Mobile IP is a technology that allows a "mobile node" (MN) to change
its point of attachment to the Internet while communicating with the
"correspondent node" (CN) using IP. The formal description of Mobile
IP can be found in [1, 6]. Mobile IP primarily addresses the correct
routing of packets to MN's current point of attachment with the
Internet.
It is also essential to provide proper Quality of Service (QoS)
forwarding treatment to the packets sent by or destined to MN as they
propagate along different routes in the network due to node mobility.
This document will identify the requirements that Mobile IP places on
an IP QoS mechanism.
When an MN using Mobile IP undergoes handover from one access router
to another, the path traversed by MN's packet stream in the network
may change. Such a change may be limited to a small segment of the
end-to-end path near the extremity, or it could also have an end-to-
end impact. Further, the packets belonging to MN's ongoing session
may start using a new care-of address after handover. Hence, they
may not be recognized by some forwarding functions in the nodes even
along that segment of the end-to-end path that remains unaltered
after handover. Finally, handover may occur between the subnets that
are under different administrative control.
Chaskar Informational [Page 2]
RFC 3583 Mobile IP QoS Requirements September 2003
In the light of this scenario, it is essential to establish proper
QoS support for the MN's packet stream along the new packet path.
There are four important steps involved in solving the QoS problem
for Mobile IP. They are as follows: (1) List the requirements that
Mobile IP places on the QoS mechanism, (2) Evaluate current IP QoS
solutions against these requirements, (3) Decide if current solutions
need to be extended, or if new ones need to be defined, and (4)
Depending on the result of step 3, define new solutions or fix the
old ones.
Of these, the first step, i.e., the requirements step, is addressed
in this document. The last three steps are not dealt with here in
detail. However, so as to create useful insight into the Mobile IP
QoS problem, at times this document highlights the shortcomings of
some well known current proposals for establishing QoS support for
the packet stream in the network, when directly used with Mobile IP.
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 BCP 14, RFC 2119 [2].
This section describes the requirements for a QoS solution for its
satisfactory operation with Mobile IP. Conversely, note that only
Mobile IP-specific requirements are described here. We do not assume
any particular version (4 or 6) of IP while describing the
requirements. Solutions can be designed for IPv4 and IPv6
independently, or a single solution can be designed to work with both
versions.
In this document, we assume that the target access router for MN's
handover is already pinned down by other protocols. For example,
Seamoby working group has started work on the candidate access router
discovery protocols [7]. Thus, any QoS-capability specific
negotiations that may affect the handover decision are outside the
scope of QoS solution as such, rather need to be performed by
candidate and target access router selection protocols.
Chaskar Informational [Page 3]
RFC 3583 Mobile IP QoS Requirements September 2003
1. Minimize the interruption in QoS at the time of handover:
At the time of handover, interruption in QoS would occur if the
packets sent by or destined to the MN arrive at the intermediate
node in the new packet path without that node having information
about their QoS forwarding requirement. Then, those packets will
receive default forwarding treatment. Such QoS interruption MUST
be minimized. A good metric for this performance is the number of
packets that may potentially get served with the "default" QoS at
the time of handover. The number of such packets MUST be
minimized.
As an example, this performance metric is computed in [8] for the
case of end-to-end RSVP signaling [3] with Mobile IPv6. It is
shown there that when the end-to-end path of packets changes at
large after handover or when the care-of address changes after
handover, OPWA (One Pass With Advertisement) model of reservation
used by RSVP causes the latency of about one round-trip time
between the MN and the CN before QoS can be established along the
new packet path. In other words, the packets using the new care-
of address that would be released by the MN or the CN during one
round-trip time, after these nodes are ready to use the new care-
of address, may get default forwarding treatment at the
intermediate nodes. Such a latency in QoS programming may be
acceptable at the time of session initiation, but it is not
acceptable in the middle of an active session as would be the case
with handover.
2. Localize the QoS (re)establishment to the affected parts of the
packet path in the network:
In many cases, handover changes only a small segment of the end-
to-end path of MN's packet stream near the extremity. Then, the
QoS mechanism MUST limit the extent of QoS (re)establishment to
the affected segment of the end-to-end path only.
However, note that handover may sometimes cause the end-to-end
path of MN's packet stream in the network to change at large.
This may happen, for example, in the case of handover between
different administrative domains. If the QoS mechanism used to
establish QoS support for the MN's packet stream along the new
packet path in the network is based on the explicit end-to-end
provisioning as such, it MUST perform so at the time of such
handover.
