**3.1.1 IntServ**

12 Will-be-set-by-IN-TECH

With the evolution of the 3GPP standards, operators want to provide end-to-end QoS enabled services in UMTS. The end-to-end behavior provided by a series of network elements is an assured level of bandwidth that produces a delay-bounded service with no queueing loss for all conforming packet data (RFC2212, 1997). Assuming the network is functioning correctly,

• A very high percentage of transmitted packets will be successfully delivered by the network to the receiving end-nodes. (The percentage of packets not successfully delivered must closely approximate the basic packet error rate of the transmission medium). • The transit delay experienced by a very high percentage of the delivered packets will not greatly exceed the minimum transmit delay experienced by any successfully delivered packet. (This minimum transit delay includes speed-of-light delay plus the fixed

The end-to-end QoS architecture is provided in Figure 1 in section 2. IP level mechanisms are necessary in providing end-to-end QoS services by interacting TE/MT local bearer service, GPRS bearer service and external bearer service. In this section, how to implement end-to-end

Quality of service refers to the nature of the packet delivery service provided, as described by parameters such as achieved bandwidth, packet delay, and packet loss rates (RFC2216, 1999). The Internet, as originally conceived, offers only a very simple quality of service (QoS), point-to-point best-effort data delivery. It means the network just offered available bandwidth and delay characteristics dependent on instantaneous network load. Before real-time applications such as remote video, multimedia conferencing, visualization, and virtual reality can be broadly used, the Internet infrastructure must be modified to support real-time QoS, which provides some control over end-to-end packet delays. From the view of applications, QoS is realized by adequate provisioning of the network infrastructure. In contrast, a network with dynamically controllable quality of service allows individual application sessions to request network packet delivery characteristics according to their perceived needs, and may provide different qualities of service to different applications. There

• Resource reservation (integrated services): network resources are apportioned according

• Prioritization (differentiated services): network traffic is classified and apportioned

The both types of QoS can be applied to individual application 'flow' or to flow aggregates,

• Per flow: A 'flow' is defined as an individual, uni-directional data stream between two clients (caller and callee), uniquely identified by a 5-tuple (transport protocol, source

• Per aggregate: An aggregate is simply two or more flows. Usually the flows have something in common (e.g. any one or more of 5-tuple parameters, a label or a priority

address, source port number, destination address, and destination port number).

to an application's QoS requirement, subject to bandwidth management policy.

are two basic types of QoS available (qodwhitepaper, 1999):

so there are two other methods to characterize types of QoS:

number, or perhaps some authentication information).

network resources according to bandwidth management policy.

processing time in routers and other communications devices along the path.)

**3. End-to-end IP QoS over UMTS**

these applications may assume that (**?**):

IP QoS is described.

**3.1 QoS mechanisms in IP**

The Internet integrated services (IntServ) framework provides the ability for applications to choose among multiple, controlled levels of delivery service for their data packets. It can provide hard QoS guarantee to individual traffic flows. To support this capability, two things are required (**?**):


In the integrated services framework the first function is provided by QoS control services such as Controlled-Load (RFC2211, 1997) and Guaranteed (RFC2212, 1997). The second function may be provided in a number of ways, but is frequently implemented by a resource reservation setup protocol such as RSVP (RFC2205, 1997).

The controlled load service is intended to support a broad class of applications which have been developed for use in today's Internet, but are highly sensitive to overloaded conditions. Important members of this class are the "adaptive real-time applications" currently offered by a number of vendors and researchers. These applications have been shown to work well on unloaded nets, but to degrade quickly under overloaded conditions. It is equivalent to "best effort service under unloaded conditions". The controlled-load service is intentionally minimal, in that there are no optional functions or capabilities in the specification. The service offers only a single function. It is better than best effort, but cannot provide strictly bounded service as guaranteed service.

The controlled-load service can be implemented by using evolving scheduling and admission control algorithms. The implementations are highly efficient in the use of network resources.

Guaranteed service guarantees that datagrams will arrive within the guaranteed delivery time and will not be discarded due to queue overflows, provided the flow's traffic stays within its specified traffic parameters. It is similar to emulate a dedicated virtual circuit. This service is intended for applications which need a firm guarantee that a datagram will arrive no later than a certain time after it was transmitted by its source. For example, some audio and video "play-back" applications are intolerant of any datagram arriving after their play-back time. Applications that have hard real-time requirements will also require guaranteed service.

Fig. 5. RSVP setup flow

each reservation.

a RSVP session.

**3.1.3 DiffServ**

the previous source address of the PATH message.

confirmation message back to the receiver.

