RFC 7837






Internet Engineering Task Force (IETF)                       S. Krishnan
Request for Comments: 7837                                      Ericsson
Category: Experimental                                     M. Kuehlewind
ISSN: 2070-1721                                               ETH Zurich
                                                              B. Briscoe
                                              Simula Research Laboratory
                                                                C. Ralli
                                                              Telefonica
                                                                May 2016


        IPv6 Destination Option for Congestion Exposure (ConEx)

Abstract



   Congestion Exposure (ConEx) is a mechanism by which senders inform
   the network about the congestion encountered by packets earlier in
   the same flow.  This document specifies an IPv6 destination option
   that is capable of carrying ConEx markings in IPv6 datagrams.

Status of This Memo



   This document is not an Internet Standards Track specification; it is
   published for examination, experimental implementation, and
   evaluation.

   This document defines an Experimental Protocol for the Internet
   community.  This document is a product of the Internet Engineering
   Task Force (IETF).  It represents the consensus of the IETF
   community.  It has received public review and has been approved for
   publication by the Internet Engineering Steering Group (IESG).  Not
   all documents approved by the IESG are a candidate for any level of
   Internet Standard; see Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc7837.














Krishnan, et al.              Experimental                      [Page 1]

RFC 7837                ConEx Destination Option                May 2016


Copyright Notice



   Copyright (c) 2016 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents



   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   3
   2.  Conventions Used in This Document . . . . . . . . . . . . . .   3
   3.  Requirements for the Coding of ConEx in IPv6  . . . . . . . .   4
   4.  ConEx Destination Option (CDO)  . . . . . . . . . . . . . . .   5
   5.  Implementation in the Fast Path of ConEx-Aware Routers  . . .   8
   6.  Tunnel Processing . . . . . . . . . . . . . . . . . . . . . .   8
   7.  Compatibility with Use of IPsec . . . . . . . . . . . . . . .   9
   8.  Mitigating Flooding Attacks by Using Preferential Drop  . . .   9
   9.  Security Considerations . . . . . . . . . . . . . . . . . . .  11
   10. IANA Considerations . . . . . . . . . . . . . . . . . . . . .  11
   11. References  . . . . . . . . . . . . . . . . . . . . . . . . .  11
     11.1.  Normative References . . . . . . . . . . . . . . . . . .  11
     11.2.  Informative References . . . . . . . . . . . . . . . . .  12
   Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . .  13
   Authors' Addresses  . . . . . . . . . . . . . . . . . . . . . . .  13



















Krishnan, et al.              Experimental                      [Page 2]

RFC 7837                ConEx Destination Option                May 2016


1.  Introduction



   Congestion Exposure (ConEx) [RFC7713] is a mechanism by which senders
   inform the network about the congestion encountered by packets
   earlier in the same flow.  This document specifies an IPv6
   destination option [RFC2460] that can be used for performing ConEx
   markings in IPv6 datagrams.

   This document specifies the ConEx wire protocol in IPv6.  The ConEx
   information can be used by any network element on the path to, for
   example, do traffic management or egress policing.  Additionally,
   this information will potentially be used by an audit function that
   checks the integrity of the sender's signaling.  Further, each
   transport protocol that supports ConEx signaling will need to
   precisely specify when the transport sets ConEx markings (e.g., the
   behavior for TCP is specified in [RFC7786]).

   This document specifies ConEx for IPv6 only.  Due to space
   limitations in the IPv4 header and the risk of options that might be
   stripped by a middlebox in IPv4, the primary goal of the working
   group was to specify ConEx in IPv6 for experimentation.

   This specification is experimental to allow the IETF to assess
   whether the decision to implement the ConEx Signal as a destination
   option fulfills the requirements stated in this document, as well as
   to evaluate the proposed encoding of the ConEx Signals as described
   in [RFC7713].

   The duration of this experiment is expected to be no less than two
   years from publication of this document as infrastructure is needed
   to be set up to determine the outcome of this experiment.
   Experimenting with ConEx requires IPv6 traffic.  Even though the
   amount of IPv6 traffic is growing, the traffic mix carried over IPv6
   is still very different than over IPv4.  Therefore, it might take
   longer to find a suitable test scenario where only IPv6 traffic is
   managed using ConEx.