Chaskar Informational [Page 4]
RFC 3583 Mobile IP QoS Requirements September 2003
When the care-of address changes upon handover, it may be required
to perform some signaling even over the unchanged part of the
end-to-end path if the path contains any QoS mechanisms that use
IP address as a key to forwarding functions. Examples are FILTER
SPECs in the IntServ nodes or packet classifiers at the edges of
DiffServ networks. However, double provisioning of resources over
the unchanged part of the packet path MUST be avoided.
Note that the QoS signaling protocol such as RSVP [9] can localize
the QoS signaling to the affected parts of the end-to-end path if
the care-of address does not change upon handover. However, if
the care-of address changes upon handover, RSVP as currently
defined [4] fails to localize the QoS signaling. In addition, it
will cause double reservations on the part of end-to-end path that
remains unchanged after handover.
3. Releasing after handover the QoS state (if any) along the old
packet path:
The QoS mechanism MUST provide some means (explicit or timer-
based) to release any QoS state along the old packet path that is
not required after handover. It is desirable that the unwarranted
QoS states, if any, along the old path are released as quickly as
possible at the time of handover. Note that, during handover, the
MN may not always get a chance to send explicit tear down message
along the old path because of the loss of link layer connectivity
with the old access router.
1. Interoperability with mobility protocols:
A number of mobility protocols that are complementary to Mobile IP
are already defined or may be defined in future in IETF,
particularly in Mobile IP and Seamoby working groups. Examples
are fast handover [10, 11], localized mobility management [12,
13], context transfer [5] etc. The QoS mechanism for Mobile IP
SHOULD take advantage of these mobility protocols for the
optimized operation. However, the QoS scheme MUST have provisions
to accomplish its tasks even if one or more of these mobility
protocols are not used.
2. Interoperability with heterogeneous packet paths as regards QoS
paradigms:
The new path after handover, of the MN's packet stream, may
traverse network domains employing different QoS paradigms
compared to those along the old path. The QoS mechanism for
Chaskar Informational [Page 5]
RFC 3583 Mobile IP QoS Requirements September 2003
Mobile IP SHOULD be able to establish proper QoS forwarding
treatment for the MN's packet stream along the packet paths
deploying different QoS paradigms (best current practices), in a
manner consistent with the QoS mechanism deployed along those
paths.
As an illustration, suppose that the MN is currently attached to
an access router which is the edge router of a DiffServ network,
and that the packet classifier and traffic policer for the MN's
flows are presently programmed in this access router. Now,
suppose that the MN needs to be handed over to the access router
which is at the edge of an IntServ network. The new access
network would expect the exchange of RSVP messages so that proper
QoS forwarding treatment can be established for the MN's packet
stream in that access network. QoS mechanism for Mobile IP SHOULD
have provisions to handle such heterogeneity as regards the QoS
mechanisms deployed along different packet paths.
1. QoS support along multiple packet paths:
After MN undergoes handover from one access router to another,
potentially, there could be multiple paths over which MN's packet
may propagate. Examples of these path are: route-optimized path
between the MN and its CN, triangle route via Home Agent (HA),
temporary tunnel between old and new access routers, reverse
tunnel from the new access router (Foreign Agent) to HA etc. A
QoS mechanism SHOULD be able to support QoS along the different
potential packet paths. However, whether all paths are supported
or only a subset of them is supported will be determined by
external mechanisms such as mobility management, policy, location
privacy requirement and so on. Further, the same QoS mechanism
may not be able to support all these paths.
2. Interactions with wireless link-layer support for QoS:
Since a vast number of devices using Mobile IP will be connected
to the Internet via wireless links, the QoS mechanism for Mobile
IP MAY provide some information to the wireless link layers for
them to support the required QoS.
An example scenario that may benefit from such information is that
of the two UDP streams associated with the same media, but
requiring different levels of error protection at the wireless
link layer due to certain characteristics of their respective
encoding schemes. The packets of these two streams are equally
delay sensitive (so as to maintain playout synchronization at the
Chaskar Informational [Page 6]
RFC 3583 Mobile IP QoS Requirements September 2003
receiver), and hence, may be treated equally (as regards queuing)
by IP layer. But they may need to be transmitted on wireless
channels of different error characteristics (say different FEC
coding or power levels).
The QoS information included for the benefit of wireless link
layers SHOULD be such that it is meaningful both ways: to
applications that reside over IP so that they can choose the IP
service of certain QoS characteristics and to wireless link QoS
managers so that they can then map this information to the details
of lower layer mechanisms and their parameters.
In the example scenario described above, such a QoS information
could be expressed as the acceptable loss rate of IP packets in
the UDP stream. This parameter enables the UDP application to
choose the IP service having QoS that matches its requirements,
and it also enables the wireless link QoS managers to choose the
right wireless channel to transmit the packets of this UDP stream.
