TSVWG Naotaka Morita
Internet-Draft NTT Corporation
Expires: December 22, 2003 Gunnar Karlsson
KTH
June 23, 2003
Framework of Priority Promotion Scheme
draft-morita-tsvwg-pps-00
Status of this Memo
This document is an Internet-Draft and is in full conformance with
all provisions of Section 10 of RFC2026.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that other
groups may also distribute working documents as Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at http://
www.ietf.org/ietf/1id-abstracts.txt.
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.
This Internet-Draft will expire on December 22, 2003.
Copyright Notice
Copyright (C) The Internet Society (2003). All Rights Reserved.
Abstract
This document describes a framework of a new scheme for traffic
control to achieve end-to-end QoS for interactive multimedia
services. The scheme is based on end-to-end measurement of network
resources by end systems. The network is assumed to fully support
the priority control scheme specified in the Diffserv architecture
for QoS and SIP [1] for session control. Since the scheme relies on
the behavior of the end systems, this document also touches on
mechanisms for monitoring end-system behavior.
Conventions used in this document
Morita & Karlsson Expires December 22, 2003 [Page 1]
Internet-Draft Framework of Priority Promotion Scheme June 2003
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC-2119 [2].
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. The target service - Interactive multimedia services . . . . . 4
3. Motivation to focus on an end-system oriented
measurement-based approach . . . . . . . . . . . . . . . . . . 6
4. Basic concepts behind the Priority Promotion Scheme . . . . . 7
5. Variation of specific usage of the Priority Promotion
Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
6. Functional architecture of the Priority Promotion Scheme . . . 10
7. Requirements of the Priority Promotion Scheme . . . . . . . . 11
7.1 Routers . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
7.2 End system . . . . . . . . . . . . . . . . . . . . . . . . . . 12
7.3 SIP proxy . . . . . . . . . . . . . . . . . . . . . . . . . . 12
7.4 Edge router . . . . . . . . . . . . . . . . . . . . . . . . . 12
7.5 Media monitoring server . . . . . . . . . . . . . . . . . . . 13
8. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 14
9. Security Considerations . . . . . . . . . . . . . . . . . . . 15
10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16
11. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 17
References . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . 18
Intellectual Property and Copyright Statements . . . . . . . . 19
Morita & Karlsson Expires December 22, 2003 [Page 2]
Internet-Draft Framework of Priority Promotion Scheme June 2003
1. Introduction
Emerging services such as VoIP, video chat, and video conferencing
require session-based QoS. A few schemes for providing the needed
QoS have been put forward, but they either require per-flow
management of routers within the network or handle the provision of
QoS on a per-class basis by allocating high capacity resources. In
this document a framework for a new QoS scheme is proposed. The
scheme is suitable for session-based interactive multimedia and adds
less complexity to the network than previous approaches, while
delivering per-flow QoS.
Karlsson [3][4] originally proposed this concept. Based on his
ideas, we clarify the requirements to routers, introduce enhancements
to session control using SIP, and show some alternatives to monitor
end-system behavior. We refer to this scheme as the "Priority
Promotion Scheme".
One of the key functions of the Priority Promotion Scheme is routers
behavior. We introduce a new MF-PHB (Measurable Forwarding Per Hop
Behavior) to represent such function. MF-PHB should be verified
whether it is feasible by exiting equipment.
This framework is intended as a guide for device manufacturers,
network administrators and operators who need a way to provide QoS
for Interactive Multimedia services. It is not intended, in its
current state, for use by the majority of networks in the Internet.
The reason that this proposal is being made at this time is that we
feel that the only way to achieve a long-term solution for
inter-domain QoS is to start with practicing on intra-domain
solutions, and incrementally expand the scope of the work as more
experience is gained in deployment.
In this document, we introduce a framework for such a Priority
Promotion. We describe a target service category, which we refer to
as "Interactive Multimedia Services", in section 2. In section 3, we
explain our motivation in focusing on an end-system oriented
measurement based approach. The basic concepts behind the Priority
Promotion Scheme are then explained in section 4. In section 5,
variations of specific usage of the Priority Promotion Scheme are
presented to show the potential of this scheme. The functional
architecture of the scheme is described in section 5, and finally
requirements for the individual functional entities are summarized in
section 6.
Morita & Karlsson Expires December 22, 2003 [Page 3]
Internet-Draft Framework of Priority Promotion Scheme June 2003
2. The target service - Interactive multimedia services
The major targeted services of the Priority Promotion Scheme are for
multimedia and interactive communication software tools running on
PCs, and operated by human being. We call such services interactive
multimedia (IMM) services. The typical examples of IMM are VoIP,
video chat, and video conferencing. IMM services have several
characteristics that differentiate them from existing data services.