2.  Conventions Used in This Document



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









Krishnan, et al.              Experimental                      [Page 3]

RFC 7837                ConEx Destination Option                May 2016


3.  Requirements for the Coding of ConEx in IPv6



   A set of requirements for an ideal concrete ConEx wire protocol is
   given in [RFC7713].  The ConEx working group recognized that it will
   be difficult to find an encoding in IPv6 that satisfies all
   requirements.  The choice in this document to implement the ConEx
   information in a destination option aims to satisfy those
   requirements that constrain the placement of ConEx information:

   R-1:  The marking mechanism needs to be visible to all ConEx-capable
         nodes on the path.

   R-2:  The mechanism needs to be able to traverse nodes that do not
         understand the markings.  This is required to ensure that ConEx
         can be incrementally deployed over the Internet.

   R-3:  The presence of the marking mechanism should not significantly
         alter the processing of the packet.  This is required to ensure
         that ConEx-Marked packets do not face any undue delays or drops
         due to a badly chosen mechanism.

   R-4:  The markings should be immutable once set by the sender.  At
         the very least, any tampering should be detectable.

   Based on these requirements, four solutions to implement the ConEx
   information in the IPv6 header have been investigated: hop-by-hop
   options, destination options, using IPv6 header bits (from the flow
   label), and new extension headers.  After evaluating the different
   solutions, the ConEx working group concluded that the use of a
   destination option would best address these requirements.

   Hop-by-hop options would have been the best solution for carrying
   ConEx markings if they had met requirement R-3.  There is currently
   some work ongoing in the 6MAN working group to address this very
   issue [HBH-HEADER].  This new behavior would address R-3 and would
   make hop-by-hop options the preferred solution for carrying ConEx
   markings.

   Choosing to use a destination option does not necessarily satisfy the
   requirement for on-path visibility, because it can be encapsulated by
   additional IP header(s).  Therefore, ConEx-aware network devices,
   including policy or audit devices, might have to follow the chaining
   (extension-) headers into inner IP headers to find ConEx information.
   This choice was a compromise between fast-path performance of ConEx-
   aware network nodes and visibility, as discussed in Section 5.

   Please note that the IPv6 specification [RFC2460] does not require or
   expect intermediate nodes to inspect destination options such as the



Krishnan, et al.              Experimental                      [Page 4]

RFC 7837                ConEx Destination Option                May 2016


   ConEx Destination Option (CDO).  This implies that ConEx-aware
   intermediate nodes following this specification need updated
   extension header processing code to be able read the destination
   options.

4.  ConEx Destination Option (CDO)



   The CDO is a destination option that can be included in IPv6
   datagrams that are sent by ConEx-aware senders in order to inform
   ConEx-aware nodes on the path about the congestion encountered by
   packets earlier in the same flow or the expected risk of encountering
   congestion in the future.  The CDO does not have any alignment
   requirements.

    0                   1                   2
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |  Option Type  | Option Length |X|L|E|C|  res  |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                 Figure 1: ConEx Destination Option Layout

   Option Type

      8-bit identifier of the type of option.  Set to the value 30
      (0x1E) allocated for experimental work.

   Option Length

      8-bit unsigned integer.  The length of the option in octets
      (excluding the Option Type and Option Length fields).  Set to the
      value 1.

   X Bit

      When this bit is set, the transport sender is using ConEx with
      this packet.  If it is not set, the sender is not using ConEx with
      this packet.

   L Bit

      When this bit is set, the transport sender has experienced a loss.

   E Bit

      When this bit is set, the transport sender has experienced
      congestion signaled using Explicit Congestion Notification (ECN)
      [RFC3168].



Krishnan, et al.              Experimental                      [Page 5]

RFC 7837                ConEx Destination Option                May 2016


   C Bit

      When this bit is set, the transport sender is building up
      congestion credit in the audit function.

   Reserved (res)

      These four bits are not used in the current specification.  They
      are set to zero by the sender and are ignored by the receiver.