1. When a sender wants to set up a traffic link, it will generate the traffic specification (TSpec), which describes data traffic, such as upper/lower bounds of bandwidth, delay, and jitter. Then RSVP sends out a PATH message containing TSpec to the receiver(s) (unicast or multicast). Along the route, each RSVP-enabled router trigger a "path-state" that includes

End to End Quality of Service in UMTS Systems 113

2. After receiver receives the PATH message, the receiver sends a RESV message "upstream" to make a resource reservation. The RESV message includes a request specification (RSpec) which indicates what type of IntServ required âA ¸ ˘ S either Controlled Load or Guaranteed, a filter specification (filter spec) (indicating e.g. the transport protocol and port number). The RSpec and filter spec represent a flow-descriptor that RSVP routers use to identify

3. Along the RSVP upstream, RSVP routers use the admission control to authenticate the resource reservation request and allocate the necessary resources when the routers receive the RESV message. If the request cannot be met (due to no enough bandwidth or authorization failure), the RSVP router returns an error back to the receiver. If the request

4. When the last router receives the RESV message and accepts the request, it sends out a

5. There is an explicit tear-down process for a reservation when sender or receiver terminate

The Integrated Services/RSVP model relies upon traditional datagram forwarding in the default case, but allows sources and receivers to exchange signaling messages which establish additional packet classification and forwarding state on each node along the path between

is accepted, the router forwards the RSVP message to the next router.

Guaranteed service does not attempt to minimize the jitter (the difference between the minimal and maximal datagram delays); it merely controls the maximal queueing delay. Because the guaranteed delay bound is a firm one, the delay has to be set large enough to cover extremely rare cases of long queueing delays. Several studies have shown that the actual delay for the vast majority of datagrams can be far lower than the guaranteed delay. Therefore, authors of playback applications should note that datagrams will often arrive far earlier than the delivery deadline and will have to be buffered at the receiving system until it is time for the application to process them.

Guaraneteed service represents one extreme end of delay control for networks. Most other services providing delay control provide much weaker assurances about the resulting delays. In order to provide this high level of assurance, guaranteed service is typically only useful if provided by every network element along the path (i.e. by both routers and the links that interconnect the routers). Moreover, as described in the Exported Information section, effective provision and use of the service requires that the set-up protocol or other mechanism used to request service provides service characterizations to intermediate routers and to the endpoints.

Integrated Services routers uses admission control and resource allocation method to offer QoS guarantee. A token-bucket model is used to characterize the input/output queueing algorithm. It can smooth the flow of outgoing traffic. The IntServ parameters include (qodwhitepaper, 1999):


#### **3.1.2 RSVP**

For offering IntServ, a way to communicate the application's requirements to network elements along the path and to convey QoS management information between network elements and the application must be provided. A resource reservation setup protocol called RSVP (rfc2205, 1997) is implemented for this purpose. It is a signaling protocol that can provide reservation setup and control to enable the integrated services by using a variety of QoS control, a variety of setup mechanisms.

A simplified RSVP working flow is shown in Figure 5

14 Will-be-set-by-IN-TECH

Guaranteed service does not attempt to minimize the jitter (the difference between the minimal and maximal datagram delays); it merely controls the maximal queueing delay. Because the guaranteed delay bound is a firm one, the delay has to be set large enough to cover extremely rare cases of long queueing delays. Several studies have shown that the actual delay for the vast majority of datagrams can be far lower than the guaranteed delay. Therefore, authors of playback applications should note that datagrams will often arrive far earlier than the delivery deadline and will have to be buffered at the receiving system until it

Guaraneteed service represents one extreme end of delay control for networks. Most other services providing delay control provide much weaker assurances about the resulting delays. In order to provide this high level of assurance, guaranteed service is typically only useful if provided by every network element along the path (i.e. by both routers and the links that interconnect the routers). Moreover, as described in the Exported Information section, effective provision and use of the service requires that the set-up protocol or other mechanism used to request service provides service characterizations to intermediate routers and to the

Integrated Services routers uses admission control and resource allocation method to offer QoS guarantee. A token-bucket model is used to characterize the input/output queueing algorithm. It can smooth the flow of outgoing traffic. The IntServ parameters include

Token rate ( r): The continually sustainable bandwidth (bytes/second) requirement for a flow. It represents the average data rate into the bucket, and the target shaped data rate

Token-bucket rate ( b ) : the extent to which the data rate can exceed the sustainable average for short periods of time, or the amount of data sent cannot exceed rT+b (where T is any

Peak rate ( p ): It is the maximum send rate (bytes/second) if known and controlled. At any

Minimum policed size ( m): The size (byte) of the smallest packet (data payload only) can be generated by the sending application. The size m is not an absolute number. If the percentage of small packets is small, the number m should increased to reduce the

Maximum packet size ( M ): The biggest size of a packet (bytes). The M is absolute number. Any packets (size > M) are considered out of spec and may not receive QoS controlled

For offering IntServ, a way to communicate the application's requirements to network elements along the path and to convey QoS management information between network elements and the application must be provided. A resource reservation setup protocol called RSVP (rfc2205, 1997) is implemented for this purpose. It is a signaling protocol that can provide reservation setup and control to enable the integrated services by using a variety

time period (T), the amount sent data cannot exceed M+pT.

of QoS control, a variety of setup mechanisms.