The QoS solution for Mobile IP SHOULD satisfy standard requirements
such as scalability, security, conservation of wireless bandwidth,
low processing overhead on mobile terminals, providing hooks for
authorization and accounting, and robustness against failures of any
Mobile IP-specific QoS components in the network. While it is not
possible to set quantitative targets for these desirable properties,
the QoS solution MUST be evaluated against these criteria.
The QoS (re)establishment triggered by node mobility MUST be guarded
against security attacks. Such attacks could be launched by
malicious nodes that spoof the QoS signaling to make it appear to the
intermediate nodes that the MN has undergone handover. Such an
attack could disrupt the QoS offered to MN's ongoing sessions as the
intermediate nodes may then tear down the QoS along some segments of
the true packet paths between MN and CN. The malicious nodes may
also request a reduced level of QoS or supply fake packet
classifiers, thereby affecting QoS over some segments (e.g., that do
not get affected by the spoofed handover) of the true packet paths
between MN and CN. Further, network resources may be wasted or used
in an unauthorized manner by the malicious nodes that spoof MN's
handover. To prevent this, QoS mechanism MUST provide means for
intermediate nodes to verify the authenticity of handover-induced QoS
(re)establishment.
Chaskar Informational [Page 7]
RFC 3583 Mobile IP QoS Requirements September 2003
In this document, we described the requirements for a QoS solution
for its satisfactory operation with Mobile IP. The expectation is
that the appropriate working group will use this requirements
document to provide a QoS solution for Mobile IP.
I would like to acknowledge the participants of the mailing list that
was set up to discuss the above requirements. Their contributions
and active participation in the discussion on the mailing list were
very useful in the preparation of this document. Special thanks are
due to, in alphabetical order, Brian Carpenter (IBM), Marc Greis
(Nokia), Glenn Morrow (Nortel), Phil Roberts (Megisto) and Michael
Thomas (Cisco) for their input during the preparation of this
document.
[1] Perkins, C., Ed., "IP mobility support for IPv4", RFC 3344,
August 2002.
[2] Bradner, S., "Key words for use in RFCs to Indicate Requirement
Levels", BCP 14, RFC 2119, March 1997.
[3] Bernet, Y., Ford, P., Yavatkar, R., Baker, F., Zhang, L., Speer,
M., Braden, R., Davie, B., Wroclawski, J. and E. Felstaine, "A
Framework for Integrated Services Operation over Diffserv
Networks", RFC 2998, November 2000.
[4] Braden, R., Ed., Zhang, L., Berson, S., Herzog, S. and S. Jamin,
"Resource ReSerVation Protocol (RSVP) -- Version 1 Functional
Specification", RFC 2205, September 1997.
[5] Kempf, J., Ed., "Problem description: Reasons for performing
context transfers between nodes in an IP Access Network", RFC
3374, September 2002.
[6] Johnson, D., Perkins, C. and J. Arkko, "Mobility support in
IPv6", Work in Progress, May 2003.
Chaskar Informational [Page 8]
RFC 3583 Mobile IP QoS Requirements September 2003
[7] Trossen, D., et al., "Issues in Candidate Access Router
discovery for seamless IP handoffs", Work in Progress, October
2002.
[8] Chaskar, H. and R. Koodli, "QoS support in Mobile IP version 6",
IEEE Broadband Wireless Summit (Networld+Interop), May 2001.
[9] Thomas, M., "Analysis of Mobile IP and RSVP interactions", Work
in Progress, February 2001.
[10] MIPv4 Handoffs Design Team, "Low latency handoffs in Mobile
IPv4", Work in Progress, June 2002.
[11] Koodli, R., Ed., "Fast handovers for Mobile IPv6", Work in
Progress, March 2003.
[12] Williams, C., Ed., "Localized mobility management requirements",
Work in Progress, March 2003.
[13] Soliman, H., et al., "Hierarchical MIPv6 mobility management
(HMIPv6)", Work in Progress, October 2002.
The working group can be contacted via the current chair:
John Loughney
Nokia Research Center
11-13 Italahdenkatu
00180 Helsinki
Finland
EMail: john.loughney@nokia.com
Questions about this memo can be directed to the editor:
Hemant Chaskar
Nokia Research Center
5 Wayside Road
Burlington, MA 01803, USA
Phone: +1 781-993-3785
EMail: hemant.chaskar@nokia.com
Chaskar Informational [Page 9]
RFC 3583 Mobile IP QoS Requirements September 2003
Copyright (C) The Internet Society (2003). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assignees.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFC Editor function is currently provided by the
Internet Society.
Chaskar Informational [Page 10]