c>	<c>	4.1.2	</c>
<c>	Number of Peers	</c>	<c>	hundreds	</c>	<c>	thousands	</c>
<c>	Number of Users	</c>	<c>	millions	</c>	<c>	millions	</c>
<c>	Overlay spans administrative domains	</c>	<c>	no	</c>	<c>	yes	</c>
<c>	Multicast Available	</c>	<c>	no	</c>	<c>	no	</c>
<c>	P2P Client Support	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Interaction with CS-SIP	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Non-stop Operation	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Centralized Operations and Management	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Centralized ID Control	</c>	<c>	yes	</c>	<c>	no	</c>
<c>	Supports Anonymous Users	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Carrier-Class Robustness	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	NATs within a single overlay	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	DNS available	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	End-to-end SIP Encryption	</c>	<c>	yes	</c>	<c>	yes	</c>
	        <postamble> 	Public P2P VoIP Service Providers and Open Global P2P VoIP Network	</postamble>					
</texttable>													

        </section>

        <section title="Open Global P2P VoIP Network">
          <t>This is a global P2P VoIP network in which there is no central
          authority such as a single service provider. Anyone can join and
          leave the network freely and anyone can implement the software to
          participate in the overlay network. In such a system, the protocols
          used must be based on open standards. This P2P VoIP network
          resembles the global Internet itself in that it has distributed
          management and growth, enables anyone to reach anyone else in the
          overlay network, and any device supporting the standard protocols
          can be used. </t>

        <t><xref target="table_4.1.1"> </xref> provides a high-level overview of this category.</t>					

        </section>

        <section title="Wide Area Networks of Consumer Electronics Devices">
          <t>Instant messaging application
          software provides presence, text and media messaging, and 
          file transfer capabilities between online users. As more and
          more multimedia 
          consumer electronics devices such as cameras, camcorders and
          televisions become network aware, instant sharing of multimedia
          content such as photos and video clips between family members and
          friends will be desirable. VoIP may not be needed on some of these
          consumer electronics devices, however in other cases such as
          gaming, voice communication between users may be highly
          desirable. As consumer electronics providers may desire to
          provide these capabilities without investing in extensive
          server capabilities, a global P2P network supporting presence is an 
          important infrastructure component for this application scenario.</t>

        <t><xref target="table_4.1.3"> </xref> provides a high-level overview of this category.</t>

    <texttable anchor='table_4.1.3'>					
<ttcol align='center'>	Application Scenario	</ttcol>	<ttcol align='center'>	Wide Area Networks of Consumer Electronics Devices	</ttcol>
<c>	Section in Draft	</c>	<c>	4.1.3	</c>
<c>	Number of Peers	</c>	<c>	thousands	</c>
<c>	Number of Users	</c>	<c>	millions	</c>
<c>	Overlay spans administrative domains	</c>	<c>	yes	</c>
<c>	Multicast Available	</c>	<c>	no	</c>
<c>	P2P Client Support	</c>	<c>	yes	</c>
<c>	Interaction with CS-SIP	</c>	<c>	no	</c>
<c>	Non-stop Operation	</c>	<c>	no	</c>
<c>	Centralized Operations and Management	</c>	<c>	maybe	</c>
<c>	Centralized ID Control	</c>	<c>	maybe	</c>
<c>	Supports Anonymous Users	</c>	<c>	no	</c>
<c>	Carrier-Class Robustness	</c>	<c>	no	</c>
<c>	NATs within a single overlay	</c>	<c>	yes	</c>
<c>	DNS available	</c>	<c>	yes	</c>
<c>	End-to-end SIP Encryption	</c>	<c>	no	</c>
	        <postamble> 	Wide Area Networks of Consumer Electronics Devices	</postamble>		
</texttable>									

        </section>

        <section title="Multimedia content sharing via Application Layer Multicasting (Content Providers or Ad Hoc)">
          <t>IP-layer multicasting is not generally available beyond the
          boundary of single IP subnet. Application layer multicasting has
          become a plausible alternative to IP-layer multicasting. In
          application layer multicasting, the nodes that need to receive the
          content from the same source form a distribution network, typically
          of a tree-like topology, and relay the received content to other
          nodes in the distribution network. This technique can be used to
          multicasting video or audio stream to a number of nodes distributed
          over the Internet (or across multiple IP subnets). </t>
	    <t>Note that this application scenario covers two types of deployments - 
          large-scale commercial audio or video
          distribution/broadcasting services such as Internet radio or TV
          services ("Content Provider") or to ad-hoc video sharing among a group of friends ("Ad Hoc").</t>

