WO2010069373A1

WO2010069373A1 - Method of managing quality of service in a communication network

Info

Publication number: WO2010069373A1
Application number: PCT/EP2008/067777
Authority: WO
Inventors: Vesa Pauli Hellgren; Serafim Petsis
Original assignee: Nokia Siemens Networks Oy
Priority date: 2008-12-17
Filing date: 2008-12-17
Publication date: 2010-06-24

Abstract

A method of managing quality of service in a communication network is provided, wherein the method comprises receiving an accounting request message at a network access server, determining whether a secondary packet data protocol context is active, and in case a secondary packet data protocol context is active determining an aggregated quality of service value based on quality of service requirements of active primary and secondary packet data protocol contexts.

Description

DESCRIPTION

Method of Managing Quality of Service in a Communication Net- work

Field of Invention

The present invention relates to a method of managing quality of service in a communication network, in particular in a mobile network. Moreover it relates to a network element for a communication network, in particular to a network access server adapted to manage quality of service in the communica- tion network. Furthermore, the invention relates to a program element, and a computer-readable medium.

Art Background

In a public land mobile network (PLMN) , a service aware (SA) node can be deployed behind a gateway GPRS support node (GGSN) , corresponding to a so-called Gi-deployment, to enable operators to offer services in a more flexible and efficient way. The communication between GGSN and SA node is performed through the network access server (NAS) interface. NAS interface uses radius authentication dial in user service (RADIUS) authentication for allocating the IP address (if needed) and RADIUS accounting for providing the IP address, international mobile subscriber identity (IMSI) and mobile subscriber ISDN (MSISDN) to the SA node. The NAS can be for example a 3rd party GGSN that acts as a RADIUS client and a SA node that acts as a RADIUS server. The user plane traffic is carried in plain IP, generic routing encapsulation (GRE) or IP-in-IP tunnels and the RADIUS signalling is used to set up the user plane traffic in a similar way as in the GPRS/3G networks. The NAS interface can be referred to as being the southbound Gi interface, as an alternative to the Gn interface.

During the packet data protocol (PDP) context activation procedure, an "accounting request start" message, including an access point name (APN) , is sent from GGSN towards the SA node to inform it about the subscriber identification (IMSI, MSISDN) so that user profile can be determined. Upon receiving of the message a NAS context is established. Thus, the start message is used to inform SA node about new sessions, so that SA node can monitor them properly. An "accounting request stop" message can be used by GGSN to inform the SA node about a possible PDP context termination, thus SA has to delete the already established NAS context.

RADIUS accounting request start and stop messages may be used during both primary and secondary PDP context activation and deactivation procedures, respectively. Secondary PDP contexts are established based on an existing PDP context so to share the same user equipment (UE) IP address but with different quality of service (QoS) requirements. If the SA node is used for IP address assignment, then, upon reception of a RADIUS "accounting request stop" message for all PDP contexts associated to a session defined by APN and IMSI or MSISDN, the SA node (server) may make the associated IP address available for assignment.

In order to avoid race conditions, the GGSN shall include a 3GPP vendor-specific sub-attribute session stop indicator when it sends the "accounting-request stop" for the last PDP context of a PDP session and the PDP session is terminated (i.e. the IP address and all GTP tunnels can be released) . The SA server shall not assume the PDP session terminated until an "accounting-request stop" with the session stop indicator is received. In 3GPP reference architecture, the SA function corresponds to flow based charging (FBC) , which is implemented in Release 6 TPF (traffic plane function) and in Release 7 PCEF (policy and charging enforcement function) . 3GPP does not define interface between GGSN and TPF/PCEF. SA function can be integrated as part of GGSN or it can be implemented separate node connected to GGSN via Gi interface defined in 3GPP specification 29.061, which corresponds to the NAS interface described above.

Even though 3GPP reference architecture does not specify details about the case where GGSN and TPF/PCEF are implemented as separate nodes, this alternative is often used in practice. For operators, this approach has the benefit that SA can be introduced without making changes to existing GGSN nodes .

