Internet Draft Cellular RAN Resource Issues June 2002
Internet Engineering Task Force D. Partain (ed)
INTERNET-DRAFT G. Karagiannis
Expires December 2002 P. Wallentin
L. Westberg
Ericsson
June 2002
Resource Reservation Issues in Cellular Radio Access Networks
draft-westberg-rmd-cellular-issues-01.txt
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.
Abstract
This memo describes resource management issues that are relevant to
the use of IP transport in cellular radio access networks (RANs).
The document describes the particular characteristics of these kinds
of networks, the requirements applicable to a resource reservation
scheme in a cellular RAN, and provides a brief analysis of the
applicability of existing solutions to this problem space.
D. Partain, ed. Expires December 2002 [Page 1]
Internet Draft Cellular RAN Resource Issues June 2002
Table of Contents
1 Introduction ................................................. 3
2 Terminology .................................................. 4
3 Background and motivation .................................... 4
3.1 IP transport in radio access networks ...................... 4
3.2 Motivation for this memo ................................... 8
4 Network characteristics of cellular access networks today .... 9
4.1 General aspects of the network structure ................... 9
4.2 High cost for transmission ................................. 10
4.3 Transportation of radio frames ............................. 11
4.4 Mobility aspect of radio frame transport ................... 13
5 Requirements on a Resource Reservation Scheme ................ 13
5.1 Main requirement on resource reservation scheme ............ 13
5.2 IP must provide same service behavior as the transport
networks used today ....................................... 14
5.3 Efficient network utilization .............................. 14
5.4 Handover performance requirements on resource reservation
scheme
...................................................... 15
5.5 Edge-to-edge reservations, not end-to-end .................. 16
5.6 Reservation functionality in edge nodes versus interior
routers ................................................... 17
5.7 On-demand and dynamic allocation of resources .............. 18
5.8 Unicast and not multicast .................................. 18
5.9 A single operator in the RAN ............................... 19
5.10 Minimal impact on router performance ...................... 19
5.11 Scalably Manageable ....................................... 20
5.12 Bi-directional reservations ............................... 20
5.13 Support for non-RAN specific traffic ...................... 20
6 Evaluation of existing strategies ............................ 21
6.1 End-to-end per-flow resource reservation protocol .......... 21
6.2 Integrated Services over Differentiated Services ........... 23
6.3 Statically-assigned trunk reservations based on DifferenÂ
tiated Services ........................................... 25
6.4 Dynamic trunk reservations with Aggregated RSVP ............ 25
7 Conclusion ................................................... 27
8 References ................................................... 28
9 Acknowledgements ............................................. 29
10 Authors' Addresses .......................................... 29
D. Partain, ed. Expires December 2002 [Page 2]
Internet Draft Cellular RAN Resource Issues June 2002
1. Introduction
The rapidly growing popularity of IP and its flexibility make it a
good candidate to be used for transmission in cellular networks.
Using IP-based transport on the wired transmission links in the
cellular networks gives operators an opportunity to upgrade their
transport network to a packet-based one. When compared with a
traditional STM-based system, the gain is seen in the statistical
aggregation of traffic that can be done. This results in increased
transmission efficiency and reduced leasing cost for the operator.
A radio access network (RAN) provides the radio access (e.g., GSM,
CDMA, or WCDMA) to mobile stations in a cellular network. To
accomplish this, radio frames are transported on the wired links
between different cellular-specific nodes in the RAN. The majority of
the traffic (up to 100%) is delay-sensitive traffic.
The cellular user is unaware of the IP-based transport network
underneath, and the service must work the same way as the user has
come to expect the cellular services to work in an STM-based
transport network. In addition to this requirement, the situation is
further complicated by the fact that the RAN is large in terms of its
geographic size, the number of inter-connected nodes, and the
proportion of real-time traffic.
To satisfy the above requirements, it is absolutely critical that we
have a simple and scalable bandwidth resource management scheme for
real-time traffic in this type of network.
In order for real-time services to function satisfactorily in an IP-
based RAN, we need to ensure that there are adequate transport
resources on the links available in the RAN to handle this particular
instance of the service (e.g., a phone call). Note that, in rest of
this draft, whenever the term "resources" is used, it refers to
bandwidth on the links.
If the RAN is bandwidth-limited and does not use any mechanism to
limit the usage of the network resources, congestion might occur and
degrade the network performance. For example, speech quality might
degrade due to packet losses.
D. Partain, ed. Expires December 2002 [Page 3]
Internet Draft Cellular RAN Resource Issues June 2002
2. Terminology
The following terminology is used in this memo:
* BSC: Base Station Controller
* RNC: Radio Network Controller
* MSC: Mobile Services Switching Center
* GGSN: Gateway GPRS Support Node
* SGSN: Serving GPRS Support Node
* GPRS: General Packet Radio Service, the packet-switched access
scheme and service provided in GSM.
* GSM: Global System for Mobile Communications
* UMTS: Universal Mobile Telecommunications System, the third
generation (3G) mobile system based on WCDMA and GSM,
specified by 3GPP (third generation partnership project).
* radio frame: a short data segment coded/decoded and
transmitted/received by the radio base station.
It originates from a mobile station or the BSC/RNC.
* WCDMA: Wideband Code Division Multiple Access, the
radio transmission technology used in UMTS
3. Background and motivation
The context of the issues described in this document is the cellular
radio access network (RAN). This section briefly discusses two
examples of radio access networks (for GSM - Global System for Mobile
Communication - and for WCDMA - Wideband Code Division Multiple
Access) and then outlines the motivation for this memo.
3.1. IP transport in radio access networks
This section introduces the radio access network and its use of IP
transport. A radio access network (RAN) provides the radio access
D. Partain, ed. Expires December 2002 [Page 4]
Internet Draft Cellular RAN Resource Issues June 2002
(e.g., GSM, CDMA, or WCDMA) to mobile stations for a cellular system.