   All packets sent over a ConEx-capable TCP connection or belonging to
   the same ConEx-capable flow MUST carry the CDO.  The chg bit (the
   third-highest-order bit) in the CDO Option Type field is set to zero,
   meaning that the CDO option is immutable.  Network devices with
   ConEx-aware functions read the flags, but all network devices MUST
   forward the CDO unaltered.

   The CDO SHOULD be placed as the first option in the Destination
   Option header before the AH [RFC4302] and/or Encapsulating Security
   Payload (ESP) [RFC4303] (if present).  The IPsec Authentication
   Header (AH) MAY be used to verify that the CDO has not been modified.

   If the X bit is zero, all the other three bits are undefined and thus
   MUST be ignored and forwarded unchanged by network nodes.  The X bit
   set to zero means that the connection is ConEx-capable but that this
   packet MUST NOT be counted when determining ConEx information in an
   audit function.  This can be the case if no congestion feedback is
   (currently) available, e.g., in TCP if one endpoint has been
   receiving data but sending nothing but pure ACKs (no user data) for
   some time.  This is because pure ACKs do not advance the sequence
   number, so the TCP endpoint receiving them cannot reliably tell
   whether any have been lost due to congestion.  Pure TCP ACKs cannot
   be ECN-marked either [RFC3168].

   If the X bit is set, any of the other three bits (L, E, or C) might
   be set.  Whenever one of these bits is set, the number of bytes
   carried by this IP packet (including the IP header that directly
   encapsulates the CDO and everything that IP header encapsulates)
   SHOULD be counted to determine congestion or credit information.  In
   IPv6, the number of bytes can easily be calculated by adding the
   number 40 (length of the IPv6 header in bytes) to the value present
   in the Payload Length field in the IPv6 header.

   The credit signal represents potential for congestion.  If a
   congestion event occurs, a corresponding amount of credit is consumed
   as outlined in [RFC7713].  A ConEx-enabled sender SHOULD, therefore,
   signal sufficient credit in advance of any congestion event to cover
   the (estimated maximum) amount of lost or CE-marked bytes that could



Krishnan, et al.              Experimental                      [Page 6]

RFC 7837                ConEx Destination Option                May 2016


   occur in such a congestion event.  This estimation depends on the
   heuristics used and aggressiveness of the sender when deciding the
   appropriate sending rate (congestion control).  Note that the maximum
   congestion risk is that all packets in flight get lost or CE-marked;
   therefore, this would be the most conservative estimation for the
   congestion risk.  After a congestion event, if the sender intends to
   take the same risk again, it just needs to replace the consumed
   credit as non-consumed credit does not expire.  For the case of TCP,
   this is described in detail in [RFC7786].

   If the L or E bit is set, a congestion signal in the form of a loss
   or an ECN mark, respectively, was previously experienced by the same
   connection.

   In principle, all of these three bits (L, E, or C) might be set in
   the same packet.  In this case, the packet size MUST be counted once
   for each respective ConEx information counter.

   If a network node extracts the ConEx information from a connection,
   it is expected to hold this information in bytes, e.g., comparing the
   total number of bytes sent with the number of bytes sent with ConEx
   congestion marks (L or E) to determine the current whole path
   congestion level.  Therefore, a ConEx-aware node that processes the
   CDO MUST use the Payload Length field of the preceding IPv6 header
   for byte-based counting.  When a ratio is measured and equally sized
   packets can be assumed, counting the number of packets (instead of
   the number of bytes) should deliver the same result.  But an audit
   function must be aware that this estimation can be quite wrong if,
   for example, different sized packed are sent; thus, it is not
   reliable.

   All remaining bits in the CDO are reserved for future use (which are
   currently the last four bits of the eight bit option space).  A ConEx
   sender SHOULD set the reserved bits in the CDO to zero.  Other nodes
   MUST ignore these bits and ConEx-aware intermediate nodes MUST
   forward them unchanged, whatever their values.  They MAY log the
   presence of a non-zero Reserved field.

   The CDO is only applicable on unicast or anycast packets (for
   reasoning, see the note regarding item J on multicast at the end of
   Section 3.3 of [RFC7713]).  A ConEx sender MUST NOT send a packet
   with the CDO to a multicast address.  ConEx-capable network nodes
   MUST treat a multicast packet with the X flag set the same as an
   equivalent packet without the CDO, and they SHOULD forward it
   unchanged.