In general, traffic of each PDP context (primary or secondary) is assigned to a separate GTP-U tunnel and GGSN is able to map UL packets to the right PDP context based on the GTP-U tunnel it belongs to. For downlink (DL) packets, traffic flow template (TFT) is used by GGSN to distinguish between differ- ent user payloads and transmit packets with different QoS requirements via different PDP contexts but to the same PDP address. Thus, TFT is really required and so is sent from the mobile station (MS) via the SGSN (transparently) to the GGSN in order to enable packet classification and policing for DL data transfer.

However, there may be a need for a method of quality of services managing having an improved performance and better utilization of resources when using SA nodes, in particular, this method should support the use of secondary PDP context.

Summary of the Invention

This need may be met by a method of managing quality of service in a communication network, a network element for a communication network, a program element, and a computer- readable medium according to the independent claims. Advantageous embodiments of the present invention are described by the dependent claims .

According to an exemplary aspect a method of managing quality of service in a communication network is provided, wherein the method comprises receiving an accounting request message at a network element, determining whether a secondary packet data protocol context is active, and in case a secondary packet data protocol context is active determining an aggre- gated quality of service value based on quality of service requirements of active primary and secondary packet data protocol contexts.

In particular, the determination may comprise whether a sec- ondary packet data protocol (PDP) context exists in a network access server (NAS) context. In particular, the accounting request may be a RADIUS accounting request. Furthermore, the accounting request may be send from a GGSN. The GGSN may operate according to the 3GPP protocol .

According to an exemplary aspect a network element, in particular an access server, is provided which comprises a receiving unit adapted to receive an accounting request message, a determining unit adapted to determine an active sec- ondary packet data protocol context, and a processing unit adapted to determine an aggregated quality of service value based on quality of service requirements of active primary and secondary packet data protocol contexts in case a secon- dary packet data protocol context is active.

According to an exemplary aspect a communication network is provided which comprises a network element according to an exemplary aspect of the invention, a gateway GPRS support node, and a user equipment. Furthermore, the communication network may comprise additional network elements, e.g. a SGSN node or the like.

According to an exemplary aspect of the invention a program element is provided, which, when being executed by a processor, is adapted to control or carry out a method according to an exemplary aspect of the invention.

According to an exemplary aspect of the invention a computer- readable medium, in which a computer program is stored which, when being executed by a processor, is adapted to control or carry out a method according to an exemplary aspect of the invention .

The network element may preferably be a network access server. However, it may also be a base station, a GGSN node or the like. By providing a method according to an exemplary aspect of the invention it may be possible to ensure that received packets may be mapped to right secondary PDP contexts as all contexts of the same PDP session may have source IP address even in cases a SA node is used and receiving uplink packets via IP or GRE/IP-IP tunnels while still using the best suitable quality of service. Thus, it may be possible that restrictions imposed on the way how to handle secondary contexts which restrictions are caused by a deployment of a SA node may be managed in an improved way when using a method according to an exemplary aspect of the invention in which an aggregated quality of service value or profile may be deter- mined. In particular, it may be possible to determine the most suitable QoS profile for given requirements of primary and secondary PDP context.

A "network access server (NAS) context" may be established for each network access server session when an accounting request start or accounting start request is sent from a gateway, e.g. a GGSN toward a service aware (SA) node to inform the SA node about subscriber identification (IMSI, MSISDN) so that a user profile may be determined. However, a new NAS context may not always be created when accounting request start is received. A new NAS context may only be created if accounting request start is about a new PDP session, i.e. received tuple (IMSI, MSISDN, IP address, APN) is unique. If accounting request start is about existing PDP session and thus related to some existing NAS context in SA node, then the accounting session Id may be recorded to this NAS context (if accounting session Id is received in the message) . In general there may be a one-to-one relation between PDP session and NAS context, i.e. NAS context can be single primary PDP context or NAS context is a set of secondary PDP contexts belonging to the same PDP session. When PDP session contains more than one PDP context, all those PDP contexts may be called secondary PDP contexts and if PDP session contains one PDP context, this PDP context may be called primary PDP con- text.