The boundaries for a RAN are the radio transmission and reception
access points (terminated by base stations) at one end and, at the
other end, the interfaces to the gateways (e.g., MSC and SGSN/GGSN),
which in turn provide connections to the fixed public network.
The radio access network consists of a number of nodes as shown in
the Figure 1 below.
IP IP
<---> <--->
|---------| |---------| |---------|
Mobile v | Base | | | | MSC/ | Fixed
stations |--| station |-----| BSC/RNC |-----| SGSN/ |--- public
^ | | ^ | | ^ | GGSN | network
| |---------| | |---------| | |---------|
| | |
Wireless Wired Wired
interface interface interface
Radio access network
<------------------------------->
Figure 1: Typical radio access network and its boundaries
The base station provides the radio channel coding/decoding and
transmission/reception function to and from mobile stations in its
coverage area, which is called a cell.
The BSC/RNC controls a number of base stations including the radio
channels and the connections to mobile stations. For a WCDMA radio
access network, the BSC/RNC provides soft handover combining and
splitting between streams from different base stations belonging to
the same mobile station. Furthermore, the BSC/RNC is also
responsible for the allocation of transport resources within the
radio access network. The transport is either between the base
station and the BSC/RNC, between multiple BSC/RNCs, or between the
BSC/RNC and the MSC/SGSN.
The MSC provides, among others things, support for circuit-switched
services towards mobile stations including mobility management,
access control and call control as well as interworking with external
circuit-switched networks such as the public switched telephony
D. Partain, ed. Expires December 2002 [Page 5]
Internet Draft Cellular RAN Resource Issues June 2002
network (PSTN). The SGSN/GGSN provide, amongst other things, support
for packet switched services towards mobile stations, including
mobility management, access control and control of packet data
protocol contexts. In addition, the GGSN provides interworking with
external packet-switched networks such as the public Internet.
The radio access network consists of potentially thousands of base
stations and a significant number of BSCs/RNCs. The traffic volume,
in terms of voice-traffic, generated by these nodes can vary from a
few up to fifty voice calls per base station, and up to several
thousand simultaneous calls (Erlang) per BSC/RNC site. Therefore, a
router in the network has to handle many thousands of simultaneous
flows.
The transmission between base stations and the BSC/RNC is usually on
leased lines, and this part (due to the wide area coverage of the
cellular network) is usually extremely expensive when compared to the
cost of transmission in the backbone. Due the number of base
stations, the cost for these leased lines can be quite significant.
Dimensioning using over-provisioning might therefore be prohibitively
expensive, and it is unlikely that the network will be dimensioned
without using the statistical properties of traffic aggregation
(e.g., Erlang trunking).
D. Partain, ed. Expires December 2002 [Page 6]
Internet Draft Cellular RAN Resource Issues June 2002
|--------|
| Upper |<--------------------------------------------> towards the
| layers | MSC/SGSN/GGSN
|--------| |---------|--------|
| Radio | | Radio | |
| Link |<----------------------------------->| Link | |
| Layer | | Layer | |
|--------| |---------| |
| | |Radio | |<-
| | |Physical | Frame |
| | |Layer | transp.|
| | |-------|--------| |---------| layer |
| | | | Frame |<-------------->| Frame | |
| | | | transp.| | transp. | |
|Radio | | Radio |--------| |---------|--------|
|Physical| | Phys. | UDP |<-------------->| UDP | UDP |
|Layer |<->| Layer |--------| |--------| |---------|--------|
| | | | IP |<->| IP |<->| IP | IP |
| | | |--------| |--------| |---------|--------|
| | | | Link | | Link |<->| Link | Link |
| | | | Layer |<->| Layer | | Layer | Layer |
| | | |--------| |--------| |---------|--------|
| | | |Physical| |Physical| |Physical |Physical|
| | | | Layer |<->| Layer |<->| Layer | Layer |
|--------| |-------|--------| |--------| |------------------|
Mobile Base station Router BSC/RNC
Station ^ ^ ^
| | |
Wireless Wired Wired
interface interface interface
Radio access network
<----------------------------------------------------->
Figure 2: Example of a protocol stack in the
radio access network (simplified)
Figure 2 shows a simplified example of protocol layering when using
IP transport in the radio access network.
The radio physical layer performs radio transmission and reception
functions, including soft handover splitting and combining in case of
WCDMA.
The frame transport layer is used to transmit radio frames between
D. Partain, ed. Expires December 2002 [Page 7]
Internet Draft Cellular RAN Resource Issues June 2002
the base station and the BSC/RNC. A radio frame is a short data
segment coded/decoded and transmitted/received by the radio base
station at a given point in time. The radio frames must be delivered
in a timely fashion with limited delay. Otherwise, the frames are
discarded by the base station or RNC/BSC. The traffic is therefore
very sensitive to delays.
The radio link layer performs segmentation/re-assembly,
retransmission and multiplexing/scheduling functions as well as radio
resource control. The adaptation of user data performed by the radio
link layer depends on the type of radio channel and the type of
service. In one case, that very small radio frames might be
transferred, while in other cases the packets are significantly
larger.
Introducing IP in the radio access network implies that an IP QoS-
capable domain, e.g. a Differentiated Services domain, will have to
be introduced and managed in the radio access network. This domain
consists of edge and interior nodes, where the edge nodes are the
nodes located at the boundary of the domain. All the nodes which are
part of this QoS-capable domain and are not edge nodes are defined as
interior nodes.
An edge node can be defined as an ingress node, or a node that
handles the traffic as it enters the QoS-capable domain.
Alternatively, an edge node might be an egress node, or a node that
handles the traffic as it leaves the QoS-capable domain. In this
memo, an edge node (ingress or egress) is denoted as the first hop
router that the base station or BSC/RNC is connected to. The first
hop router might be a part of the base station or BSC/RNC.