Web browsing and some of file retrieval are based on client/server
models and the data transfers speeds required are not so high in
general. On the contrary, IMM services are any-to-any and relatively
high speed at the range of less than 1 Mbps to a few Mbps. These
IMM-inherent characteristics may cause big fluctuations of traffic
patterns and may not be predictable in advance.
Other important characteristics of IMM services are the requirements
for bandwidth guarantees and the real time nature in terms of QoS.
This is because normal codecs are sensitive to the fluctuation of
bandwidth and the degradation of QoS. There are several codecs that
adjust their information rates according to the available bandwidth,
but they impose high processing load on the end system and can never
avoid noticeable and maybe also annoying fluctuation in the
perceptual quality. This is why we need to assume that there will be
bandwidth guarantees. This implies that once the session is
established, the bandwidth is to be guaranteed. In other words, if
the required bandwidth is not available, the session should not be
established. It should be noted that a more extended interpretation
of this concept is that once the bandwidth is guaranteed at a certain
level, it should be maintained until the end of the session.
Improvement is acceptable but degradation is not acceptable.
Finally, IMM services are set up on-demand and may last for a period
of time in the order of minutes.
Taking into account the characteristics or requirements of IMM as
described above, explicit admission control on per-flow basis becomes
necessary. There is an argument that simple over-provisioning is
capable of meeting these kinds of requirements. But, as described
above, IMM has the characteristics of relatively high bandwidth,
unpredictability of traffic pattern and strict QoS needs. Therefore,
we need session based admission control to delivery QoS for IMM
services.
It should be emphasized that admission control has a completely
different goal from existing TCP base functionality. The goal is to
provide bandwidth guarantees with the appropriate QoS for a certain
maximum number of sessions. For example, if the network resource is
100 Mbps and 100 users request sessions with guarantees of 1 Mbps,
Morita & Karlsson Expires December 22, 2003 [Page 4]
Internet-Draft Framework of Priority Promotion Scheme June 2003
nearly 100 sessions should be established. If 1000 users request the
same 1 Mbps guarantees, still only around 100 sessions should be
established. This is quite different from existing data services,
typically operated using TCP. With TCP, the network resources are
shared in a "fair" manner among existing sessions at that time. If
the network resource is 100 Mbps and 100 users request sessions, 100
sessions should be established with 1 Mbps throughput. If 1000 users
request, all 1000 sessions should be established with around 0.1 Mbps
throughput.
SIP has become a suitable way to control IMM services. Although we
focus on SIP in this description, session-control protocols for the
Priority Promotion Scheme are not restricted to it.
The application of a QoS policy, which means any differentiation
based on the identity of a caller or callee in the session, needs to
be studied. There are issues such as the competition between a VIP
call with an ordinary call, or between a preferential call and an
ordinary call in case of a disaster. If such policy is applied along
with simple admission control based on the resource availability,
policy credential information from SIP or other signaling methods may
needs to be incorporate into this framework.
Morita & Karlsson Expires December 22, 2003 [Page 5]
Internet-Draft Framework of Priority Promotion Scheme June 2003
3. Motivation to focus on an end-system oriented measurement-based
approach
As IP-based networks proliferate, the overall network configuration
becomes increasingly complex. In terms of bandwidth available in the
access network, DSL alone includes many variants. 12-Mbps ADSL is
quite popular in Japan and higher speed ADSL services will be
deployed in the near future, but the actual throughput is completely
dependent on conditions that have nothing to do with the access
network, such as the distance from the central office and
interference among lines.
Another point is the variations in the network configurations of
customers, including broadband routers. The broadband routers
initially offered for use with higher-speed access lines may not be
capable of providing maximum throughput. A customer's PC may impose
similar restrictions. The network to which the customer is connected
adds a lot of variables.
In such a complicated situation, end-to-end guarantees of QoS are
difficult to achieve and the role of the end system becomes more
important, because only the end system knows the actual communication
conditions. In the Priority Promotion Scheme, the end systems
measure, monitor, or probe network resources to set up and maintain a
media stream with the required QoS.
We refer to terminal points of the media stream, i.e. PCs or
residential gateways, as end systems.
Morita & Karlsson Expires December 22, 2003 [Page 6]
Internet-Draft Framework of Priority Promotion Scheme June 2003
4. Basic concepts behind the Priority Promotion Scheme
As is described in the previous section, we will take end-system
oriented measurement-based approach.
