WO2014180949A1

WO2014180949A1 - A method for operating a network and an according network

Info

Publication number: WO2014180949A1
Application number: PCT/EP2014/059434
Authority: WO
Inventors: Gottfried Punz; Andreas Maeder
Original assignee: Nec Europe Ltd.
Priority date: 2013-05-08
Filing date: 2014-05-08
Publication date: 2014-11-13

Abstract

For allowing a very efficient traffic handling by simple means a method for operating a network, particularly a radio network, is claimed, wherein a management of traffic from a plurality of User Equipments, UEs, is performed by applying one or more definable policy rules on UE data sessions. The method is characterized in that at least one generalized policy rule for application on UE data sessions is defined based on at least one condition and that the at least one generalized policy rule is applied on UE data sessions matching said condition. Further, an according network is claimed, preferably for carrying out the above mentioned method.

Description

A METHOD FOR OPERATING A NETWORK AND AN ACCORDING

NETWORK

The present invention relates to a method for operating a network, particularly a radio network, wherein a management of traffic from a plurality of User Equipments, UEs, is performed by applying one or more definable policy rules on UE data sessions. Further, the present invention relates to a network, particularly a radio network, comprising means for performing a management of traffic from a plurality of User Equipments, UEs, by applying one or more definable policy rules on UE data sessions.

Abbreviations:

ADC Application Detection Control

A-SDF Aggregated SDF

EPC Evolved Packet Core

IP-CAN IP Connectivity Access Network

LTE Long Term Evolution

NW Network

PCC Policy and Charging Control

PCEF Policy and Charging Enforcement Function

PCRF Policy and Charging Rule Function

P-GW Packet Data Network Gateway

RAN Radio Access Network

RAT Radio Access Network Technology

SDF Service Data Flow

SPR Subscription Profile Repository

TDF Traffic Detection Function

UDR User Data Repository

UMTS Universal Mobile Telecommunication System

UPCON User Plane Congestion Management

In case of user plane congestion in a mobile cellular radio network, such as 3GPP UMTS or LTE, the Packet Data Network Gateway, P-GW, and the co-located Policy Control Enforcement Function, PCEF, as the traffic aggregation point in the network have to carry and manage high traffic volumes from potentially many - tens of - thousands of subscribers at line speed. Certain classes of applications generate mainly short-lived data flows, e.g. web surfing, Apps, instant messaging, etc., which cannot effectively be managed by means of traffic shaping, rate limitation, etc. on a per-flow basis. For this type of traffic it is more appropriate to implement congestion mitigation schemes, which operate on traffic aggregates.

3GPP has developed an elaborate framework for policy handling in a mobile network with Policy and Charging Control, PCC, see 3GPP TS 23.203, "Policy and charging control architecture". It relies on the concept of IP-CAN sessions and Service Data Flows, SDFs, i.e. the level of granularity is per established access NW connectivity per UE - in selected cases a UE may have established connectivity via several access NWs in parallel - and the SDFs within an IP-CAN session. Such granularity was adequate so far, because in normal operation mobile terminals - and thus their data flows - need to be treated individually, according to their subscribed and requested services. Two models are supported:

1. IP-CANs supporting bearer characteristics, e.g. 3GPP and 3GPP2 RANs;

2. bearer-less IP-CANs, e.g. WLAN access NWs.

The concept is visualized in Fig. 1. For every IP-CAN session in PCEF, that is, in the gateway in the user data plane, a corresponding session exists in the Policy Control and Rule Function, PCRF, and allows to control the former via the Gx interface. Gx is DIAMETER based, i.e. we see also a corresponding number of DIAMETER sessions. PCRF may interface to SPR/UDR and thus has available detailed information on individual subscribers.

In this example, 4 different SDFs are transported on bearer 1 within IP-CAN session 1. IP-CAN session 2 does not have bearers, but transports another SDF, as does IP-CAN session 3 on yet another - it's only - bearer. IP-CAN sessions 2 and 3 both transport only one SDF each. The aspect of bearer handling is not important for the rest of this invention.

Within this current concept, no efficient handling of IP-CAN sessions is possible.