Furthermore, the base station and BSC/RNC must be able to handle
algorithms used for purposes other than edge node functionality that
are many times more complex than the algorithms required for handling
the edge node functionality. Therefore, the edge node functionality
will only have minimal impact on the complexity of the base station
or BSC/RNC.
3.2. Motivation for this memo
The issues described in this document concern only the cellular radio
access network (RAN).
D. Partain, ed. Expires December 2002 [Page 8]
Internet Draft Cellular RAN Resource Issues June 2002
The architecture of the RAN and the nature of the transported data
mean that the RAN has different characteristics when compared with
other IP-based networks. However, those differences, which are the
motivation for this memo, are limited to the domain of the RAN and do
not extend into the backbone of the IP network.
In order for the transport within the RAN to function satisfactorily,
even if the transport network is IP-based, we need to ensure that
there are adequate resources in the transport network to meet the
needs of the data flows between the nodes within the RAN.
Based upon the characteristics of the RAN (described in Section 4
below), the current strategies for resource management do not meet
the requirements for an appropriate resource management strategy
within a RAN. This document seeks to initiate a dialog on how to
correct that situation.
4. Network characteristics of cellular access networks today
Cellular RANs today have a unique set of characteristics compared to
other kinds of IP networks. These characteristics result in a set of
requirements on any resource reservation scheme that might be used in
the RAN.
4.1. General aspects of the network structure
The network structure for cellular radio access networks can be
described as having the following characteristics:
* Operator relationship
The RAN is typically controlled by a single cellular operator
with full control over the network. The IP network used to
transport radio frames might be leased from another operator
or be built by the same cellular operator. This network
could be thought of as an "intranet".
* Size of network
RANs can be very large routed networks. Networks including
thousands of nodes are certainly within reason.
D. Partain, ed. Expires December 2002 [Page 9]
Internet Draft Cellular RAN Resource Issues June 2002
* Traffic volume
The traffic from a large number of radio base stations
needs to be supported by the same transport network. Even
if a single radio base station generates a modest volume of
traffic, the total number of flows for radio frame transport
in the radio access network is very large.
* Transmission sharing
The network between the BSC/RNC and the base stations is
built to support transmission sharing between different
nodes even if they are geographically distributed.
One transmission link (possibly with redundancy) can
support more than one base station. In other words, one
piece of hardware can serve more than one base station and
therefore can support more cells in one location, such as
a three sector site. This means that the cells that are
located at the same location will have to compete for the
same transmission resources.
* Unicast transport of radio frames
The transport of radio frames in the radio access network is
point-to-point transport. Even if the soft handover splitting
in the BSC/RNC is multicast of radio frames in some sense,
this is handled above the IP layer by the frame transport
protocol. For each radio channel in each base station the
frame transport protocol needs a separate flow. Therefore,
the frame transport protocol requires unicast transmission
from the IP layer.
4.2. High cost for transmission
The transmission between base stations and the BSC/RNC is usually on
leased lines, and this part (due to the wide area coverage of the
cellular network) is usually extremely expensive when compared to the
cost of transmission in the backbone. Due to the number of base
stations, the cost for these leased lines can be quite significant.
Even if the cost for the leased line decreases over the years, the
"last mile" to the base station is likely to be expensive due to the
location of the base station.
Cellular RANs are built over a very wide geographic area. There are,
D. Partain, ed. Expires December 2002 [Page 10]
Internet Draft Cellular RAN Resource Issues June 2002
of course, many different networks that cover a wide geographic area
(e.g., across USA, and the world), but the dispersion of nodes over
the area in the RAN case is different. Due to the fact that the base
stations are positioned based on a radio network perspective, i.e.,
radio coverage, and not based on a transmission perspective, a large
proportion of nodes are distributed throughout rural and urban areas,
not close to installed high-capacity transmission hubs. Even worse,
the the base stations could be positioned far out in the countryside.
The peak bitrate of the multirate radio channels is selected on-
demand. To utilize the bandwidth of the expensive transmission links
used for radio frame transport efficiently, dimensioning using over-
provisioning might therefore be prohibitively expensive, and it is
unlikely that the network will be dimensioned for peak allocation.
Dynamic allocation and optimization to reduce the cost are therefore
a fundamental requirement. Resource reservations make it possible to
have high utilization of the network for real-time sensitive traffic
as well as avoiding congestion in the network.
4.3. Transportation of radio frames
The traffic in cellular access networks is dramatically different
from the Internet in general. The Internet primarily supports best-
effort traffic today, while the traffic on a RAN is (at least today)
largely real-time traffic. The network from the base station to the
BSC/RNC is the part of the network that has the highest volume of
real-time traffic and where delay must be minimized as much as
possible. The reasons for the this are:
* End-to-end delay for speech traffic consists of delays in the
mobile stations, the RAN and in the MSC (see Figure 1 in
Section 3.1). The major portion of delay in the RAN is caused
by the radio-related functionality (e.g., interleaving and
coding in the base station and adaptation in the BSC/RNC).
Therefore, the combined delay in all parts (MSC, radio
network, and mobile stations) must be minimized as much
as possible to give the end user proper speech quality.
* Handover is a major issue. For GSM, with typically multiple
handovers per call, excessive delay in the control, e.g.
radio network control traffic, of the radio network will
cause a longer handover interruption period. The majority
of handovers are also made within the radio network.
D. Partain, ed. Expires December 2002 [Page 11]
Internet Draft Cellular RAN Resource Issues June 2002
* The transport of radio frames is very delay-sensitive. In
the direction from the BSC/RNC to the base station, a radio
frame is a short segment of data (payload) to be coded and
transmitted on a given radio channel by the base station at a
given point in time. In the direction from the base station
to the BSC/RNC, a radio frame is a short segment of data
(payload) that was received and decoded by the base station
at a given point in time and potentially needs to be combined
in the BSC/RNC with radio frames received by other base
stations at the same point in time as this particular frame.