That is, a NAS context may be established at the beginning of a new session, service or communication relating to a new PDP session and may be used in order to identify the specific PDP session .

Next, further exemplary embodiments of the method are de- scribed. However, these embodiments also apply to the, the network element, the program element, and the computer- readable medium.

According to another exemplary embodiment the method further comprises defining an aggregated quality of service profile based on the aggregated quality of service value. In particular, the aggregated QoS profile may be used or may be handled as an additional parameter of a NAS context data structure.

According to another exemplary embodiment of the method the aggregated quality of service profile is used as an in-use quality of service profile for services of the network element. In particular, the in-use QoS profile may be used for further services of a network access server (NAS) . For exam- pie, the in-use QoS may define or may be handled as an additional parameter of a NAS context data structure. In particular, the NAS may provide policing and/or shaping based on the QoS profile and/or QoS value.

According to another exemplary embodiment of the method the determination of the aggregated quality of service value is based on one function out of the group of functions consisting of the allocation retention priority, delivery order, delivery of erroneous SDU, maximum bit rate, maximum SDU size, residual bit error rate, SDU error ration, traffic class, transfer delay, traffic handling priority, and guaranteed bit rate . In particular, in the case of the allocation retention priority (ARP) a value may be used as QoS value which indicates highest priority. In the case of using the delivery order this function may be enabled if any of the QoS profiles are enabled, i.e. any of the QoS profiles relating to a primary or secondary context. In the case of using the delivery of erroneous SDU this function may be enabled if any of the QoS profiles are enabled, i.e. any of the QoS profiles relating to a primary or secondary context. In case of the maximum bit rate or the maximum SDU size the highest value, respectively may be used. In case of using the residual bit error rate or the SDU error ratio, the lowest value, respectively may be used. In case of using the traffic class the most demanding value, i.e. the value corresponding to the service having the highest requirements, may be used, e.g. streaming may be more demanding than background class. In case of using the transfer delay the lowest value may be used, and in case of using the traffic handling priority the highest priority may be select, while in case of using the guaranteed bit rate the maximum value may be used. However, in the latter three cases this selection may be only used in case that the traffic class is not set to "background" .

According to another exemplary embodiment of the method the accounting request message comprises a packet data protocol quality of service profile. In particular, this PDP QoS profile may be stored in the NAS, e.g. under the NAS context data structure.

According to another exemplary embodiment of the method the packet data protocol quality of service profile is used as an in-use quality of service profile for services of the network element in case no secondary packet data protocol context is active . According to another exemplary embodiment the method further comprises sending the aggregated quality of service value to a further network element. In particular, the further network element may be another network element distinct to the network element, e.g. a network element external to the NAS constituting an important entity of the communication network or of another communication network. Additionally or alternatively to the aggregated quality of service value a corre- sponding aggregated quality of service profile may be send to the further network element. Furthermore, resources may be reserved internally in the NAS based on the aggregated quality of service value and/or profile. It should be noted that the term "resource" or "communication resource" may particu- larly denote or relate to traffic resources i.e. bandwidth and packet priority.

According to another exemplary embodiment of the method the accounting request message is an accounting request start message.

According to another exemplary embodiment of the method the accounting request message is an accounting request interim message. That is, the accounting request message may be a message initiating an update of a PDP QoS profile after an accounting request start message is sent but before an accounting request stop message is sent.

According to another exemplary embodiment of the method the accounting request message is an accounting request stop message .

According to another exemplary embodiment the method further comprises removing a packet data protocol quality of service profile from a network access server context data structure, wherein the packet data protocol quality of service profile is correlated to the accounting request stop message.

Summarizing an exemplary aspect of the invention may be seen in providing a method of managing quality of service in a communication network wherein the method provides a solution that may always use the most appropriate/possible QoS profile when secondary PDPs are established in an SA node. A princi- pie idea may be shortly characterized in the following:

• A data structure of the NAS context may always store the QoS profiles of all the related PDP contexts.