Fig. 2 depicts the high-level flow of information for control of an IP-CAN session by PCRF, according to 3GPP TS 29.213, "Policy and Charging Control signaling flows and Quality of Service (QoS) parameter mapping"; in this case even the more general situation with standalone Traffic Detection Function, TDF, a capability to detect Service Data Flows in the user plane, is shown. But note that the Application Detection Control, ADC, function could also be collocated with PCEF.

Policy control through PCEF/TDF is performed as following:

1. each PCC rule is associated to exactly one IP-CAN session, see 3GPP TS 23.203, "Policy and charging control architecture".

2. SDFs are detected and bound to IP-CAN sessions by means of SDF templates, e.g. by using IP-5-tuples <src-ip; dst-ip; src-port; dst-port; protocols

3. Policy control, e.g. traffic limitation, is applied according to the PCC rule for each SDF individually.

This means that in the worst case for each SDF the PCEF/TDF has to request a new policy from the PCRF.

In summary, both policy enforcement - in PCEF/TDF - and policy control - in PCRF - are in current systems performed on the granularity of SDFs, creating unnecessary overhead on the control plane and reducing efficiency of traffic handling for certain types of traffic, e.g. short-lived flows. The description above describes the established state of the art according to 3GPP TS 23.203 and TS 29.212. Further established art is found in WO 2010/086013 A1 , "Group Session management for policy control", which describes a group session concept between PCRF and PCEF, characterized in the following:

- groups are defined dynamically in either PCRF or PCEF;

- after creation, PCEF keeps an association of IP-CAN sessions belonging to a group in the bearer plane;

- if a group is created at the PCEF, it informs PCRF to create the group session via Gx; in this case one or more "criteria" is indicated from the PCRF to PCEF, otherwise the criterion is applied only in PCRF; - if a group is created at the PCRF, it informs PCEF about to create the group session via Gx;

- different criteria are mentioned such as location area, congestion, and subscription.

As such, this method is based on group Gx sessions, which are created and maintained by explicitly adding, confirming, and removing individual user sessions, see Fig. 3. This includes also reporting for group events. 3GPP has found that, due to the general strong increase of data transmission by e.g. smart phones, and for the benefit of optimal control in special situations like disasters or mass events, means are required to prioritize, reduce and limit data traffic. Requirements related to user plane congestion management, UPCON, have been included in 3GPP TS 22.101 , "Service aspects; Service principles" and 3GPP SA2 has started a feasibility study with the aim to enhance the system architecture correspondingly, see 3GPP TS 23.705, "System Enhancements for User Plane Congestion Management". Solutions of widely different characteristics are currently being proposed; one class of solutions is based on congestion notification from the RAN and intends to apply congestion related policies in PCEF. In view of the shortcomings of the known solutions it is an object of the present invention to improve and further develop a method for operating a network and an according network for allowing a very efficient traffic handling or management by simple means.

In accordance with the invention, the aforementioned object is accomplished by a method comprising the features of claim 1 and by a network comprising the features of claim 23. According to claim 1 the method is characterized in that at least one generalized policy rule for application on UE data sessions is defined based on at least one condition and that the at least one generalized policy rule is applied on UE data sessions matching said condition. According to claim 23 the network is characterized by means for defining at least one generalized policy rule for application on UE data sessions based on at least one condition and by means for applying the at least one generalized policy rule on UE data sessions matching said condition. According to the invention it has been recognized that it is possible to allow a very effective handling or management of traffic by using at least one generalized policy rule for application on UE data sessions. Such a generalized policy rule is defined based on at least one condition which can be selected for a suitable influence on traffic management. In a further step the at least one generalized policy rule is applied on UE data sessions matching said condition. Thus, the approaches of managing traffic only on a per UE data session basis is enhanced by a concept of aggregated handling of policies by at least one generalized policy rule. The defined generalized policy rule will only be applied on UE data sessions matching said condition. An explicit identifying and associating of a policy rule with a specific UE data session is no longer necessary. As a result, a very efficient aggregated traffic handling is possible by simply defining a suitable generalized policy rule based on at least one suitable condition. Depending on the individual situation the at least one generalized policy rule can be applied on a set of individual UE data sessions or on an aggregate of UE data sessions matching said condition. This means that UE data sessions can be seen and handled as an aggregate of UE data sessions matching said condition or as a set of individual UE data sessions each matching said condition.