Note that the data segment in a radio frame may contain user
data but also control signalling information and the same
type of synchronized frame transport is needed for almost
all kinds of radio channels and is generally not coupled
to the type of service. Therefore, even if an end to end
application is best effort, the transport of the radio frames
originating from this application might be treated as real-time
within the radio access network.
The real-time traffic on the RANs is today almost exclusively voice
(up to 90% with 10% signalling), but the cellular systems are
evolving to provide capabilities for other kinds of real-time traffic
(e.g., video). Nonetheless, voice continues to be one of the most
important sources of revenue in most cellular environments today. The
transport resources are today allocated when the call is accepted,
and the radio frame transport over an IP network has to provide the
same guarantees. If real-time traffic cannot be engineered to work
correctly, the primary revenue stream will disappear.
Some of the sources, such as video-based services and gaming, will be
able to send data at a variable bitrate at higher rates. For radio,
the rates of the radio channels are selected on-demand. In reality,
the radio network can support a wide range of partitioning of the
radio resources among the different radio channels. A rather large
portion of the transmission resources between the base station and
BSC/RNC will have to be allocated for such services. To be able to
utilize the bandwidth used for radio frame transport efficiently, the
same flexibility is required in assignment of the transport resources
as in the air-interface. Therefore, statically-assigned resources
will induce a cost which is too high for the operator.
D. Partain, ed. Expires December 2002 [Page 12]
Internet Draft Cellular RAN Resource Issues June 2002
4.4. Mobility aspect of radio frame transport
The mobility of the mobile stations imposes strong requirements on
managing the transmission resources available in the RAN,
Furthermore, this also implies that there are strong requirements on
the RAN's internal signaling and not only on transferring of packets
sent by the mobile station.
Hard handover is one of the issues. For GSM, with typically multiple
handovers per call, excessive delay in the control of the radio
network will cause a longer handover interruption period. Typically,
most of the handovers will be made between base stations controlled
by the same BSC and therefore extensive delay between base station
and BSC will degrade more than delays in the MSC and SGSN/GGSN
network.
Moreover, for maximal utilization of radio spectrum in WCDMA (and
also in CDMA), fast and frequent (soft) handover operations between
radio channels and radio base stations are required. The frequency of
handover events is therefore typically higher in WCDMA radio access
networks than in GSM and means even higher performance requirements
on the transport solution. If the soft handover cannot be performed
fast enough, spectrum cannot be utilized efficiently, which will
cause degradation of the radio network capacity. At each handover
event, resource reservation is needed, and therefore resource
reservation needs to be fast and will be used very frequently.
The impact of mobility in the radio access network has therefore two
major differences compared to the fixed network:
(1) High volume of resource reservation events
(2) Requirement on short response time for reservations
5. Requirements on a Resource Reservation Scheme
This section outlines what we believe are the fundamental
requirements placed on any resource reservation scheme in a cellular
radio access network. Later sections will outline how current schemes
match these requirements.
5.1. Main requirement on resource reservation scheme
One of the primary requirements that real-time applications impose on
D. Partain, ed. Expires December 2002 [Page 13]
Internet Draft Cellular RAN Resource Issues June 2002
any resource reservation scheme is the provisioning of good QoS
(delay and packet loss) guarantees. This can only be achieved if the
network can be utilized while avoiding congestion and without having
too high packet losses. The level of utilization depends on network
topology, traffic mix, scheduling discipline, delay and packet loss
requirements. The utilization is given by network dimensioning but
should be as high as possible.
The resource reservation scheme must be able to keep the real-time
traffic under a certain pre-defined network utilization in order to
bound congestion. Otherwise, it is impossible to guarantee percentile
bounds on the QOS requirements for real-time traffic.
5.2. IP must provide same service behavior as the transport networks
used today
Today's commercial IP networks are mainly optimized to carry best-
effort traffic. As explained above and also discussed in [WeLi99],
the transport of radio frames in the radio access network puts real-
time requirements on the underlying transport network. All of these
characteristics are fulfilled by the connection-oriented transport
networks (STM and ATM) used by cellular networks today. By, at a
minimum, meeting these same requirements, the IP networks will be
capable of providing the same behavior as the transport networks that
are currently used by cellular systems while gaining the advantages
of IP networking. It should be noted that IP networks will be able to
meet these requirements only if the following two constraints are
met:
(1) that service guarantees are percentiles, see Section 5.1
(2) strictly limited to a given operator's IP network, see
Section 5.9.
5.3. Efficient network utilization
Due to the high cost of the leased transmission, we must utilize the
network to the highest degree possible, and this must be facilitated
by the resource reservation scheme.
D. Partain, ed. Expires December 2002 [Page 14]
Internet Draft Cellular RAN Resource Issues June 2002
However, in considering a resource reservation scheme, its impact
upon the performance and scalability of the network as a whole must
also be taken into account. For example, the performance and
scalability impact on the edge and internal routers is a very
important consideration.
5.4. Handover performance requirements on resource reservation scheme
Whatever reservation scheme is used must be highly performant for at
least the following reasons:
* Handover rates
In the GSM case, mobility usually generates an average
of one to two handovers per call. For third generation
networks (such as WCDMA and cdma2000), where it is
necessary to keep radio links to several cells
simultaneously (macro-diversity), the handover rate is
significantly higher (see for example [KeMc00]).
Therefore, the admission control process has to cope
with far more admission requests than call setups alone
would generate.
* Fast reservations
Handover can also cause packet losses. If the processing
of an admission request causes a delayed handover to the
new base station, some packets might be discarded, and
the overall speech quality might be degraded
significantly.
Furthermore, a delay in handover may cause degradation
for other users. This is especially true for radio access
technologies using macro-diversity, such as WCDMA and CDMA,
where a handover delay will cause interference for other
users in the same cell. Furthermore, in the worst case
scenario, a delay in handover may cause the connection
to be dropped if the handover occurred due to bad radio
link quality.