• An additional parameter "in-use QoS profile" may also be used in the data structure that defines the currently used QoS that may be used both internally in SA node and externally towards other important entities of the communication network.

• When secondary PDPs are established, the "in-use QoS Profile" parameter may take the value of the "aggregate QoS", which will provide the so-called aggregate QoS of all the

PDPs established under the NAS context. There may be several ways to compute the "aggregate QoS".

• Whenever a secondary PDP is deactivated, the corresponding QoS profile may be removed from the NAS context data structure. If there are still multiple PDP contexts active under the PDP session, then "aggregate QoS" may be updated and the "in-use QoS profile" may use the QoS value defined by "aggregate QoS". If only one PDP context remains in the PDP session, the actual QoS profile of that remaining PDP may overwrite the "in-use QoS profile" value.

It has to be noted that embodiments and aspects of the invention have been described with reference to different subject- matters. In particular, some embodiments and aspects have been described with reference to apparatus type claims whereas other embodiments have been described with reference to method type claims. However, a person skilled in the art will gather from the above and the following description that, unless other notified, in addition to any combination of features belonging to one type of subject-matter also any combination between features relating to different subject- matters, in particular between features of the apparatus type claims and features of the method type claims is considered to be disclosed with this application.

The aspects and exemplary embodiments defined above and further aspects of the invention are apparent from the example of embodiment to be described hereinafter and are explained with reference to these examples of embodiment.

Brief Description of the Drawings

Fig. 1 schematically shows a SA node in Gi deployment.

Fig. 2 schematically shows a SA node in Gi deployment and indicates the traffic separation.

Fig. 3 schematically depicts handling in PDP context activation, de-activation and modification.

Detailed Description

The illustration in the drawing is schematic With reference to Figs. 1 and 2 information is given which may be helpful in understanding the present invention.

As already mentioned above a public land mobile network (PLMN) 100 may comprise a service aware (SA) node 101 which can be deployed behind a gateway GPRS support node (GGSN) 102, corresponding to a so-called Gi-deployment, to enable operators to offer services in a more flexible and efficient way. A mobile phone 103 may be used for transferring data or data packets in the PLMN network and may communicate with the GGSN which is indicated by arrow 104 in Fig. 1.

The communication between GGSN and SA node may be performed through the network access server (NAS) interface. NAS interface uses radius authentication dial in user service (RADIUS) authentication for allocating the IP address (if needed) and RADIUS accounting for providing the IP address, international mobile subscriber identity (IMSI) and mobile subscriber ISDN (MSISDN) to the SA node. That is, the GGSN 102 as well as the SA 101 may comprise an interface adapted to perform the necessary communication for a RADIUS service. In Fig. 1 the communication relating to the RADIUS service is indicated by the arrow 105. For example, the NAS interface can be formed by a 3rd party GGSN that acts as a RADIUS client and a SA node that acts as a RADIUS server. The user plane traffic is carried in plain IP, generic routing encapsulation (GRE) or IP-in-IP tunnels which is indicated schematically by the pipe 106 and the RADIUS signalling is used to set up the user plane traffic in a similar way as in the GPRS/3G networks. The NAS interface can be referred to as being the southbound Gi interface, as an alternative to the Gn interface. Fig. 2 schematically shows a SA node in Gi deployment and indicates the traffic separation.

As already mentioned above, In general, traffic of each PDP context (primary or secondary) is assigned to a separate GTP- U tunnel 201 and 202, respectively and GGSN 203 is able to map UL packets to the right PDP context based on the GTP-U tunnel it belongs to. For DL packets, TFT is used by GGSN to distinguish between different user payloads and transmit packets with different QoS requirements via different PDP contexts but to the same PDP address. Thus, TFT may be really required and so is sent from the mobile station (MS) via the service GPRS support node (SGSN, transparently) to the GGSN in order to enable packet classification and policing for DL data transfer.

