PCRF the Liberator for Mobile Data Charging

PCRF – Policy and Charging Rules Function – It is not too old concept introduced in the last quarter of the year 2007. When standards for the 3GPP Policy Charging Control (PCC) architecture were published. According to the 3GPP standard architecture up to Release 9, a Charging System was not supposed to interact directly with a Policy Server. The Charging System; responsible for rating and charging, while the Policy Server is responsible for determining the right policy depending on the kind of traffic.

PCRF – An important Entity in LTE Network

Prior to launching a new service like VoLTE, MNO’s need to test and confirm their policy rules within the PCRF and PCEF to make sure the services are delivered with integrity and to make sure that there is enough capacity to offer the requested services.   Charging rules are very similar and must be validated. MNO or data service provider may need to carry out a multitude of charging rules for each service; and these rules may differ based on a variety of conditions, such as Customer Service Level Agreement, time of day, or network conditions.

Gx interface between PCRF and PCEF. PCRF generates PCC rules and sends the generated PCC rules towards PGW/PCEF to enforce the PCC rules at PGW for an end User. Gy interface between PCEF to OCS. It is used for transfer data uses a report from PGW/PCEF to OCS. So both the interfaces meet different goals.

The way PCRF performs in 4G networks, most likely in 5G the same job will be done by PCF i.e. to provide policy rules for control plane functions.  The main job would be as below

•  Network slicing
• Roaming and mobility management.

For example, service providers can use PCRF to charge subscribers based on their volume of usage.  This usage will be driven out of high-bandwidth applications, charge extra for QoS guarantees. The subscriber can limit app usage while he/she user is on roaming, or lower the bandwidth of wireless subscribers using heavy-bandwidth apps during peak usage times.

Right Policy vs Wrong Policy – Pattern Detection

In reality, it seems that the decision about the right policy might be influenced by some real-time subscriber information basis on machine learning algorithms to detect patterns. This might also be relevant for charging and therefore stored in the Charging System. The goal of this post is to understand and take feedback for learning on what such an interface would look like, based on an actual implementation. Learn a scenario where the policy should change in real-time during a data session because a volume threshold crossed.

Mobile Network Operators are facing two challenges (thanks to the deployment of 3G/LTE/4G cellular networks VoLTE is around the corner like Christmas gift) how to optimize and monetize their networks as voice revenue continues on a southbound journey with excellent speed. As per the Google research, mobile broadband traffic is projected to increase by up to 35 times by end 2015/2016, the response to these two issues could mean the difference between significantly growing business over the next five years or fall behind the competition.

As mentioned before many of the MNOs have, in search of revenue actually cannibalised their own voice revenue by encouraging revenue share based VAS services from VAS service providers as usage on the network remains the same at 1 to 10 USD on average. Under PCRF new revenue creation and monetization of the network is easily driven with the ability to dynamically change the quality of service (QoS) for an individual subscriber who creates incremental revenue i.e Tiered Plans and etc.

Patterns detections Supervised and Unsupervised Learning Method

Worldwide predictions on Mobile Data usage are hinting towards data tsunami. The demand for mobile device data is many folds higher compared old days of desktop or laptop is driven data which clearly proves mobile broadband traffic is increasing. Mobile network operators (MNOs) have no choice but to offer more data, more speed but sadly with much less cost. Machine Learning algorithm plays a big role here to cut cost and increase speed.

Packet cores need to deliver innovative, faster & high revenue-generating services to stay in business. PCRF does excellent network resource management as traffic can be prioritized to meet customer service expectations and SLAs; eg: bandwidth congestion. Also, personalization of services are super easy by the value of network assets maximization by offering services that meet the personal needs of subscribers; eg: high bandwidth for a time period

Diameter interfaces give connection among Diameter nodes to enable essential services provider network functions such as authentication, online and offline billing, and policy and charging. PCRF – Dedicated policy equipment standardize in 3GPP that enables the policy function for bandwidth and charging on multimedia networks. The functional element that comprises flow based charging control and policy control decision functionalities.

Policy and charging rules function

The PCRF stipulates network control about the service data flow detection; gating, QoS and flow based charging (except credit management) towards the PCEF. PCRF receives session and media related information from the application function and informs the application function for data traffic. The PCEF is the functional element that encompasses policy enforcement and follows based charging functionalities.

The PCRF interfaces with the main packet gateway and takes charging enforcement decisions on its behalf. The centralized device can act as a policy decision point (PDP) for the wireless operator and gets as granular as each subscriber. Gx and Gy are two main and key interfaces in existence for different purposes. Every service has unique bandwidth requirements. Policy control within the PCRF and PCEF ensures that the right amounts of bandwidth are dynamically allocated to each service in real-time, thus making the most efficient use of network resources.

Monetizing data by leveraging the PCRF

An intelligent policy and charging control solution helps assure the proper allocation of network resources based on what subscribers have purchased and what the network can deliver. In this context, a direct interface between the Charging System and the Policy Server might be required.

At the beginning of Gx/Gy specification (Release 6) it was meant to allow a simple solution, i.e. by letting the OCS control some PCC stuff, i.e. including QoS bearer change (e.g. fair usage policy use case). No extra Gx/PCRF etc. were needed in that case. This option as standards solution was immediately ceased in further 3GPP releases as PCC got more complex (Gx, PCRF functions, LTE concepts, etc.).

However when credit authorization(Gy) and charging rule provision(Gx) are required simultaneously, then it will use Gx over Gy application such that it will avoid signalling load on Network. If we do not use this mechanism then we have to send Gx and Gy message separately. DPI and PCRF help to make sure security & revenue leakage. Attackers are usually looking for ways to get free mobile data services. Without a doubt, these worms do exist and always find loopholes in the charging policy.