Therefore, it is critical that an admission control
D. Partain, ed. Expires December 2002 [Page 15]
Internet Draft Cellular RAN Resource Issues June 2002
request for handover be carried out very quickly. Since
the processing of an admission control request is only
one of many tasks performed during handover, the time to
perform admission control should be a fraction of the
time available for handover.
Furthermore, in the situation that the transport network
in the RAN is over-utilised it is preferable to keep
the reservation on already established flows while new
requests might be blocked. Therefore, the handover
requests for resource reservation should be treated
with a higher priority than the new requests for
resource reservation.
5.5. Edge-to-edge reservations, not end-to-end
Real-time applications require a high level of quality of service
(QoS) from the underlying transmission network. This can only be
achieved by accomplishing the QoS management on an end-to-end basis
(i.e., end user to end user), from application to application,
potentially across many domains.
However, this does not mean that the resource reservation protocol
must be applied end-to-end. The end-to-end QoS management
architecture may consist of many interoperable edge-to-edge QoS
management architectures where each of them might use a different
edge-to-edge resource reservation protocol. In fact, this is far
more likely to be the case than that a global signaling structure
will be available across all different domains in an end-to-end
perspective. This will increase the flexibility and the openness of
the transmission network since various access networks that are using
the same transmission network and different edge-to-edge QoS
management architectures will be able to interoperate.
It is critical that the appropriate mechanisms for providing the
service guarantees needed in the radio access network be put in place
independently of solving the more difficult problem of end-to-end
QoS.
In our case, the access network is simply an intranet in which we
D. Partain, ed. Expires December 2002 [Page 16]
Internet Draft Cellular RAN Resource Issues June 2002
need to solve a local QoS problem. This implies that a general
solution which handles the end-to-end QoS problem is unnecessarily
complicated for solving the intranet problem in the cellular access
network.
5.6. Reservation functionality in edge nodes versus interior routers
In our network, it is important that the reservation mechanism be as
simple as possible to implement in the interior nodes since in most
cases there might be more interior routers (<= 10 depending on
network structure) in the path than there are edge nodes.
Typically, in a RAN there are two edge nodes located in a
communication path. Moreover, the average number of interior nodes in
a communication path within a RAN depends on the chosen network
topology by the RAN operator. As such, the scheme must be optimized
for the interior nodes and not for the edge nodes, thus reducing the
requirements placed on the functionality of the interior routers.
This means that we can have complicated mapping of traffic parameters
at the edges and a simplified model in the interior nodes, and that
the necessary set of parameters required for setting up reservations
shall be based upon their effect on interior nodes and not on edge
nodes.
The edge routers typically have to perform per-session
management/control, and hence complex per flow handling is not a
significant burden.
However, interior nodes do not need to have per flow
responsibilities. We must therefore optimize for simple QoS
mechanisms on these interior devices, and use more complex mechanisms
in the edge devices.
In our case, edge device functionality is implemented in the first
hop router that the base station or BSC/RNC is connected to (see
Section 3.1). In this way we optimize for simple QoS mechanisms on
the interior devices, while the more complex mechanisms are applied
on the edge devices, e.g., base station, BSC/RNC.
This emphasis on simplicity is due to performance requirements listed
above. We need to make sure that we understand the minimal level of
functionality required in the reservation scheme in order to
D. Partain, ed. Expires December 2002 [Page 17]
Internet Draft Cellular RAN Resource Issues June 2002
guarantee the performance of real-time traffic.
5.7. On-demand and dynamic allocation of resources
Real-time services require that a portion of network resources is
available to them. These resources can be reserved on a static or
dynamic basis, or potentially based on some kind of measurement of
network load.
In the first situation, this may result in an poorly utilized
network. This is mainly due to the fact that the network resources
are typically reserved for peak real-time traffic values. Mobility
in the network makes static configuration even less desirable as the
resources will be used even less effectively.
If using dynamic allocation, this problem will be avoided since the
resources are reserved on demand. However, the load from resource
reservation will be much higher than if static allocation of
resources is used. If the dynamic allocation of the resources is
done on a micro-flow basis, the resulting network load from resource
reservation might be quite high.
We might use other methods, such as measurement-based admission
control, to simplify the reservation protocol, as long as these
methods can fulfill the requirements (now or in the future).
What is important is that all of these mechanisms can be used for
solving part of the network utilization problem, and, as such, any
reservation scheme must have the flexibility to provide both on-
demand reservations as well as measurement-based admission control.
As high bitrate and variable bitrate applications enter the cellular
space, the need for on-demand reservations of resources will become
even more acute.
5.8. Unicast and not multicast
The majority of the traffic in the RAN is point-to-point unicast
transport of radio frames between the base station and the BSC/RNC.
As such, the resource reservation scheme need not to be optimized for
D. Partain, ed. Expires December 2002 [Page 18]
Internet Draft Cellular RAN Resource Issues June 2002
multicast.
5.9. A single operator in the RAN
It is realistic to assume that end-to-end communication in IP
networks as well as the end-to-end QoS management architectures will
be managed by more than one operator.
Furthermore, it also realistic to assume that an edge-to-edge
resource reservation protocol can be managed by one single operator.
As such, it is reasonable to limit reservation scheme to a single
operator domain. This will ensure that each operator can optimize the
edge-to-edge QoS management architecture for their needs. Moreover,
this limitation (a single operator domain) means that the reservation
scheme does not need to handle the issues inherent in a multi-
operator domain, thus simplifying the scheme.
5.10. Minimal impact on router performance
The performance of each network node that is used in an end-to-end
communication path has a significant impact on the end-to-end
performance of this communication path. Therefore, the end-to-end
performance of the communication path can be optimized by optimizing
the router performance. It is absolutely critical that the
introduction of QoS mechanisms and signaling does not overly impact
the performance of the infrastructure. Obviously, you cannot
introduce new things that need to be done by networking
infrastructure without impacting its performance, but that impact
must be minimized to the greatest extent possible.