For the case where SA functionality is required and a SA node 204 is deployed behind a legacy GGSN node, NAS interface may be used and thus traffic reaches the SA node using either IP or GRE/IP-IP tunnels. The aforementioned deployment may generate technical restrictions on the way secondary PDP contexts are handled in the SA node. Actually, when the SA node receives UL packets via all PDP context traffic tunnels 205, it may not be able to map the received packets to the right secondary PDP context as all the contexts of the same PDP session have same source IP address. By definition, we can not either assume that TFT can be applied also to UL traffic. Thus, secondary PDP contexts of a session are all mapped to one NAS context in the SA node. Traffic received from all those contexts is also mapped to the same NAS context and consequently there is no need to distinguish traffic for each secondary context. Due to the fact that only one NAS Context is established for all PDPs that belong to the same session (secondary) , the need for appropriate QoS handling internally in the SA node and externally towards other entities is required. The pro- posed solution will be described later and specifies an appropriate QoS profile that has to be used internally and also conveyed towards external nodes (i.e., AAA, CG, OCS etc) in the case where more than one PDP contexts exists per session. This is schematically indicated by the four tunnels 206, 207, 208, and 209 in Fig. 2. In addition to that, it should remember each secondary's PDP QoS, use it internally for policing and shaping purposes and also pass that info to all required external servers when this is the last context of the session (case where all other have been deactivated) .

In particular, Fig. 3a schematically shows a handling of ap- propriate QoS profile in the case of context activation. After receiving of an accounting request start message 301 in an SA node, the QoS profile of the received PDPs is stored under the NAS context data structure 302. Then a determination is made whether secondary PDPs exist in the NAS context 303. In case secondary PDPs exist an aggregated QoS profile is calculated based on the newly received QoS profile and the already present QoS profiles 304. Possible ways to calculate or determine the aggregated QoS profiles will be described later in more detail. After the calculation of the aggregated QoS profile said aggregated QoS profile will be used as the in-use QoS profile 305, i.e. will be defined as the QoS profile used for further services of said SA node for that specific NAS context. Additionally, the SA node may reserve resources internally and may provide policing/shaping based on said in-use QoS profile or parameter 306. Furthermore, the SA node may inform/update external entities with the newly calculated aggregated QoS profile in case an already established NAS context QoS profile is different than the newly calcu- lated aggregated QoS 307.

In case no secondary PDPs exist the actual received PDPs QoS profile is used for the whole NAS context, i.e. the received QoS forms the in-use QoS profile 308. Furthermore, the SA node may reserve resources internally and may provide policing/shaping based on said in-use QoS profile or parameter 309. Additionally, external entities may be informed based on said parameter or in-use QoS profile 310.

Additionally, a schematic NAS context data structure is depicted in Fig. 3a in the right upper part. The NAS context data structure comprises a plurality of PDP context parameters, e.g. 1 to 11, which may include QoS values. An additional parameter "in-use QoS profile" may be included into the NAS context data structure.

In particular, Fig. 3b schematically shows a handling of appropriate QoS profile in the case of context modification which handling is similar to the case of context activation shown in Fig. 3a.

After receiving of an accounting request interim message 311 in an SA node, the QoS profile of the received PDPs is used to update the NAS context data structure 312. Then a determi- nation is made whether secondary PDPs still exist in the NAS context 313. In case secondary PDPs exist an aggregated QoS profile is calculated based on the newly received QoS profile and the already present QoS profiles 314. After the calculation of the aggregated QoS profile said aggregated QoS pro- file will be used as the in-use QoS profile 315, i.e. will be defined as the QoS profile used for further services of said SA node for that specific NAS context. Additionally, the SA node may reserve resources internally and may provide polic- ing/shaping based on said in-use QoS profile or parameter

316. Furthermore, the SA node may inform/update external entities with the newly calculated aggregated QoS profile in case an already established NAS context QoS profile is different than the newly calculated aggregated QoS 317.