        <t><xref target="table_4.1.4"> </xref> provides a high-level overview of this category.</t>

    <texttable anchor='table_4.1.4'>								
<ttcol align='center'>	Application Scenario	</ttcol>	<ttcol align='center'>	Multimedia content sharing via Application Layer Multicasting (Content Providers)	</ttcol>	<ttcol align='center'>	Multimedia content sharing via Application Layer Multicasting (Ad Hoc)	</ttcol>
<c>	Section in Draft	</c>	<c>	4.1.4	</c>	<c>	4.1.4	</c>
<c>	Number of Peers	</c>	<c>	hundreds	</c>	<c>	hundreds	</c>
<c>	Number of Users	</c>	<c>	thousands	</c>	<c>	thousands	</c>
<c>	Overlay spans administrative domains	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Multicast Available	</c>	<c>	application multicast	</c>	<c>	application multicast	</c>
<c>	P2P Client Support	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Interaction with CS-SIP	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Non-stop Operation	</c>	<c>	yes	</c>	<c>	no	</c>
<c>	Centralized Operations and Management	</c>	<c>	yes	</c>	<c>	no	</c>
<c>	Centralized ID Control	</c>	<c>	yes	</c>	<c>	no	</c>
<c>	Supports Anonymous Users	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Carrier-Class Robustness	</c>	<c>	yes	</c>	<c>	no	</c>
<c>	NATs within a single overlay	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	DNS available	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	End-to-end SIP Encryption	</c>	<c>	yes	</c>	<c>	no	</c>
	        <postamble> 	Multimedia content sharing via Application Layer Multicasting (Content Provider and Ad Hoc) </postamble>						
</texttable>								

        </section>

      </section>

      <section title="Environments with Limited Connectivity to the Internet or Infrastructure">
        <t>When there is no physical network available for stable deployment
        of client server SIP or an instant deployment of real-time
        communication systems is required, the P2P approach may be the only
        feasible solution. Examples of such environment are isolated wireless
        ad-hoc networks with no connection to the Internet or ad-hoc networks
        with limited connectivity to the Internet in situations like outdoor
        public events, emergencies, and battlefields. Any type of manual
        configuration is difficult to achieve because technical support is not
        readily available in such environment. In some cases, connectivity to
        the global Internet may be available, but be very expensive, of
        limited capacity, or unstable, such as satellite connections. In such
        cases, it is preferable to localize communications as much as
        possible, reducing dependency on any infrastructure in the global
        Internet.</t>

        <t><xref target="table_4.2"> </xref> provides a high-level overview of this category.</t>	

<texttable anchor='table_4.2'>														
<ttcol align='center'>	Application Scenario	</ttcol>	<ttcol align='center'>	Ad-Hoc and Ephemeral Groups	</ttcol>	<ttcol align='center'>	Extending the Reach of Mobile Devices	</ttcol>	<ttcol align='center'>	Impeded Access	</ttcol>	<ttcol align='center'>	Local Area Networks of Consumer Electronics Devices	</ttcol>
<c>	Section in Draft	</c>	<c>	4.2.1	</c>	<c>	4.2.2	</c>	<c>	4.2.3	</c>	<c>	4.2.4	</c>
<c>	Number of Peers	</c>	<c>	tens	</c>	<c>	hundreds	</c>	<c>	hundreds	</c>	<c>	tens	</c>
<c>	Number of Users	</c>	<c>	tens	</c>	<c>	hundreds	</c>	<c>	hundreds	</c>	<c>	tens	</c>
<c>	Spans administrative domains	</c>	<c>	no	</c>	<c>	no	</c>	<c>	yes	</c>	<c>	no	</c>
<c>	Multicast Available	</c>	<c>	link multicast	</c>	<c>	link multicast	</c>	<c>	no	</c>	<c>	link multicast	</c>
<c>	P2P Client Support	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Interaction with CS-SIP	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Non-stop Operation	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Centralized Operations and Management	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Centralized ID Control	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Supports Anonymous Users	</c>	<c>	yes	</c>	<c>	yes	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Carrier-Class Robustness	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	NATs within a single overlay	</c>	<c>	no	</c>	<c>	no	</c>	<c>	yes	</c>	<c>	no	</c>
<c>	DNS available	</c>	<c>	no	</c>	<c>	no	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	End-to-end SIP Encryption	</c>	<c>	no	</c>	<c>	no	</c>	<c>	yes	</c>	<c>	no	</c>
	        <postamble> 	Environments with Limited Connectivity to the Internet or Infrastructure	</postamble>											
</texttable>																																													