One of the factors that can contribute to this optimization is the
minimization of the resource reservation signaling protocol load on
each router. When the dynamic allocation of the resources is on a per
micro-flow basis, the resource reservation signaling protocol could
easily overload a router located in a core network, causing severe
router performance degradation. Furthermore, any mechanisms defined
must be such that it is reasonable to implement them in hardware
which will increase the scalability of the solution.
D. Partain, ed. Expires December 2002 [Page 19]
Internet Draft Cellular RAN Resource Issues June 2002
5.11. Scalably Manageable
Any strategy for resource management in a RAN must be done in such a
way that it is easily manageable in a very large network. This
implies as little "laying on of hands" as possible and as much
automation as possible. In networks made up of many thousands of
routers, changing of even a single parameter in all routers may be
prohibitively difficult. Minimizing the involvement of the operator
(or the operator's management tools) is therefore an important
requirement.
5.12. Bi-directional reservations
In current RANs, the BSC/RNC is responsible for initiation of
reservation of resources in the transport plane. Therefore, via the
resource reservation signaling protocol, the BSC/RNC has to support
the initiation and management of the resource reservations for both
directions, both to and from the base station, simultaneously. In
this way a simpler edge-proxy resource reservation functionality will
be implemented in the base station, decreasing its complexity.
5.13. Support for non-RAN specific traffic
Any strategy for resource management used in a cellular RAN must be
able to support any type of traffic (RAN-specific or non-RAN
specific) in the same way, as long as the traffic belongs to the same
traffic class.
The RAN-specific traffic is the traffic that is transported through
the RAN and is generated or used by specific entities belonging to
the same cellular technology as the one used in the RAN.
The non-RAN specific traffic is the traffic that is transported
through the RAN but is neither generated by nor used by any specific
entity belonging to the same cellular technology as the one used in
the RAN.
D. Partain, ed. Expires December 2002 [Page 20]
Internet Draft Cellular RAN Resource Issues June 2002
6. Evaluation of existing strategies
In order to understand whether technology exists today which will
allow us to manage the resources in cellular networks, we briefly
look at the protocols that currently exist which address parts or all
of these requirements.
6.1. End-to-end per-flow resource reservation protocol
An end-to-end per-flow resource reservation signaling protocol is
applied in an end-to-end IP communication path, and it can be used by
an application to make known and reserve its QoS requirements to all
the network nodes included in this IP communication path. This type
of protocol is typically initiated by an application at the beginning
of a communication session. A communication session is typically
identified by the combination of the IP destination address,
transport layer protocol type and the destination port number. The
resources reserved by such a protocol for a certain communication
session will be used for all packets belonging to that particular
session. Therefore, all resource reservation signaling packets will
include details of the session to which they belong.
The end-to-end per-flow resource reservation signaling protocol most
widely used today is the Resource Reservation Protocol (RSVP) (see
[RFC2210], [RFC2205]). The main RSVP messages are the PATH and RESV
messages. The PATH message is sent by a source that initiates the
communication session. It installs states on the nodes along a data
path. Furthermore, it describes the capabilities of the source. The
RESV message is issued by the receiver of the communication session,
and it follows exactly the path that the RSVP PATH message traveled
back to the communication session source. On its way back to the
source, the RESV message may install QoS states at each hop. These
states are associated with the specific QoS resource requirements of
the destination. The RSVP reservation states are temporary states
(soft states) that have to be updated regularly. This means that PATH
and RESV messages will have to be retransmitted periodically. If
these states are not refreshed then they will be removed. The RSVP
protocol uses additional messages either to provide information about
the QoS state or explicitly to delete the QoS states along the
communication session path. RSVP uses in total seven types of
messages:
D. Partain, ed. Expires December 2002 [Page 21]
Internet Draft Cellular RAN Resource Issues June 2002
* PATH and RESV messages
* RESV Confirm message
* PATH Error and RESV Error messages
* PATH Tear and RESV Tear messages
An overview of the functionality of the RSVP functionality includes:
* End-to-end reservation with aggregation of path
characteristics such as fixed delay.
* The same type of reservation functionality in all
routers. Only policy handling separates the edge of the
domain from other routers.
* Multicast and unicast reservations with receiver initiated
reservations. RSVP makes reservations for both unicast and
many-to-many multicast applications, adapting dynamically
to changing routes as well as to group membership.
* Shared reservations for multiple flows.
* Support for policy handling to handle multi-operator
situations since more than one operator will be
responsible for RSVP's operation.
* Flexible object definitions. RSVP can transport and maintain
traffic and policy control parameters that are opaque to
RSVP. Each RSVP message may contain up to fourteen classes of
attribute objects. Furthermore, each class of RSVP objects
may contain multiple types to specify further the format
of the encapsulated data. Moreover, the signaling load
generated by RSVP on the routers is directly proportional
to the flows processed simultaneously by these routers.
Furthermore, processing of the individual flows in the
networking infrastructure may impose a significant processing
burden on the machines, thus hurting throughput. These
issues make it reasonable to question the scalability and
performance in a large cellular radio access network.
* support for uni-directional reservations, not bi-directional.
D. Partain, ed. Expires December 2002 [Page 22]
Internet Draft Cellular RAN Resource Issues June 2002
In the situation where a mobile moves or the connection moves from
one base station to another, it could force the communication path to
change its (source/dest) IP address. The change of IP address will
require that RSVP establish a new RSVP session through the new path
that interconnects the two end points involved in the RSVP session
and release the RSVP session on the old path. During this time, the
end-to-end data path connection is incomplete (i.e., QoS disruption)
and it will negatively affect the user performance.
This approach includes much more functionality and complexity than is
required in the cellular RAN. Our problem is significantly simpler to
solve. The trade-off between performance and functionality is one of
the key issues in the RAN. In our case, the majority of the
functionality in RSVP is not required. This is true for four
reasons:
* Unicast reservations are much less complex than multicast.