   As stated in [RFC7713] (see Section 3.3, item N on network-layer
   requirements), protocol specs should describe any warning or error



Krishnan, et al.              Experimental                      [Page 7]

RFC 7837                ConEx Destination Option                May 2016


   messages relevant to the encoding.  There are no warnings or error
   messages associated with the CDO.

5.  Implementation in the Fast Path of ConEx-Aware Routers



   The ConEx information is being encoded into a destination option so
   that it does not impact forwarding performance in the non-ConEx-aware
   nodes on the path.  Since destination options are not usually
   processed by routers, the existence of the CDO does not affect the
   fast-path processing of the datagram on non-ConEx-aware routers,
   i.e., they are not pushed into the slow path towards the control
   plane for exception processing.

   ConEx-aware nodes still need to process the CDO without severely
   affecting forwarding.  For this to be possible, the ConEx-aware
   routers need to quickly ascertain the presence of the CDO and process
   the option if it is present.  To efficiently perform this, the CDO
   needs to be placed in a fairly deterministic location.  In order to
   facilitate forwarding on ConEx-aware routers, ConEx-aware senders
   that send IPv6 datagrams with the CDO SHOULD place the CDO as the
   first destination option in the Destination Option header.

6.  Tunnel Processing



   As with any destination option, an ingress tunnel endpoint will not
   normally copy the CDO when adding an encapsulating outer IP header.
   In general, an ingress tunnel SHOULD NOT copy the CDO to the outer
   header as this would change the number of bytes that would be
   counted.  However, it MAY copy the CDO to the outer header in order
   to facilitate visibility by subsequent on-path ConEx functions if the
   configuration of the tunnel ingress and the ConEx nodes is
   coordinated.  This trades off the performance of ConEx functions
   against that of tunnel processing.

   An egress tunnel endpoint SHOULD ignore any CDO in the outer header
   on decapsulation of an outer IP header.  The information in any inner
   CDO will always be considered correct, even if it differs from any
   outer CDO.  Therefore, the decapsulator can strip the outer CDO
   without comparison to the inner.  A decapsulator MAY compare the two
   and MAY log any case where they differ.  However, the packet MUST be
   forwarded irrespective of any such anomaly, given an outer CDO is
   only a performance optimization.

   A network node that assesses ConEx information SHOULD search for
   encapsulated IP headers until a CDO is found.  At any specific
   network location, the maximum necessary depth of search is likely to
   be the same for all packets between a given set of tunnel endpoints.




Krishnan, et al.              Experimental                      [Page 8]

RFC 7837                ConEx Destination Option                May 2016


7.  Compatibility with Use of IPsec



   A network-based attacker could alter ConEx information to fool an
   audit function in a downstream network into discarding packets.  If
   the endpoints are using the IPsec Authentication Header (AH)
   [RFC2460] to detect alteration of IP headers along the path, AH will
   also detect alteration of the CDO header.  Nonetheless, AH protection
   will rarely need to be introduced for ConEx, because attacks by one
   network on another are rare if they are traceable.  Other known
   attacks from one network on another, such as TTL expiry attacks, are
   more damaging to the innocent network (because the ConEx audit
   discards silently) and less traceable (because TTL is meant to
   change, whereas CDO is not).

   Section 4 specifies that the CDO is placed in the Destination Option
   header before the AH and/or ESP headers so that ConEx information
   remains in the clear if ESP is being used to encrypt other
   transmitted information in transport mode [RFC4301].  In general, a
   Destination Option header inside an IPv6 packet can be placed in two
   possible positions, either before the Routing header or after the
   ESP/AH headers as described in Section 4.1 of [RFC2460].  If the CDO
   was placed in the latter position and an ESP header was used with
   encryption, ConEx-aware intermediate nodes would not be able to view
   and interpret the CDO, effectively rendering it useless.

   The IPv6 protocol architecture currently does not provide a mechanism
   for new headers to be copied to the outer IP header.  Therefore, if
   IPsec encryption is used in tunnel mode, ConEx information cannot be
   accessed over the extent of the ESP tunnel.