        <section title="Ad-Hoc and Ephemeral Groups">
          <t>Groups of individuals meeting together have need for
          collaborative communications systems that are ephemeral in nature,
          have minimum (ideally zero) configuration, and do not depend on
          connectivity to the Internet. These scenarios require an arbitrary
          number of users to connect communications devices. These
          can include cases where Internet connectivity due to
          remote location, inability to pay for connectivity, or
          following a natural disaster where service is interrupted.</t>

          <t>Example: A group gets together for a meeting, but there is no
          Internet connectivity. If the users establish a wireless ad hoc
          network or have a base station, all users may connect and establish
          chat sessions using an IM protocol with no need for server
          configuration.</t>

          <t>Example: Following a disaster, the local fire department arrives.
          Each fire fighter has a wireless handset, and one or more trucks
          have wireless base stations. When a nearby locality sends additional
          rescuers, their wireless handsets should be able to instantly join
          the communications network and communicate, without the
          need for central configuration.</t>

        </section>

        <section title="Extending the Reach of Mobile Devices">

          <t><cref>Spencer has misgivings about "Extending the Reach of Mobile Devices", based on 
		(1) relaying at the link layer or at the 
		network layer would make more sense, and would work for devices that do not support P2PSIP, and (2)
		although small networks might work well enough when peers simply forward a request "around the overlay", 
		in larger networks 
		the problem space morphs from forwarding traffic in *the* direction of 
		a destination to routing traffic in the *best* direction to the destination - a much harder problem.
	    </cref></t>

          <t>A network of mobile devices can relay traffic between themselves
          to reach a base station, even if the base station is out of reach of
          that device.</t>

          <t>Example: A user has a handset for communication that cannot reach
          a base station. Some other user is within range of both that user
          and a base station. This intermediate user can serve as a relay for
          the caller who is out of range. A system might make this feature
          optional for standard communication and mandatory for E911.</t>

        </section>

        <section title="Impeded Access">

          <t>Certain groups may have their ability to communicate impeded.
          These users should be able to communicate without the need to
          connect to any centralized servers, which may be blocked by
          providers upstream of the user. A fully decentralized system cannot
          be completely disconnected without removing connectivity at the
          basic Internet level.</t>

          <t>Example: A user wishes to use an IP telephony service to
          communicate PC to PC with a friend, but the ports commonly used by
          these services, or the servers used for authentication, are
          blocked by the ISP because the ISP also offers communications
          systems and have a vested interest in denying access to external communications systems.</t>

          <t>Example: A user with an Internet enabled PDA devices wishes to connect
          with colleagues, but traditional services are blocked to ensure that
          SMS or voice minutes are used (at additional cost) instead.</t>

        </section>

        <section title="Local Area Networks of Consumer Electronics Devices">
          <t>In addition to consumer devices sharing information with
          other users across the Internet, having devices that can
          locate each other and exchange information within the local
          LAN of a particular user may also be an attractive
          application. In this case, devices could use P2PSIP to
          locate multimedia resources available on other devices and
          stream the information between the devices.
	    </t> 

          <t>Example: A user wishes to share content among consumer electronics devices within a home network.</t>
        </section>

      </section>

      <section title="Managed, Private Network Environments">
        <t>A corporate network or a school campus network is an example of the
        managed, private network environment. Most likely client server SIP
        can be used and managed for real-time communication applications in
        these environments. However, in certain scenarios, P2PSIP may be used
        instead or as a complementary means, to achieve various goals such as
        cost and management overhead reduction, scalability, and system
        robustness.</t>

        <t><xref target="table_4.3"> </xref> provides a high-level overview of this category.</t>	
										