The at least on condition responsible for definition of the at least one generalized policy rule can be selected depending on individual aims of management of traffic from the UEs. Concretely, the condition can comprise at least one action and/or event to be executed on a set of individual UE data sessions or on an aggregate of UE data sessions matching said condition.

Further preferred, a set of individual UE data sessions and/or an aggregate of UE data sessions can be created based on the existence of a corresponding action and/or event in the generalized policy rule. Further steps of the inventive method can be performed regarding such a created set or aggregate of UE data sessions. This simplifies the inventive method.

An action operating on individual UE data sessions can comprise triggering user locating reporting, user plane congestion reporting, WLAN offloading of data traffic, packet marking, configuring of bearer properties or Quality of Service, QoS, enforcement. Depending on individual situations further kinds of action can be defined. An action operating on an aggregate of UE data sessions can comprise limiting a data traffic to a definable bandwidth or prioritizing a definable type of traffic before an other definable type of traffic. Depending on individual requirements an appropriate action can be part of the method. The action and/or the event can be executed in different functional entities. Within a preferred embodiment the action and/or event can be executed in a Policy and Charging Enforcement Function, PCEF. An event operating on individual UE data sessions can be triggered if the accumulated number of individual UE data sessions of monitored UE data sessions exceeds a definable threshold. However, also other kinds of events can be defined for operating on individual UE data sessions.

An event operating on an aggregate of UE data sessions can be triggered if a metric based on a data traffic aggregate exceeds a definable threshold. This kind of event is a preferred kind of event. However, other kinds of events can be used for operating on an aggregate of UE data sessions depending on individual requirements.

Within a preferred embodiment each generalized policy rule can be assigned a precedence or priority value. A higher priority means that a generalized policy rule is applied before a generalized policy rule with a lower priority value.

Within a further preferred embodiment the management of traffic can be performed in the control plane and/or user plane. The selection of the location of performance can be dependent on individual requirements and circumstances. Within a further preferred embodiment a generalized PCRF session per connected PCEF can be provided to handle the at least one generalized policy rule. Such a generalized PCRF session can be created at start-up of communication between PCEF and PCRF. However, other creation processes are possible. Such a generalized PCRF session can be used for installation, modification and/or removal of generalized policy rules and/or for generalized event reporting. There can be provided different use cases for such a generalized PCRF session in an advantageous way. A generalized event reporting can be used for monitoring a status of an aggregate of UE data sessions and/or for reporting the congestion level or a derived metric to the PCRF. This will enhance the efficiency of traffic handling. Within a further preferred embodiment the PCRF can update the generalized policy rule via a corresponding generalized PCRF session after having received an event report from the PCEF. Thus, a generalized policy rule can be kept in an actual status corresponding to varying traffic situations.

Within a further preferred embodiment congestion management can be performed by the inventive method and network.

Within a concrete method step a PCEF can evaluate the condition over all matching UE data sessions. Thus, a reliable evaluation of the individual condition or conditions can be provided by the inventive method.

Preferably, the PCEF can send action and/or event triggering messages selectively to all UEs with matching UE data sessions. This will also enhance the reliability of evaluation of the condition or conditions.

Further preferred, the PCEF can monitor data traffic of UE data sessions and can derive at least one metric for triggering an action and/or event. Such a monitoring function further enhances reliability and efficiency of the inventive method.

Within concrete preferred embodiments a UE data session can be an Internet Protocol Connectivity Access Network, IP-CAN, session, a Service Data Flow, SDF, a single traffic flow or a bearer. Further preferred the generalized policy rule can be a generalized Policy and Charging Control, PCC, rule.

Important aspects of embodiments of the present invention can be summarized as follows:

There is defined at least one generalized policy or PCC rule defined by one or a plurality of conditions, actions and events, executed in the PCEF. Such actions and events operate either on a data traffic aggregate or on a set of individual data sessions matched by said conditions. The corresponding data traffic aggregates and/or set of data sessions can be created based on the existence of a corresponding action or event in the generalized policy or PCC rule. Events based on aggregation metrics either for the traffic aggregate or the set of data sessions can trigger a notification towards the PCRF.