   The destination IP stack will not usually process the CDO; therefore,
   the sender can send a CDO without checking if the receiver will
   understand it.  The CDO MUST still be forwarded to the destination IP
   stack, because the destination might check the integrity of the whole
   packet, irrespective of whether it understands ConEx.

8.  Mitigating Flooding Attacks by Using Preferential Drop



   The ideas in this section are aspirational, not being essential to
   the use of ConEx for more general traffic management.  However, once
   CDO information is present, the CDO header could optionally also be
   used in the data plane of any IP-aware forwarding node to mitigate
   flooding attacks.

   Please note that ConEx is an experimental protocol and that any kind
   of mechanism that reacts to information provided by the ConEx
   protocol needs to be evaluated in experimentation as well.  This is




Krishnan, et al.              Experimental                      [Page 9]

RFC 7837                ConEx Destination Option                May 2016


   also true, or especially true, for the preferential drop mechanism
   described below.

   Dropping packets preferentially that are not ConEx-capable or do not
   carry a ConEx mark can be beneficial to mitigate flooding attacks as
   ConEx-Marked packets can be assumed to be already restricted by a
   ConEx ingress policer as further described in [RFC7713].  Therefore,
   the following ConEx-based preferential dropping scheme is proposed:

   If a router queue experiences a very high load so that it has to drop
   arriving packets, it MAY preferentially drop packets within the same
   DiffServ Per-Hop Behavior (PHB) using the preference order given in
   Table 1 (1 means drop first).  Additionally, if a router implements
   preferential drop based on ConEx, it SHOULD also support ECN marking.
   Even though preferential dropping can be difficult to implement on
   some hardware, if nowhere else, routers at the egress of a network
   SHOULD implement preferential drop based on ConEx markings (stronger
   than the MAY above).

                 +----------------------+----------------+
                 |                      |   Preference   |
                 +----------------------+----------------+
                 | Not-ConEx or no CDO  | 1 (drop first) |
                 | X (but not L,E or C) |       2        |
                 | X and L,E or C       |       3        |
                 +----------------------+----------------+

                Table 1: Drop Preference for ConEx Packets

   A flooding attack is inherently about congestion of a resource.  As
   load focuses on a victim, upstream queues grow, requiring honest
   sources to pre-load packets with a higher fraction of ConEx marks.

   If ECN marking is supported by downstream queues, preferential
   dropping provides the most benefits because, if the queue is so
   congested that it drops traffic, it will be CE-marking 100% of any
   forwarded traffic.  Honest sources will therefore be sending 100%
   ConEx E-marked packets (and subject to rate-limiting at an ingress
   policer).

   Senders under malicious control can either do the same as honest
   sources and be rate-limited at ingress, or they can understate
   congestion and not set the E bit.

   If the preferential drop ranking is implemented on queues, these
   queues will reserve E/L-marked traffic until last.  So, the traffic
   from malicious sources will all be automatically dropped first.
   Either way, malicious sources cannot send more than honest sources.



Krishnan, et al.              Experimental                     [Page 10]

RFC 7837                ConEx Destination Option                May 2016


   Therefore, ConEx-based preferential dropping as described above
   discriminates against attack traffic if done as part of the overall
   policing framework as described in [RFC7713].

9.  Security Considerations



   [RFC7713] describes the overall audit framework for assuring that
   ConEx markings truly reflect actual path congestion and [CONEX-AUDIT]
   provides further details on the handling of audit signals.  This
   section focuses purely on the security of the encoding chosen for
   ConEx markings.

   The CDO Option Type is defined with a chg bit set to zero as
   described in Section 4.  If IPsec AH is used, a zero chg bit causes
   AH to cover the CDO option so that its end-to-end integrity can be
   verified, as explained in Section 4.

   This document specifies that the Reserved field in the CDO must be
   ignored and forwarded unchanged even if it does not contain all
   zeroes.  The Reserved field is also required to sit outside the
   Encapsulating Security Payload (ESP), at least in transport mode (see
   Section 7).  This allows the sender to use the Reserved field as a
   4-bit-per-packet covert channel to send information to an on-path
   node outside the control of IPsec.  However, a covert channel is only
   a concern if it can circumvent IPsec in tunnel mode and, in the
   tunnel mode case, ESP would close the covert channel as outlined in
   Section 7.