* Edge-to-edge with one operator does not require policy
handling in the interior routers.
* Path characteristics and flexible traffic parameters and
QoS definitions could be solved by network dimensioning
and edge functionality.
* Per microflow states in intermediate routers cause severe
scalability problems. Furthermore, receiver-initiated
reservations impose high complexity in the states due to
reverse-direction routing of the RESV messages. A scheme
based on sender oriented reservation (see e.g., [AhBe99])
decreases the complexity of the per microflow states due
to the fact that no reverse-direction routing is
required.
6.2. Integrated Services over Differentiated Services
The IntServ over DiffServ framework addresses the problem of
providing end-to-end QoS using the IntServ model over heterogeneous
networks. In this scenario, DiffServ is one of these networks
providing edge-to-edge QoS. This is similar to the underlying
architecture for this draft, where the specific network is the
cellular RAN, and where the end-to-end model is unspecified. As such,
D. Partain, ed. Expires December 2002 [Page 23]
Internet Draft Cellular RAN Resource Issues June 2002
the problem addressed by IntServ over DiffServ is similar in nature
to the problem described here, although the specific requirements
(such as network utilization and performance) are different.
The IntServ over DiffServ framework discusses two different possible
deployment strategies. The first is based on statically allocated
resources in the DiffServ domain. In this strategy the Diffserv
domain is statically provisioned (see Section 6.3). Furthermore, in
this strategy the devices in the Diffserv network region are not RSVP
aware. However, it is considered that each edge node in the customer
network is consisting of two parts. One part of a node is a standard
Intserv that interfaces to the customer's network region and the
other part of the same node interfaces to the Diffserv network
region. Any edge node in the customer network maintains a table that
indicates the capacity provisioned per SLS (Service Level
Specification) at each Diffserv service level. This table is used to
perform admission control decisions on Intserv flows that cross the
Diffserv region. A disadvantage of this approach is that the edge
nodes in the customer network will not be aware of the traffic load
in the nodes located within the Diffserv domain. Therefore, a
congestion situation on a communication path within the Diffserv
domain cannot be predicted by any of these edge nodes. Due to the
"Efficient network utilization" requirement explained in Section 5.3,
the RAN is dimensioned such that it may have performance bottlenecks
which are not visible to the edges. More advantages and disadvantages
of this approach are discussed in Section 6.3.
The second possible strategy is based on dynamically allocated
resources in the DiffServ domain. According to [RFC2998], this can be
done using RSVP-aware DiffServ routers. However, this approach has
most of the drawbacks described in Section 6.1, and per-microflow
state information is kept in the intermediate routers.
Alternatively, resources in the DiffServ domain can be dynamically
allocated using Aggregated RSVP. This will be discussed in Section
6.4.
Other approaches related to the bandwidth broker concept are still
very immature and will not be discussed here.
D. Partain, ed. Expires December 2002 [Page 24]
Internet Draft Cellular RAN Resource Issues June 2002
6.3. Statically-assigned trunk reservations based on Differentiated
Services
A significant problem in deploying an end-to-end per-flow resource
reservation signaling scheme is its scalability. This can be solved
by aggregating (trunking) several individual reservations into a
common reservation trunk. The reservation trunks can be either
statically or dynamically configured. When the reservation trunks
are statically configured, no signaling protocol is required for
performing the reservation of network resources but is likely to be a
difficult management problem. However, due to the different mobility
requirements (such as handover) and QoS requirements (such as
bandwidth) that the multi-bitrate applications impose on the RAN, it
will be difficult to configure the trunked reservations statically
and utilize the RAN efficiently.
6.4. Dynamic trunk reservations with Aggregated RSVP
The reservation trunks can be dynamically configured by using a
signaling protocol that manages various mechanisms for dynamic
creation of an aggregate reservation, classification of the traffic
to which the aggregate reservation applies, determination of the
bandwidth needed to achieve the requirement and recovery of the
bandwidth when the sub-reservations are no longer required.
The first router that handles the aggregated reservations could be
called an Aggregator, while the last router in the transit domain
that handles the reservations could be called a Deaggregator.
The Aggregator and Deaggregator functionality is located in the edge
nodes. In particular, an Aggregator is located in an ingress edge
node, while a Deaggregator is located in an egress edge node,
relative to the traffic source.
The aggregation region consists of a set of aggregation capable
network nodes. The Aggregator can use a policy that can be based on
local configuration and local QoS management architectures to
identify and mark the packets passing into the aggregated region.
For example, the Aggregator may be the base station that aggregates a
set of incoming calls and creates an aggregate reservation across the
edge-to-edge domain up to the Deaggregator. In this situation the
call signaling is used to establish the end-to-end resource
D. Partain, ed. Expires December 2002 [Page 25]
Internet Draft Cellular RAN Resource Issues June 2002
reservations. Based on policy, the Aggregator and Deaggregator will
decide when the Aggregated states will be refreshed or updated.
One example of a protocol that can be used to accomplish QoS dynamic
provisioning via trunk reservations is the RSVP Aggregation signaling
protocol specified in [BaIt00].
With regards to aggregated RSVP, even if the reservation is based on
aggregated traffic, the number of re-negotiations of the allocated
resources due to mobility (handover) does not decrease and each re-
negotiation of resources has the same performance requirements as the
per-flow reservation procedure.
Note that the aggregated RSVP solution may use a policy to maintain
the amount of bandwidth required on a given aggregate reservation by
taking account of the sum of the underlying end to end reservations,
while endeavoring to change it infrequently. However, such solutions
(policies) are very useful assuming that the cost of the
overprovisioned bandwidth is not significant, since this implies the
need for a certain "slop factor" in bandwidth needs. In a RAN, where
overprovisioning is not preferable due to high costs of transmission
links, a more dynamic QoS provisioning solution is needed.