Installing/modification/updating of generalized policy or PCC rules in the PCRF can be performed by means of a single generalized Gx session per rule.

Further, there can be provided a prioritization concept to avoid conflicts with existing policy or PCC rules.

Within preferred embodiments policy or PCC rules are based on matching conditions for aggregated handling, independent of IP-CAN sessions.

Further, triggering of actions on traffic aggregates, e.g. for congestion mitigation, is possible. Further, generalized PCRF sessions between PCEF and PCRF can be realized.

The inventive method and network provide a significant reduction of the amount of control signaling on Gx interface. An aggregated modelling of UEs and/or bearers is allowed. Also, an efficient handling of traffic aggregates in PCC is allowed. There are several ways how to design and further develop the teaching of the present invention in an advantageous way. To this end, it is to be referred to the patent claims subordinate to patent claim 1 on the one hand and to the following explanation of preferred examples of embodiments of the invention, illustrated by the drawing on the other hand. In connection with the explanation of the preferred embodiments of the invention by the aid of the drawing, generally preferred embodiments and further developments of the teaching will be explained. In the drawings Fig. 1 is showing a visualization of a known and simplified PCC concept according to 3GPP TS 23.203,

Fig. 2 is illustrating a high level flow of information for control of a known IP-CAN session by PCRF,

Fig. 3 is illustrating a grouping model according to the state of the art,

Fig. 4 is illustrating an embodiment of a method for operating a network according to the invention wherein an A-SET/A-SDF creation based on existence of A-SET/A-SDF actions/events is shown,

Fig. 5 is illustrating the concept of generalized PCC rules and generalized PCRF sessions according to an embodiment of the invention,

Fig. 6 is illustrating an example of PCRF controlled congestion handling based on aggregate policy handling according to an embodiment of the invention and Fig. 7 is showing an example of PCRF controlled non-3GPP offload control based on aggregate policy handling according to an embodiment of the invention.

In the following there is explained the inventive aggregated handling of policies and how this can work and which enhancements in PCRF, Gx and PCEF can be advantageously provided in order to achieve this within preferred embodiments of the invention. Concept of "Generalized policy or PCC rules":

For aggregated handling of policies it is no longer possible - in most embodiments not desirable - to indicate exactly at PCRF for which IP-CAN sessions and for which SDFs a policy shall be applied. Rather than explicitly identifying and associating a PCC rule with an IP-CAN session or with a group of IP-CAN sessions, in this concept a PCC rule is identified by its matching conditions and enforcement actions/events. Such a rule is called "Generalized PCC rule", G-PCC rule. G-PCC rules can be applied in parallel with IP-CAN session PCC rules.

Applying a G-PCC rule in PCEF onto all IP-CAN sessions and their SDFs results in matching of a potentially large set - A-SET - of SDFs, which form a traffic aggregate, or an aggregated service data flow, A-SDF. Specifically, an A-SET is a set of IP-CAN sessions, SDFs, or bearers which matches a set of conditions of a G-PCC rule. An A-SDF is the corresponding data traffic aggregate of an A-SET.

In contrast to IP-CAN session PCC rules, G-PCC rules operate on A-SDFs and/or on A-SETs. For example, QoS control rules on maximum/guaranteed bitrates are applied to the whole traffic aggregate of an A-SDF, thus allowing efficient handling of traffic aggregates.

In the following, the term "data session" refers to an IP-CAN session, an SDF, a single traffic flow or a bearer.

Applying a G-PCC rule in PCEF may include the following procedural steps, as illustrated in Fig. 4: 1. Match all data sessions, e.g. IP-CAN sessions, SDFs, bearers, against the specified condition or conditions.

2. Depending on the type of action as defined within the G-PCC rule o build a set, A-SET, comprising the matched data sessions as target for further actions;

o and/or build a traffic aggregate - denoted as "Aggregate SDF", A- SDF - comprising the totality of the data traffic of the matched data sessions. 3. Calculate metrics based on aggregate traffic of an A-SDF or based on accumulated individual events of A-SETs. These metrics are calculated in the PCEF/TDF.