When it comes to the network, the routers and L2 switches within the
network also have their roles, and the Diffserv architecture is very
popular for these devices. Certain configurations of the priority
mechanism provided by Diffserv give us very simple ways to get
information on the availability of network resources. There are two
examples.
When expedited forwarding (EF) and best effort (BE) are applied on a
certain link and the maximum-rate limit for EF is the same as the
link capacity, the throughput of BE represents the bandwidth
remaining after allocation of bandwidth to the EF traffic.
Let us consider the other example. One assured forwarding (AF) class
with two-drop precedence might be configured as, for example, af11
with a low drop precedence and af12 with a high drop precedence. In
this case, the throughput of af12 represents the bandwidth remaining
after allocation to af11.
We can automatically achieve the required QoS by setting up
communication between the devices that play the two roles above. In
other words, the network provides two priority classes, which share
the same capacity. The end system at the transmitter's side sends
data before or during the first phase of the media stream as
lower-priority traffic, and the other end system, at the receiver
side, reports the condition of the received media to the transmitter
side. The transmitter-side system uses this information to decide
whether it will assign the higher priority to the packets for the
remainder of the media stream, give up on this and send
lower-priority packets, or stop setting up the media stream. The key
is to promote and demote the priority between the monitoring phase
and the real media transmission, or in the very early stage of stream
delivery. Changing the packet priority in this way gives us a way to
measure remaining bandwidth, while maintaining the QoS of established
media streams. If all of the end systems in a network behave this
way, per-flow QoS is inevitably achieved. We refer to this basic
concept as end-system oriented, measurement-based connection/session
admission control.
It should be noted that we are talking about scheduling priority in
the diff-serv scheduler as opposed to call control preference from a
policy perspective or drop preference in a common queue.
The measurement-based approach has many variants. Any of the end
Morita & Karlsson Expires December 22, 2003 [Page 7]
Internet-Draft Framework of Priority Promotion Scheme June 2003
systems-the proxy or home gateway, the edge router at the ingress
point of the network, or the border gateway-might have the role of
measurement entity. The items for measurement might be packet loss
and/or delay, from which we can find the remaining bandwidth.
Explicit indication by the network, e.g. of congestion, is another
possible information to use in forming measurement results. When it
comes to the media, the media characteristics are an important
factor. If the media streams are all constant bit rate, the overall
behavior of the system is quite simple. However, most actual media
streams, particularly video, flow at variable rates.
For the sake of simplicity, we would like to focus on an approach
that is 1) end-system oriented, 2) loss-rate-based, 3) includes no
explicit indications from the network, and has 4) streams which flow,
at constant bit rates, for periods of the order of minutes.
As we previously have noted, the above concept is not new. It was
originally proposed by Karlsson [3][4]. We would like to extend
Karlsson's proposal to meet the needs of a real service.
How we check the end system's behavior is an important point for a
real service. Since the Priority Promotion Scheme is completely
reliant on knowledge of the end system's behavior, incorrect
behavior, whether accidental or intentional, will affect the QoS for
other customers.
One possible solution for this is to introduce two-stage monitoring
of end-system behavior. Primary monitoring may be implemented at the
edge router and is triggered by session initiation. Secondary
monitoring might be done by a dedicated media monitoring server. The
primary monitor checks all media streams that it handles which are
controlled by the Priority-Promotion-Scheme. Example items to check
are whether the flow is allowed to enter the network and whether the
flow is less than the declared peak bit rate. The secondary monitor
checks end-systems behavior in detail. Whether or not the two
monitoring stages are really used will depend on the specific network
environment, but it should be possible to use both to good effect.
Another solution is that such checking mechanism is installed in
every edge routers.
As we describe in the next section, The Priority Promotion Scheme as
described in the following sections is quite suitable for constant-
bit- rate traffic such as voice coded by G.711 without silence
compression, but we should investigate the possibility of tackling
variable bit-rate traffic in future work. Initial investigations
have already been done by Karlsson.
Morita & Karlsson Expires December 22, 2003 [Page 8]
Internet-Draft Framework of Priority Promotion Scheme June 2003
5. Variation of specific usage of the Priority Promotion Scheme
As is described above, the Priority Promotion Scheme can be used as a
kind of admission control. However, it is not limited to the
connection/session admission control as is imagined in the legacy
telephone network. For example, if the initial trial fails, there
are options for the next actions. One is to give up the connection
establishment. This is like the ordinary admission control. Another
one is to stop sending the real media at the low level, but then try
to send it with another class. After a while, the transmitter will
retry and if this try succeeds, the real media is sent with the high
class. Another possibility is that depending on the received
condition information, the transmitter estimates the actual available
bandwidth, selects the closest bandwidth lower than the available
bandwidth and then sends the media with high priority. Another
possible action is to send media at the full rate but only the core
part of the flow is sent with high priority, and the other parts are
sent with low priority. If hierarchical coding is used, this
approach may work well (for example, in MPEG, sending I frames with
high priority and P or B frames with low priority).