In case no secondary PDPs exist the actual received PDPs QoS profile is used for the whole NAS context, i.e. the received QoS forms the in-use QoS profile 318. Furthermore, the SA node may reserve resources internally and may provide polic- ing/shaping based on said in-use QoS profile or parameter

319. Additionally, external entities may be informed based on said parameter or in-use QoS profile 320.

In particular, Fig. 3c schematically shows a handling of ap- propriate QoS profile in the case of context de-activation which handling is similar to the case of context activation shown in Fig. 3a.

After receiving of an accounting request stop message 321 in an SA node, the QoS profile of the received PDPs is removed from the NAS context data structure 322. Then a determination is made whether secondary PDPs still exist in the NAS context 323. In case secondary PDPs exist an aggregated QoS profile is calculated based on the newly received QoS profile and the already present QoS profiles 324. After the calculation of the aggregated QoS profile said aggregated QoS profile will be used as the in-use QoS profile 325, i.e. will be defined as the QoS profile used for further services of said SA node for that specific NAS context. Additionally, the SA node may reserve resources internally and may provide policing/shaping based on said in-use QoS profile or parameter 326. Furthermore, the SA node may inform/update external entities with the newly calculated aggregated QoS profile in case an al- ready established NAS context QoS profile is different than the newly calculated aggregated QoS 327.

In case no secondary PDPs exist the last/remaining PDPs QoS profile is used for the whole NAS context, i.e. the received QoS forms the in-use QoS profile 328. Furthermore, the SA node may reserve resources internally and may provide policing/shaping based on said in-use QoS profile or parameter 329. Additionally, external entities may be informed based on said parameter or in-use QoS profile 330.

In the following three methods of computing or calculating an aggregated QoS profile will be described.

Method 1 ("on/off") : "Aggregate QoS" will have value "0", which means that no QoS functionality is enabled when this QoS profile is in use. The advantage of this method is that it is very simple to implement. It means that in practice QoS functionality is disabled whenever NAS context consists of more than one PDP context. This may be very similar to the prior art solution used, although there is one crucial difference namely that the QoS is still enabled for all those NAS contexts which consist of only a single PDP context.

Method 2 ("default QoS") :

"Aggregate QoS" will have a fixed value "default QoS", which uses some meaningful values for the QoS profile. The table listed below shows a possible implementation of the "default QoS":

The method is as well simple to implement. Compared to method 1, it means that QoS is always enabled. The method 2 may be used in the SA node whenever a NAS context has been created even if the NAS context consists of only a single PDP context and no QoS profile has been defined in the request message.

Method 3 ("real aggregation") : "Aggregate QoS" will be computed from the real QoS profiles of the individual PDP contexts using some aggregation functions. Following function may be used to compute the aggregate QoS profile:

ARP: use the value, which indicates highest priority, Delivery order: enable, if enabled in any of the QoS profiles, • Delivery of erroneous SDUs: enabled, if enabled in any of the QoS profiles,

• Maximum bit rate: use the maximum value,

• Maximum SDU size: use the maximum value, • Residual Bit Error Rate: use the lowest value,

• SDU error ratio: use the lowest value,

• Traffic class: use the "most demanding" value (e.g. streaming is more demanding than background class) ,

• Transfer delay: use the lowest value (not used if traf- fie class is "background") ,

• Traffic Handling Priority: select the highest priority

(not used if traffic class is "background") , and

• Guaranteed bit rates: use the maximum values (not used if traffic class is "background")

The method 3 provides the QoS profile, which never uses values, which are higher than any of the real QoS profiles. Thus, it is fairer than the other approaches. Another advantage may be the fact that if downlink traffic which passes through SA node is not compliant with the QoS profile of the related PDP context, then the SA node may not charge traffic which is actually dropped in the GGSN node. Preferably, the aggregate QoS may be computed every time there is a change in the real QoS profiles or new real QoS profiles are added/removed.