4. Perform associated actions on A-SET and/or A-SDF, e.g. triggering of events, notification of PCRF.

Examples of matching conditions are e.g.

1. the area where a UEs is currently located;

2. type of bearer allocated for a UE;

3. congestion status of the UE, the RAN area or cell where the UE is located, or of the data path between UE and core network;

4. RAT type;

5. type of application;

6. type of service;

7. type of traffic source, e.g. unattended vs. attended traffic;

8. group identifier, e.g. for M2M type of communication.

The following types of actions are defined:

1. Actions operating on individual data sessions. One example is to switch on user location reporting for a plurality of users reporting user plane congestion for IP-CAN sessions. Another example is to drop all bearers of a QoS Class Index, QCI, of low priority in case of reported heavy user plane congestion. A further example is to trigger WLAN offloading for all users with ongoing video download.

2. Actions operating on data traffic aggregates, A-SDF. One example is limiting the data traffic of an A-SDF to a certain bandwidth for purpose of congestion mitigation. Another example is to prioritize a certain type of traffic before other type of traffic.

Furthermore, the following types of events are defined: 1. Events operating on A-SETs are triggered if the accumulated number of individual data sessions of the monitored data sessions exceeds a defined threshold. For example, an event is triggered if out of 100 sessions in an A- SET, for 20 the minimum bandwidth cannot be fulfilled.

2. Events operating on A-SDFs are triggered if a metric based on the data traffic aggregate exceeds a certain threshold, e.g. a certain data volume. For example, an event is triggered if the average minimum bandwidth of the A-SDF traffic aggregate cannot be fulfilled.

As such, there is no need for tracking group memberships at the PCRF via the Gx interface.

In order to simplifying conflict handling with IP-CAN session PCC rules, each G- PCC rule has a precedence, or priority value. A higher priority means that a G- PCC, and specifically the action triggered by it, is applied before G-PCC rules and IP-CAN session PCC rules with a lower priority value. For normal IP-CAN session PCC rules without a priority value, one can assume a default value which allows G-PCC rules with higher and lower priorities, e.g. a priority value of "5" for IP-CAN session PCC rules allows G-PCC rules with higher and lower priorities.

It is up to the operator to avoid conflicts, e.g. for triggering specific actions such as off-loading or QoS enforcement. The concept is detailed in Fig. 5. In this figure, three G- PCC rules are established in the PCEF, which are applied to a set of IP-CAN sessions and on the SDFs which they transport.

In case of Proxy Mobile IP, PMIP, on S5/S8/S2a/S2b, i.e. usage of Gxx interface and Bearer Binding and Event Reporting Function, BBERF, bearer-related functionality like bearer dropping, offload, or per-bearer rate limiting, etc. can be perfomed by involving BBERF. Alternatively, the functionality performed in the PCEF/TDF can also be implemented in the BBERF, e.g. in roaming scenarios. Concept of aggregated PCRF and Gx sessions:

Currently IP-CAN sessions on Gx level, i.e. PCRF sessions, are uniquely identified by Subscription-Id and PDN-Connection-ID, possibly together with RAT-type, see 3GPP TS 29.212, "Policy and Charging Control, PCC, over Gx/Sd reference point". On DIAMETER level the unique identification is provided by Session-Id; for every instantiation of an IP-CAN session a different DIAMETER Session-Id will be used. For the new concept described above, i.e. aggregate or aggregated policy handling, a generalized PCRF session is required per connected PCEF to handle G-PCC rules. It should be created at start-up of communication between these nodes and is used for installation, modification and removal of G-PCC rules as well as generalized event reporting.

We propose to add actions and event reporting for aggregated policy handling as defined in the above part "Concept of "Generalized policy or PCC rules"". Actions can be used to trigger specific handling in the user data plane and associated control plane for all applicable IP-CAN sessions, that is, all which fulfill the condition, e.g.:

1. trigger offload to non-3GPP access;

2. packet marking, for support of RAN-based mechanisms;

3. rate limit data traffic;

4. re-negotiate bearer properties.