Morita & Karlsson Expires December 22, 2003 [Page 9]
Internet-Draft Framework of Priority Promotion Scheme June 2003
6. Functional architecture of the Priority Promotion Scheme
Figure 1 shows the functional architecture of the Priority Promotion
Scheme. The main functional elements are the two end systems, i.e.
the transmitter and receiver, the transmitter-side edge router, core
routers, the SIP proxy, and media-monitoring server.
SIP proxy (Media-monitoring server)
+------+ +------+
/---------------| |------------| |
/ +------+ +------+
/ | //
/ | //
+------+ +------+ +------+ +------+ +------+
| |===============|Edge |======|Core |======| |======| |
+------+ +------+ +------+ +------+ +------+
End system End system
(Transmitter) (Receiver)
Figure 1: Functional architecture of the Priority Promotion Scheme
Morita & Karlsson Expires December 22, 2003 [Page 10]
Internet-Draft Framework of Priority Promotion Scheme June 2003
7. Requirements of the Priority Promotion Scheme
In this section, we describe the requirements for each functional
entity.
7.1 Routers
Although the end systems perform an important role in the Priority
Promotion Scheme, there are a few requirements put on the network.
More specifically, the queuing mechanism or the PHB (per-hop
behavior) for the PPS creates new requirements for network elements.
The Priority Promotion Scheme appears to work with the existing
Diffserv PHB, as was indicated in the introduction. However, to
clearly explain the scheme's requirements, we have to define a new
PHB. We refer to this as measurable forwarding (MF). The essential
requirements for MF are as follows.
* MF has two sub classes, MF-High (MF-H) and MF-Middle (MF-M)
* MF-H and MF-M share the same capacity
* MF-H takes priority over MF-M
In other words, the total amount of MF-H and MF-M traffic is set as a
limit, rather than having separated limits for MF-H and MF-M traffic.
However, since MF-M traffic will always defer to MF-H traffic, MF-M
traffic may experience markedly higher jitter and loss than MF-H; in
fact, one would expect MF-H traffic to experience very nominal jitter
or loss.
Another view of MF is that, if a given amount of MF-M traffic for a
particular stream passes through a router, the same amount of MF-H
traffic for that stream must also be able to pass through.
In the absence of other classes, it appears feasible to configure
existing commercially available routers to produce the desired
MF-PHB. Further requirements are as follows.
1. The MF must co-exist with other PHBs, such as the EF, AF, and
BE. Existing implementations may not be capable of satisfying
this extended requirement.
2. MF should take priority over AF and BE. This is because the
target services are IMM services, where real-time variations
in traffic characteristics are crucially important.
Morita & Karlsson Expires December 22, 2003 [Page 11]
Internet-Draft Framework of Priority Promotion Scheme June 2003
7.2 End system
The transmitter should send trial packets before or at the beginning
of a session.
The receiver should record the results of trial-packet reception and
report this information to the transmitter.
The RTCP would be the best candidate to handle reporting of the
receiving result. Some improvements might be necessary to reduce the
measurement period and to make quick decisions. Actually, the
minimum measurement period is the key factor of the usability of the
Priority Promotion Scheme. This determines the possible service
scenarios as is described in section 5.
The transmitter then decides on the next action.
* If the conditions of reception are good, the transmitter sends
the remaining packets with the higher priority.
* If conditions are not good, the transmitter gives up sending
monitor packets and either 1) sends the remaining packets with
other classes such as BE, 2) stops sending any media data and,
after a while, starts sending monitoring packets again, or 3)
terminates the session.
Synchronization between the two directions of the media stream
remains a subject for further study.
7.3 SIP proxy
In principle, SIP is not directly related to the Priority Promotion
Scheme. However, for commercial applicability, the operator would
have to be able to monitor the service subscription of the customer
before establishing the call. Furthermore, if the edge router has
the capability to monitor the user stream, the SIP proxy can send
commands to the edge router asking it to check up on the end system's
behavior.
The specific signaling sequence may depend on the chosen service
model.
If the policy is applied as is described in the previous section,
signaling is where the policy credentials can get exchanged.
7.4 Edge router
As noted above, in some networks the SIP server is able to instruct
Morita & Karlsson Expires December 22, 2003 [Page 12]
Internet-Draft Framework of Priority Promotion Scheme June 2003
the edge router to monitor the end system's behavior. The edge
router might monitor the following things:
* The transmitter should not send packets at rates above the peak
bit rate offered in the monitoring phase.