Furthermore, the aggregated RSVP scheme is receiver initiated and
cannot support bi-directional reservations.
In the aggregated RSVP scheme the resource reservation states stored
in all the RSVP aware Edge and Interior nodes represent aggregated
RSVP sessions (i.e., trunks of RSVP sessions). Therefore, the number
of the resource reservation states in the aggregated RSVP scheme
compared to the (per-flow) RSVP scheme, is decreased. However, in a
Diffserv based RAN the number of the aggregated RSVP sessions depends
on:
* the number of Aggregators/Deaggregators; Considering
that each base station and each BSC/RNC is used as
Aggregator/Deaggregator, the total number of
Aggregators/Deaggregators within the RAN is
significantly high. This is due to the fact
that the number of BSCs/RNCs is significantly
high and the number of base stations in a RAN is
in the range of thousands, see Section 3.1.
* the network topology used; The communication between
RNCs is performed in a meshed way, i.e., all to all
communication. This will imply that many communication
D. Partain, ed. Expires December 2002 [Page 26]
Internet Draft Cellular RAN Resource Issues June 2002
paths will have to be maintained by the RAN
simultaneously.
* the number of Diffserv Code Points (DSCPs) used; More
than one traffic classes will be supported
within the RAN. Therefore, the number of the Diffserv
Code Points (DSCPs) used within the RAN will probably
be higher than one.
Therefore, the number of the aggregated RSVP reservation states
within a Diffserv based RAN will be significantly large.
7. Conclusion
Cellular radio access networks and coming wireless applications
impose different requirements on reservation strategies than typical
Internet conditions.
Firstly, the reservation solution does not need to have the same
level of complexity:
* Edge-to-edge not end-to-end: The IP traffic is generated
in the network and is only transported as far as the
cellular-specific nodes (such as the base station and
BSC/RNC).
* Single operator domain and no inter-domain transport: The
transport is owned and managed by a single operator.
* Only unicast not multicast: The end-to-end payload is
transported between nodes. This transport only requires
a unicast reservation.
Furthermore, a cellular radio access network has much higher
performance requirements on the reservation strategy:
* Efficient usage of the transmission network: The transport
between the BSC/RNC and the base station represents a significant
cost for the cellular operator, and efficient usage of
the transmission network is therefore critical from a cost
point of view. The network should allow dynamic allocation
of resources to allow efficient statistical aggregation
D. Partain, ed. Expires December 2002 [Page 27]
Internet Draft Cellular RAN Resource Issues June 2002
of traffic without causing congestion.
* A wide-area network with significant volume of real-time
traffic: Real-time traffic levels up 100% must be
supported.
* The resource reservation process has to handle a
significantly higher volume of requests, and the process
has to be fast enough to avoid packet losses in the
air-interface during handover.
* The scheme must be optimal for interior nodes and not
for the edge nodes. In this way the necessary set of
parameters required for setting up reservations should be
based upon their effect on interior nodes and not on edge
nodes. This reduces the complexity on the interior routers.
Given these requirements, we believe that appropriate standardization
should take place to create the necessary protocols for edge-to-edge
resource management for a single operator domain.
8. References
[AhBe99] Ahlard, D., Bergkvist, J., Cselenyi, I., "Boomerang
Protocol Specification", Internet draft, Work in progress.
[BaIt00] Baker, F., Iturralde, C. Le Faucher, F., Davie, B.,
"Aggregation of RSVP for Ipv4 and Ipv6 Reservations",
IETF RFC 3175, 2001.
[KeMc00] Kempf, J., McCann, P., Roberts, P., " IP Mobility
and the CDMA Radio Access Network: Applicability
Statement for Soft Handoff", Internet draft, Work
in progress.
[RFC2205] Braden, R., Zhang, L., Berson, S., Herzog, A., Jamin, S.,
"Resource ReSerVation Protocol (RSVP) -- Version 1
Functional Specification", IETF RFC 2205, 1997.
[RFC2210] Wroclawski, J., "The use of RSVP with IETF Integrated
Services", IETF RFC 2210, 1997.
[WeLi99] Westberg, L., Lindqvist, M., "Realtime Traffic over
D. Partain, ed. Expires December 2002 [Page 28]
Internet Draft Cellular RAN Resource Issues June 2002
Cellular Access Networks", Internet draft, Work in
progress (expired).
[RFC2998] Bernet, Y., Yavatkar, R., Ford, P., baker, F., Zhang, L.,
Speer, M., Braden, R., Davie, B., "Felstaine, E.,
"Framework for Integrated Services operation over
Diffserv Networks", IETF RFC 2998, 2000.
9. Acknowledgements
Special thanks to Brian Carpenter, Steven Blake, Phil Chimento,
Anders Bergsten, Marc Greis, Hamad el Allali,
Nicola Blefari-Melazzi, Giuseppe Bianchi and Geert Heijenk for
providing useful comments and input.
10. Authors' Addresses
David Partain
Ericsson Radio Systems AB
P.O. Box 1248
SE-581 12 Linkoping
Sweden
EMail: David.Partain@ericsson.com
Georgios Karagiannis
Ericsson EuroLab Netherlands B.V.
Institutenweg 25
P.O.Box 645
7500 AP Enschede
The Netherlands
EMail: Georgios.Karagiannis@eln.ericsson.se
Pontus Wallentin
Ericsson Radio Systems AB
P.O. Box 1248
SE-581 12 Linkoping
Sweden
EMail: Pontus.Wallentin@era.ericsson.se
Lars Westberg
D. Partain, ed. Expires December 2002 [Page 29]
Internet Draft Cellular RAN Resource Issues June 2002
Ericsson Research
Torshamnsgatan 23
SE-164 80 Stockholm
Sweden
EMail: Lars.Westberg@era-t.ericsson.se
D. Partain, ed. Expires December 2002 [Page 30]