Event reports can be used to monitor the status of a whole, large set of IP-CAN sessions or SDFs - possibly all those currently handled on the PCEF - as an aggregate. To this end, the occurrence of user plane overload in a 3GPP RAT will be detected whenever one of the UEs currently handled in the corresponding RAT experiences congestion in the user plane. One example of handling RAN user plane congestion in such a way is to include an indication, e.g. a RAN user plane congestion information, RCI/RUCI, element in the event report to the PCRF, comprising a congestion level for a RAN cell, UE, or bearer. Procedures:

As a first example, we show in Fig. 6 a procedure by which PCRF can support congestion handling in the 3GPP RAN, based on the concept of aggregate policy handling described above.

The following steps are performed:

0. as a pre-requisite, the PCRF has requested from PCRF event reporting of type of congestion using a generalized PCRF session;

1. UEs have established their data communication via the 3GPP RAT normally;

2. At one point the 3GPP RAN runs into congestion and notifies this status to the PGW. The assumption here is that this happens in the user plane, but note that this is only one of several possibilities; a similar notification could also be achieved via the control plane - signaling via MME, S-GW to P-GW.

3. The congestion is reported up to PCRF in the generalized PCRF/Gx session.

4. The PCRF installs a policy rule with conditions "congestion" and "bearer has QCI type 2" - video - and an action "packet marking".

5. The PCEF evaluates the condition over all IP-CAN sessions.

6. Based on the result of step 5, the PCEF activates packet marking for all packets matching both conditions - UE is in congested area and the bearer is of QCI value 5.

7. Packets not matching the condition are not marked.

In this example, for simplicity, the conditions result in un-mixed packets - either all marked or all unmarked - being delivered per UE; this would be the case if all UEs in the congested areas incidentally have only video streaming ongoing. In reality many UEs would receive both marked and unmarked packets, according to how the two conditions match.

In the second example, see Fig. 7, the system likes to push offload data traffic like youtube and p2p from 3GPP access towards non-3GPP access, for multimode capable UEs that are in coverage of an alternative, non-3GPP RAT. This may be e.g. according to a well-known schedule and location, like in dense metro areas at peak traffic hours, e.g. Shinagawa station in Tokyo. It is assumed here that specific offload policies have been provisioned via Access Network Discovery and Selection Function, ANDSF, to UEs in advance and that a mechanism for triggering them from EPC is available.

The following steps are performed:

0. PCRF has established - at some earlier point in time - a generalized Gx session with PCEF/P-GW.

1. UEs have established communication / bearers via the 3GPP RAT.

2. According to a well-known schedule, PCRF detects that it is time to offload non-essential services from 3GPP access.

3. PCRF activates a generalized policy rule towards PCEF, with the condition describing in which area and for which type of traffic, e.g. youtube and p2p, the action should apply. The action is defined as "trigger offload to non- 3GPP access".

4. The PCEF/P-GW evaluates the condition and determines which UEs are affected. Note: the location information of UEs may be known when location reporting is on.

5. The PCEF/P-GW sends messages selectively to affected UEs, triggering the non-3GPP ANDSF offload policies. Here it is assumed that a corresponding capability in the control plane exists.

Further Embodiments:

The following embodiment describes the application of G-PCC rules for the purpose of user plane congestion management:

- The PCRF provides G-PCC rules at start-up time to PCEF/TDF.

The G-PCC rules contain: conditions for application type / IP flows, subscriber class, congestion level

actions for enforcing minimum, maximum, or target values of bandwidth for A-SDFs, traffic aggregates

events which are triggered by the PCEF/TDF and report to the PCRF, e.g. once the enforced actions cannot be enforced without violating minimum bandwidth limits and/or QoS targets.

The actual traffic management and scheduling is left for implementation in the PCEF/TDF. For evaluating the G-PCC rules, the PCEF/TDF monitors the traffic and derives appropriate metrics to trigger events/actions.

- The PCRF updates the G-PCC rules via the corresponding aggregated Gx sessions, e.g. as soon as it receives an event report from the PCEF.