10.  IANA Considerations



   The IPv6 ConEx destination option is used for carrying ConEx
   markings.  This document uses the experimental option type 0x1E (as
   assigned in IANA's "Destination Options and Hop-by-Hop Options"
   registry) with the act bits set to 00 and the chg bit set to 0 for
   realizing this option.  No further allocation action is required from
   IANA at this time.

11.  References



11.1.  Normative References



   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119,
              DOI 10.17487/RFC2119, March 1997,
              <http://www.rfc-editor.org/info/rfc2119>.






Krishnan, et al.              Experimental                     [Page 11]

RFC 7837                ConEx Destination Option                May 2016


   [RFC2460]  Deering, S. and R. Hinden, "Internet Protocol, Version 6
              (IPv6) Specification", RFC 2460, DOI 10.17487/RFC2460,
              December 1998, <http://www.rfc-editor.org/info/rfc2460>.

   [RFC3168]  Ramakrishnan, K., Floyd, S., and D. Black, "The Addition
              of Explicit Congestion Notification (ECN) to IP",
              RFC 3168, DOI 10.17487/RFC3168, September 2001,
              <http://www.rfc-editor.org/info/rfc3168>.

   [RFC4301]  Kent, S. and K. Seo, "Security Architecture for the
              Internet Protocol", RFC 4301, DOI 10.17487/RFC4301,
              December 2005, <http://www.rfc-editor.org/info/rfc4301>.

   [RFC4302]  Kent, S., "IP Authentication Header", RFC 4302,
              DOI 10.17487/RFC4302, December 2005,
              <http://www.rfc-editor.org/info/rfc4302>.

   [RFC4303]  Kent, S., "IP Encapsulating Security Payload (ESP)",
              RFC 4303, DOI 10.17487/RFC4303, December 2005,
              <http://www.rfc-editor.org/info/rfc4303>.

   [RFC7713]  Mathis, M. and B. Briscoe, "Congestion Exposure (ConEx)
              Concepts, Abstract Mechanism, and Requirements", RFC 7713,
              DOI 10.17487/RFC7713, December 2015,
              <http://www.rfc-editor.org/info/rfc7713>.

11.2.  Informative References



   [CONEX-AUDIT]
              Wagner, D. and M. Kuehlewind, "Auditing of Congestion
              Exposure (ConEx) signals", Work in Progress,
              draft-wagner-conex-audit-02, April 2016.

   [HBH-HEADER]
              Baker, F., "IPv6 Hop-by-Hop Options Extension Header",
              Work in Progress, draft-ietf-6man-hbh-header-handling-03,
              Marcy 2016.

   [RFC7786]  Kuehlewind, M., Ed. and R. Scheffenegger, "TCP
              Modifications for Congestion Exposure (ConEx)", RFC 7786,
              DOI 10.17487/RFC7786, May 2016,
              <http://www.rfc-editor.org/info/rfc7786>.









Krishnan, et al.              Experimental                     [Page 12]

RFC 7837                ConEx Destination Option                May 2016


Acknowledgements



   The authors would like to thank David Wagner, Marcelo Bagnulo,
   Ingemar Johansson, Joel Halpern, John Leslie, Martin Stiemerling,
   Robert Sparks, Ron Bonica, Brian Haberman, Kathleen Moriarty, Bob
   Hinden, Ole Troan, and Brian Carpenter for the discussions that made
   this document better.

Authors' Addresses



   Suresh Krishnan
   Ericsson
   8400 Blvd Decarie
   Town of Mount Royal, Quebec
   Canada

   Email: suresh.krishnan@ericsson.com


   Mirja Kuehlewind
   ETH Zurich

   Email: mirja.kuehlewind@tik.ee.ethz.ch


   Bob Briscoe
   Simula Research Laboratory

   Email: ietf@bobbriscoe.net
   URI:   http://bobbriscoe.net/


   Carlos Ralli Ucendo
   Telefonica

   Email: ralli@tid.es















Krishnan, et al.              Experimental                     [Page 13]