Finally, it should be noted that the above-mentioned embodiments illustrate rather then limit the invention, and that those skilled in the art will be capable of designing many alternative embodiments without departing from the scope of the invention as defined by the appended claims. In the claims, any reference signs placed in parentheses shall not be construed as limiting the claims. The word "comprising" and "comprises", and the like, does not exclude the presence of elements or steps other than those listed in any claim or the specification as a whole. The singular reference of an element does not exclude the plural reference of such elements and vice-versa. In a device claim enumerating several means, several of these means may be embodied by one and the same item of software or hardware. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

List of reference signs:

100 PLMN

101 SA node 102 GGSN

103 Mobile phone

104 Arrow

105 Generic Routing Encapsulation

106 IP-in-IP tunnel 201 GTP-U tunnel

202 GTP-U tunnel

203 GGSN

204 SA node

205 PDP context traffic 206-209 Tunnels

301 Accounting request

302 Storing of QoS

303 Decision whether secondary PDP exists

304 Calculating aggregated QoS 305 Defining in-use QoS profile

306 Reserving of resources

307 Informing external entities

308 Defining in-use QoS profile

309 Reserving of resources 310 Informing external entities

311 Accounting request

312 Update of QoS

313 Decision whether secondary PDP still exists

314 Calculating aggregated QoS 315 Defining in-use QoS profile

316 Reserving of resources

317 Informing external entities

318 Defining in-use QoS profile

319 Reserving of resources 320 Informing external entities

321 Accounting request

322 Removing of QoS

323 Decision whether secondary PDP exists 324 Calculating aggregated QoS

325 Defining in-use QoS profile

326 Reserving of resources

327 Informing external entities

328 Defining in-use QoS profile 329 Reserving of resources

330 Informing external entities

Claims

CLAIMS :

1. A method of managing quality of service in a communication network, the method comprising: receiving an accounting request message (301, 311, 321) at a network element (204); determining whether a secondary packet data protocol context is active (303, 313, 323); and in case a secondary packet data protocol context is active determining an aggregated quality of service value based on quality of service requirements of active primary and secondary packet data protocol contexts (304, 314, 324) .

2. The method according to claim 1, further comprising: defining an aggregated quality of service profile based on the aggregated quality of service value.

3. The method according to claim 1 or 2, wherein the aggregated quality of service profile is used as an in-use quality of service profile (305, 315, 325) for services of the network element.

4. The method according to any one of the claims 1 to 3, wherein the determination of the aggregated quality of service value (304, 314, 324) is based on one function out of the group of functions consisting of:

ARP, delivery order, delivery of erroneous SDU, maximum bit rate, maximum SDU size, residual Bit error rate,

SDU error ration, traffic class, transfer delay, traffic handling priority, and guaranteed bit rate .

5. The method according to any one of the claims 1 to 4 , wherein the accounting request message comprises a packet data protocol quality of service profile.

6. The method according to claim 5, wherein the packet data protocol quality of service profile is used as an in-use quality of service profile for services of the network element in case no secondary packet data protocol context is active.

7. The method according to any one of the claims 1 to 6, further comprising: sending the aggregated quality of service value to a network element (307, 317, 327; 310, 320, 330) .

8. The method according to any one of the claims 1 to 7, wherein the accounting request message is an accounting request start message (301) .

9. The method according to any one of the claims 1 to 7, wherein the accounting request message is an accounting request interim message (311) .

10. The method according to any one of the claims 1 to 7, wherein the accounting request message is an accounting request stop message (312) .

11. The method according to claim 10, further comprises removing a packet data protocol quality of service profile from a network access server context data structure (322), wherein the packet data protocol quality of service profile is correlated to the accounting request stop message (321) .

12. A network element (204), in particular a network access server, comprising: a receiving unit adapted to receive an accounting re- quest message; a determining unit adapted to determine an active secondary packet data protocol context; and processing unit adapted to determine an aggregated quality of service value based on quality of service requirements of active primary and secondary packet data protocol contexts in case a secondary packet data protocol context is active.

13. A communication network comprising: a network element according to claim 12, a gateway GPRS support node, and a user equipment.

14. A program element, which, when being executed by a processor, is adapted to control or carry out a method according to claim 1.

15. A computer-readable medium, in which a computer program is stored which, when being executed by a processor, is adapted to control or carry out a method according to claim 1.