* The transmitter should not pause while sending packets. This
is because, if it does this, the other end systems overestimate
the remaining network resources and incorrectly send
higher-priority packets.
7.5 Media monitoring server
In addition to primary monitoring by the edge routers, more detailed
monitoring may be required. The typical items to monitor are as
follows:
* The accuracy of packet-reception information from receivers,
and the correct reaction of transmitters to the information
from receivers should be achieved.
* If the received information indicates poor conditions, the
transmitter stops sending high priority packets. If a next
trial is allowed, a certain time interval should be maintained
between the initial trial and next trial.
Morita & Karlsson Expires December 22, 2003 [Page 13]
Internet-Draft Framework of Priority Promotion Scheme June 2003
8. Conclusion
With the architecture described above, the next step will be a
detailed specification of each relevant functional entity's actions.
Candidates to be specified include;
1. MF-based per-hop behavior;
2. A SIP signaling extension for the Priority Promotion Scheme;
3. The interface between an SIP proxy and edge router;
Although the existing Diffserv architecture may already meet the
requirements of a MF class, there is an urgent need to verify this.
This is because, although new requirements may not seem like much,
the MF PHB is essential to the realization of the Priority Promotion
Scheme.
Other items may be left to each implementation.
Morita & Karlsson Expires December 22, 2003 [Page 14]
Internet-Draft Framework of Priority Promotion Scheme June 2003
9. Security Considerations
To be described.
Morita & Karlsson Expires December 22, 2003 [Page 15]
Internet-Draft Framework of Priority Promotion Scheme June 2003
10. IANA Considerations
To be described.
Morita & Karlsson Expires December 22, 2003 [Page 16]
Internet-Draft Framework of Priority Promotion Scheme June 2003
11. Acknowledgements
The authors would like to thank Fred Baker, David Oran, Glenn Reitsma
and other Cisco technical experts for their insightful suggestions.
Morita & Karlsson Expires December 22, 2003 [Page 17]
Internet-Draft Framework of Priority Promotion Scheme June 2003
References
[1] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, A.,
Peterson, J., Sparks, R., Handley, M. and E. Schooler, "SIP:
Session Initiation Protocol", RFC 3261, June 2002.
[2] Bradner, S., "Key words for use in RFCs to Indicate Requirement
Levels", RFC 2119, BCP 14, March 1997.
[3] Karlsson, G., "Providing Quality for Internet Video Services",
in Proc. of the CNIT/IEEE 10th International Tyrrhenian Workshop
on Digital Communications, Ischia, Italy, September 1998.
[4] Elek, V., Karlsson, G. and R. Ronngren, "Providing Quality for
Internet Video Services", in Proc. IEEE INFOCOM Tel-Aviv,
Israel, March 2000.
Authors' Addresses
Naotaka Morita
Network Service Systems Laboratories
NTT Corporation
9-11, Midori-Cho 3-Chome
Musashino-shi, Tokyo 180-8585
Japan
EMail: morita.naotaka@lab.ntt.co.jp
Gunnar Karlsson
KTH, Royal Institute of Technology
Isafjordsgatan 39
P.O.Box Electrum 229
SE-164 40, Kista
Sweden
EMail: gk@imit.kth.se
Morita & Karlsson Expires December 22, 2003 [Page 18]
Internet-Draft Framework of Priority Promotion Scheme June 2003
Intellectual Property Statement
The IETF takes no position regarding the validity or scope of any
intellectual property or other rights that might be claimed to
pertain to the implementation or use of the technology described in
this document or the extent to which any license under such rights
might or might not be available; neither does it represent that it
has made any effort to identify any such rights. Information on the
IETF's procedures with respect to rights in standards-track and
standards-related documentation can be found in BCP-11. Copies of
claims of rights made available for publication and any assurances of
licenses to be made available, or the result of an attempt made to
obtain a general license or permission for the use of such
proprietary rights by implementors or users of this specification can
be obtained from the IETF Secretariat.
The IETF invites any interested party to bring to its attention any
copyrights, patents or patent applications, or other proprietary
rights which may cover technology that may be required to practice
this standard. Please address the information to the IETF Executive
Director.
Full Copyright Statement
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
Morita & Karlsson Expires December 22, 2003 [Page 19]
Internet-Draft Framework of Priority Promotion Scheme June 2003
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgment
Funding for the RFC Editor function is currently provided by the
Internet Society.
Morita & Karlsson Expires December 22, 2003 [Page 20]