<texttable anchor='table_4.3'>											
<ttcol align='center'>	Application Scenario	</ttcol>	<ttcol align='center'>	Serverless or Small Scale IP-PBX	</ttcol>	<ttcol align='center'>	P2P for Redundant SIP Servers	</ttcol>	<ttcol align='center'>	Failover for Centralized Systems	</ttcol>
<c>	Section in Draft	</c>	<c>	4.3.1	</c>	<c>	4.3.2	</c>	<c>	4.3.3	</c>
<c>	Number of Peers	</c>	<c>	hundreds	</c>	<c>	hundreds	</c>	<c>	tens	</c>
<c>	Number of Users	</c>	<c>	hundreds	</c>	<c>	hundreds	</c>	<c>	tens	</c>
<c>	Spans administrative domains	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Multicast Available	</c>	<c>	IP multicast	</c>	<c>	IP multicast	</c>	<c>	IP multicast	</c>
<c>	P2P Client Support	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Interaction with CS-SIP	</c>	<c>	yes	</c>	<c>	no	</c>	<c>	yes	</c>
<c>	Non-stop Operation	</c>	<c>	yes	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Centralized Operations and Management	</c>	<c>	maybe	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Centralized ID Control	</c>	<c>	self-cert?	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	Supports Anonymous Users	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
<c>	Carrier-Class Robustness	</c>	<c>	no	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	NATs within a single overlay	</c>	<c>	yes	</c>	<c>	no	</c>	<c>	no	</c>
<c>	DNS available	</c>	<c>	no	</c>	<c>	yes	</c>	<c>	yes	</c>
<c>	End-to-end SIP Encryption	</c>	<c>	no	</c>	<c>	no	</c>	<c>	no	</c>
	        <postamble> 	Managed, Private Network Environments	</postamble>								
</texttable>																															
	<section title="Serverless or Small Scale IP-PBX">
          <t>Many small enterprises have a need for integrated communications
          systems. These systems have slightly different requirements than
          more traditional IP PBXs. For small enterprises, there may be no
          administrator for these systems, requiring the systems to be
          essentially self-configuring and/or self-organizing. Additional
          endpoints should be able to be added with no requirements for
          configuration on central devices.</t>

          <t>These systems should offer the feature sets similar to those of
          client server type PBX systems. Connectivity to the PSTN is an
          important feature for these systems. In addition, they may support
          features such as call transfer, voice mail, and possibly even other
          communications modes such as instant messaging or media features
          such as video or conference services. There are already commercial
          products of this type.</t>

          <t>Example: Small organizations without centralized IT</t>
        </section>

        <section title="P2P for Redundant SIP Proxies">
          <t>Service providers may wish to connect a farm of proxies together
          in a transparent way, passing resources (user registrations or other
          call information) between themselves with as little configuration or
          traffic as possible. Ideally, the redundancy and exchange of
          information should require a minimum of configuration between the
          devices. P2P architecture between the proxies allows proxy farms
          to be organized and operated in this way. With this approach, it is
          easy to add more proxies with minimal service disruptions and
          increases the robustness of the system.</t>

	    <t>Example: a SIP service provider may wish to scale SIP proxies by 
	    using a P2PSIP overlay that provides <xref target="RFC3263">RFC 3263</xref> request routing services,
          instead of using either front-end load balancing devices or making 
          the structure of the proxy farm visible outside the proxy farm itself.</t>

        </section>

        <section title="Failover for Centralized Systems">
          <t>A traditional centralized SIP server, such as used in an IP-PBX,
          forms a single point of failure of an otherwise fault-independent
          network. Relying on P2PSIP as a backup to the centralized server
          allows the communications system to continue functioning normally in
          the event of planned or unplanned service interruptions of the
          central IP-PBX. When combined with a low-configuration
          P2PSIP PBX, this can provide a simple, standalone
          communications system for the developing world that allows
          local communication even when Internet connectivity is severed.</t>

          <t>Example: A small company has a central IP-PBX. When that device
          experiences a failure, the handsets are able to transparently
          continue operation for the 24 hours it takes to obtain a replacement
          switch.</t>

          <t>Example: A village in the developing world has connectivity that
          is limited by weather (microwave connection) or is solar powered. It
          would be desirable for intra-village communication to continue to
          function in the absence of Internet connectivity.</t>