A simplified RSVP working flow is shown in Figure 5

overhead estimate. All packets smaller than m are treated as size m.

is time for the application to process them.

endpoints.

(qodwhitepaper, 1999):

out of the bucket.

time period).

service.

**3.1.2 RSVP**

Fig. 5. RSVP setup flow


#### **3.1.3 DiffServ**

The Integrated Services/RSVP model relies upon traditional datagram forwarding in the default case, but allows sources and receivers to exchange signaling messages which establish additional packet classification and forwarding state on each node along the path between

**3.1.4 MPLS**

best effort IP networks, including

MPLS is a key development in IETF that will add a number of essential capabilities to today's

End to End Quality of Service in UMTS Systems 115

• Traffic Engineer, enhancing overall network utilization by creating a uniform or

• Supporting network scalability, providing IP based Virtual Private Networks (VPN)

MPLS borrows the idea from ATM switching. It remains independent of the Layer-2 and Layer-3 protocols. Besides IP, other network protocols (such as IPX, ATM, PPP or Frame-Relay) also can work with MPLS. MPLS resides on routers. When a packet flow enters a edge router of the MPLS domain, all packets are marked to clarify priority with a fixed-length label (20 bits label). The label identifies the packets routing information in this MPLS network,

A MPLS domain includes label edge routers (LERs) and label switching routers (LSRs). The route taken by an MPLS-labeled packet is called the label switched path (LSP). LST is a high-speed router in the core of a MPLS network, which participates in the establishment of LSPs. LER is a router that operates at the edge of a MPLS network. It is used to assign and

MPSL is similar to DiffServ because it also marks traffic at ingress of a MPLS network, and un-marks at egress gate. However, MPLS marking is used to decide the next hop router while

This section describes how to provide Quality of Service in UMTS for the end-to-end services through the TE/MT local bear service, GPRS bearer service and external bearer service shown in the Fig. 1. To provide end-to-end IP QoS, it is necessary to manage the QoS within each domian. An IP BS Manager is used to control the external IP bearer service. Due to the different techiniques used within the IP network,this communicates to the UMTS BS manager

At PDP context setup the user shall have access to one of the following alternatives, basic GPRS IP connectivity service or enhanced GPRS based services. To enable coordination between events in the application layer and resource management in the IP bearer layer, a logical element, the Policy and Charging Rules Function (PCRF), is used as a logical policy decision element which will be detailed in section 4. It is also possible to implement a policy decision element internal to the IP BS Manager in the GGSN. While interworking with the

QoS management functions is shown in Fig. 7 which describes how to control the external IP bearer services and how they relate to the UMTS bearer service QoS management entity.

IP BS Manager uses standard IP mechanisms to manage the IP bearer services. These mechanisms may be different from mechanisms used within the UMTS, and may have different parameters controlling the service. When implemented, the IP BS Manager

differentiated distribution of traffic throughout the network.

• Providing traffic with different Classes of Service (CoS) • Providing traffic with different Quality of Service (QoS)

remove labels when packets enter or exit the MPLS network.

DiffServ marking is used to determine priority in route itself.

**3.2 QoS management functions for end-to-end IP QoS**

external network, the RSVP, DiffServ, MPLS will be used.

through the Translation function.

also define the quality of service for the packets.

them (rfc1633, 1994). In the absence of state aggregation, the amount of state on each node scales in proportion to the number of concurrent reservations, which can be potentially large on high-speed links. This model also requires application support for the RSVP signaling protocol. Differentiated service is a simple method by classifying services of different applications (rfc2475, 1998). Currently there are two standard per hop behaviour (PHBs) define two traffic classes:


Fig. 6. DffServ architecture

DiffServ offers a simple QoS management method without signaling mechanism. The DiffServ architecture is shown in Figure 6. It includes classifier and traffic conditioner. A traffic conditioner contains the following elements: meter, maker, shaper/dropper. The differentiated services architecture is based on a simple model where traffic entering a network is classified and possibly conditioned at the boundaries of the network, and assigned to different behavior aggregates. Each behavior aggregate is identified by a single DiffServ codepoint (DSCP). Within the core of the network, packets are forwarded according to the per-hop behavior associated with the DiffServ codepoint.

When a traffic flow enters a DiffServ network, the flow is selected by a classifier, which steers the packets to a logical instance of a traffic conditioner. A meter is used to measure the traffic flow agains a traffic profile. A meter is used (where appropriate) to measure the traffic stream against a traffic profile. The state of the meter with respect to a particular packet (e.g., whether it is in- or out-of-profile) may be used to affect a marking, dropping, or shaping action. When packets exit the traffic conditioner of a DS boundary node the DiffServ codepoint of each packet must be set to an appropriate value.