Importance of Diameter

Diameter interface is defined for signalling between many core network nodes and services. Overload on these interfaces can lead to server congestion or even collapse. The impact on services can be as follows-

• Denial of services
• Persistent access restriction
• Loss of IMS and broadband services
• Location information for emergency services and lawful intercept loss
• Loss of ability to use policy control to offer service personalization
• Loss of ability to use policy control to optimize network resources and Billing errors
• Major revenue loss and leakage.

An analysis table for these interfaces and service impact is shown in the following slides. (Source: 3GPP S2-122906)

It makes good sense when the same endpoint for both PCRF and OCS functionality in the same box to use Gx over Gy that reduces signalling load over the network. OFCS and OCS are now more up-to-date with interfaces that are being used to exchange signalling data. Gx interface which is responsible for ofcs and its reference point is for provisioning and removal of policy and charging control rules from the PCRF to the PCEF and the transmission of traffic plane events from the PCEF to the PCRF.

The Gx & Gy Interfaces

The Gx reference point can also be used for charging-control, policy-control or both by applying AVPs relevant to the application. In most cases, PCEF is based inside PDNGW (Packet Data Network Gateway). Gy interface allows online credit control for service data flow based charging. HSS is a combination of HLR and AUC in the 2G/3G network, used to keep & update subscribers’ information and authentication. SPR is used to keep subscribers’ profiles/policies – for PCRF via Sp interface. Both SPR and HSS are databases keeping the subscribers’ information but both playing different roles in 3GPP LTE architecture.

HSS basically plays its main function of authenticating the LTE subscribers with IMSI as the primary key for SAE-HSS and IMPU ( IMS Public ID ) + IMPI ( IMS Private ID ) for IMS VoLTE authentication. SPR is the repository keeping the subscriber’s policies & profiles for QoS management. Both HSS and SPR are the front end element because the back-end element i.e. UDR is the databases keeping all information.

Some vendors do provide both in a single box. PCRF or a policy server or a Policy Decision Function (PDF) is the part of the Evolved Packet Core (EPC) that supports service data flow detection, policy enforcement and flow-based charging. The PCRF function is part of the larger PCC architecture, which also includes the Proxy Call Session Control Function (P-CSCF) and the Policy and Charging Enforcement Function.

Reinforcement Learning by Policy Enforcement

Operators allow free data service for certain data flow, but usually, forget or do-not enforce that the transmitted packets indeed belong to the designated free flow. Even worse, no effective mechanism is implemented to limit the traffic volume going through this free ride. Consequently, these loopholes can be exploited to enable any form of mobile data services for free. The 3G standards offer the operators flexibility to define their own charging policies.

Unfortunately, in some of MNO cases, their policies and implementations may contain serious flaws. The PCRF, PCEF and the Charging Functions in the IMS and EPC core networks drive policy and charging rules.  These elements give MNO’s with the ability to differentiate services while maximizing revenue. To validate policy, simulation of PCRF and PCEF, the application functions (AF) and the Charging Functions is required. By simulating these elements with the right tools, testing of network elements for rules implementation, error handling, and the ability to do under stress conditions can be observed well.

PCEF and PCRF

We can surely discover loopholes and simple attacks, which can be validated by experiments over operational networks. Enables real-time management of network resources in-sync with subscribers and applications. Comply with 3GPP policy and charging rules function and other industry standards. Provides real-time, in-session policy decisions. Open towards any PCEF.

The PCRF and PCEF are key to service integrity for our customers. Policy and charging rules should be verified before the launch of new services. PCRF is an important part of IMS architectures, although it is not exclusive to the 3GPP-based network in which it was certified. It works across wireless networks and can come pre-integrated in a standard IT server. Ultimately, the testing and monitoring of policy and charging functions guarantee services can be delivered appropriately, which leads to customer satisfaction and a guaranteed revenue stream.

In this work, if we conduct experiments on working 3G networks to study the security implication of such architecture and practice. Investing in the validation of policy and charging elements ensures that equipment vendors software applications do in accordance with MNO’s desired and installed policies. Testing also ensures that network elements are engineered for peaks in capacity, so MNO’s need to have correct and required tools that can drive the expected greatest loads of traffic and make sure zero revenue leakage.

Points to Note:

AI and ML are now making the most powerful tools in a marketer’s arsenal for improving return on investments. All credits if any remains on the original contributor only. We have covered all basics around mobile data models or the importance of mobile data. In the next upcoming post will talk about implementation, usage and practice experience for markets.

Books + Other readings Referred

• Research through open internet, news portals, white papers, notes made at knowledge sharing sessions and from live conferences & lectures.
• Lab and hands-on experience of  @AILabPage (Self-taught learners group) members.

Feedback & Further Question

Do you have any questions about AI, Machine Learning, Data billing/charging, Data Science or Big Data Analytics? Leave a question in a comment section or ask via email. Will try best to answer it.

Conclusion:  PCRF – DPI Compatibility Matrix and PCRF supporting Mobile data; in a big way supports the need of time as the Internet is going wireless and mobile. Two driving forces for this trend have been the explosive growth of smartphones and the rapid deployment of 3G/4G infrastructure. Unlike the wired Internet, cellular networks have implemented usage-based charging, rather than the simpler flat charging. The 3G/4G standards stipulate the accounting architecture; yet offer freedom for the MNO to define their own charging policy.

====================== About the Author =================================

Read about Author at: About Me   

Thank you all, for spending your time reading this post. Please share your feedback / comments / critics / agreements or disagreement.  Remark for more details about posts, subjects and relevance please read the disclaimer.

FacebookPage                Twitter                          ContactMe                          LinkedinPage    ==========================================================================