        </section>
      </section>
    </section>

    <section title="Changes from draft-bryan-p2psip-usecases-00">

        <t>This draft builds on the analysis done for an earlier draft, draft-bryan-p2psip-usecases-00, now expired. 
		For ease of reference, <xref target="table_changes"> </xref> shows the mapping of use cases described in 
		draft-bryan-p2psip-usecases-00 onto the application scenarios described in this document.</t>	

<texttable anchor='table_changes'>								
<ttcol align='center'>	Use Case	</ttcol>	<ttcol align='center'>	Section in Use Cases Draft	</ttcol>	<ttcol align='center'>	Application Scenario	</ttcol>
<c>	Public P2P VoIP Service Providers	</c>	<c>	3.1.1	</c>	<c>	(no change)	</c>
<c>	Open Global P2P VoIP Network	</c>	<c>	3.1.2	</c>	<c>	(no change)	</c>
<c>	Presence Using Multimedia Consumer Electronics Devices	</c>	<c>	3.1.3	</c>	<c>	Split into (Content Provider) and (Ad Hoc) scenarios	</c>
<c>	Multimedia content sharing via Application Layer Multicasting	</c>	<c>	3.1.4	</c>	<c>	(no change)	</c>
<c>	Impeded Access	</c>	<c>	3.2.1	</c>	<c>	(no change)	</c>
<c>	Anonymous Communications	</c>	<c>	3.2.2	</c>	<c>	Became an attribute of other scenarios	</c>
<c>	Security Conscious Small Organizations	</c>	<c>	3.2.3	</c>	<c>	Became an attribute of other scenarios	</c>
<c>	Ad-Hoc and Ephemeral Groups	</c>	<c>	3.3.1	</c>	<c>	(no change)	</c>
<c>	Emergency First Responder Networks	</c>	<c>	3.3.2	</c>	<c>	Merged with Ad-Hoc and Ephemeral Groups	</c>
<c>	Extending the Reach of Mobile Devices	</c>	<c>	3.3.3	</c>	<c>	(no change)	</c>
<c>	Deployments in the Developing World	</c>	<c>	3.3.4	</c>	<c>	Merged with Failover	</c>
<c>	Serverless or Small Scale IP-PBX	</c>	<c>	3.4.1	</c>	<c>	(no change)	</c>
<c>	P2P for Redundant SIP Servers	</c>	<c>	3.4.2	</c>	<c>	(no change)	</c>
<c>	Failover for Centralized Systems	</c>	<c>	3.4.3	</c>	<c>	(no change)	</c>
	        <postamble> 	Changes from draft-bryan-p2psip-usecases-00	</postamble>
</texttable>						

    </section>

    <section title="Acknowledgments">
      <t>The following persons have contributed application scenarios or ideas
      to this document:</t>

      <t>Cullen Jennings, Philip Matthews, Henry Sinnreich, Adam Roach, Robert
      Sparks, Kundan Singh, Henning Schulzrinne, K. Kishore Dhara, and Salman
      A. Baset.</t>
    </section>

    <section title="Security Considerations">
      <t>The security requirements of the various application scenarios vary tremendously.
      They should be discussed in more detail in this document.</t>
    </section>

    <section title="IANA Considerations">
      <t>This document has no IANA Considerations.</t>
    </section>
  </middle>

  <back>
    <references title="Normative References">

      <?rfc include="reference.I-D.willis-p2psip-concepts"?>

	&rfc3261;

	&rfc3263;

    </references>

    <references title="Informative References">
      <reference anchor="skypestudy">
        <front>
          <title>An Analysis of the Skype Peer-to-Peer Internet Telephony
          Protocol</title>

          <author fullname="Salman A. Baset" initials="S. A." surname="Baset">
            <organization>Columbia University</organization>
          </author>

          <author fullname="Henning Schulzrinne" initials="H."
                  surname="Schulzrinne">
            <organization>Columbia University</organization>
          </author>

          <date month="September" year="2004" />
        </front>

        <seriesInfo name="Technical Report, Department of Computer Science, Columbia University"
                    value="0309-04" />

        <format target="http://www1.cs.columbia.edu/~library/TR-repository/reports/reports-2004/cucs-039-04.pdf"
                type="PDF" />
      </reference>
    </references>
  </back>
</rfc>
<!--  LocalWords:  bryan
 -->

PAFTECH AB 2003-20262026-04-23 20:47:08