Many modifications and other embodiments of the invention set forth herein will come to mind the one skilled in the art to which the invention pertains having the benefit of the teachings presented in the foregoing description and the associated drawings. Therefore, it is to be understood that the invention is not to be limited to the specific embodiments disclosed and that modifications and other embodiments are intended to be included within the scope of the appended claims. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.

Claims

C l a i m s

1. A method for operating a network, particularly a radio network, wherein a management of traffic from a plurality of User Equipments, UEs, is performed by applying one or more definable policy rules on UE data sessions,

c h a r a c t e r i z e d in that at least one generalized policy rule for application on UE data sessions is defined based on at least one condition and that the at least one generalized policy rule is applied on UE data sessions matching said condition.

2. A method according to claim 1 , wherein the at least one generalized policy rule is applied on a set of individual UE data sessions or on an aggregate of UE data sessions matching said condition.

3. A method according to claim 1 or 2, wherein the condition comprises at least one action and/or event to be executed on a set of individual UE data sessions or on an aggregate of UE data sessions matching said condition.

4. A method according to claim 3, wherein a set of individual UE data sessions and/or an aggregate of UE data sessions is created based on the existence of a corresponding action and/or event in the generalized policy rule.

5. A method according to claim 3 or 4, wherein an action operating on individual UE data sessions comprises triggering user locating reporting, user plane congestion reporting, WLAN offloading of data traffic, packet marking, configuring of bearer properties or QoS enforcement.

6. A method according to one of claims 3 to 5, wherein an action operating on an aggregate of UE data sessions comprises limiting a data traffic to a definable bandwidth or prioritizing a definable type of traffic before an other definable type of traffic.

7. A method according to one of claims 3 to 6, wherein the action and/or event is executed in a Policy and Charging Enforcement Function, PCEF.

8. A method according to one of claims 3 to 7, wherein an event operating on individual UE data sessions is triggered if the accumulated number of individual UE data sessions of monitored UE data sessions exceeds a definable threshold.

9. A method according to one of claims 3 to 8, wherein an event operating on an aggregate of UE data sessions is triggered if a metric based on a data traffic aggregate exceeds a definable threshold.

10. A method according to one of claims 1 to 9, wherein each generalized policy rule is assigned a precedence or priority value.

1 1. A method according to one of claims 1 to 10, wherein the management of traffic is performed in the control plane and/or user plane.

12. A method according to one of claims 1 to 1 1 , wherein a generalized PCRF session per connected PCEF is provided to handle the at least one generalized policy rule.

13. A method according to claim 12, wherein the generalized PCRF session is created at start-up of communication between PCEF and PCRF.

14. A method according to claim 12 or 13, wherein the generalized PCRF session is used for installation, modification and/or removal of generalized policy rules and/or for generalized event reporting.

15. A method according to claim 14, wherein generalized event reporting is used for monitoring a status of an aggregate of UE data sessions and/or for reporting the congestion level or a derived metric to the PCRF.

16. A method according to one of claims 12 to 15, wherein the PCRF updates the generalized policy rule via a corresponding generalized PCRF session after having received an event report from the PCEF.

17. A method according to one of claims 1 to 16, wherein congestion management is performed.

18. A method according to one of claims 1 to 17, wherein a PCEF evaluates the condition over all matching UE data sessions.

19. A method according to claim 18, wherein the PCEF sends action and/or event triggering messages selectively to all UEs with matching UE data sessions.

20. A method according to claim 18 or 19, wherein the PCEF monitors data traffic of UE data sessions and derives at least one metric for triggering an action and/or event.

21. A method according to one of claims 1 to 20, wherein a UE data session is an Internet Protocol Connectivity Access Network, IP-CAN, session, a Service

Data Flow, SDF, a single traffic flow or a bearer.

22. A method according to one of claims 1 to 21 , wherein the generalized policy rule is a generalized Policy and Charging Control, PCC, rule.

23. A network, particularly a radio network, preferably for carrying out the method according to any one of claims 1 to 22, comprising means for performing a management of traffic from a plurality of User Equipments, UEs, by applying one or more definable policy rules on UE data sessions,

c h a r a c t e r i z e d by means for defining at least one generalized policy rule for application on UE data sessions based on at least one condition and by means for applying the at least one generalized policy rule on UE data sessions matching said condition.