Realizing real-time charging

In the early 1990s, UK operators attempted to offer a credit-limiting service that processed call detail records as they arrived from the network and blocked service when customers reached their credit limit.


The service was unsuccessful because processing was delayed, sometimes up to several days, failing to fulfill user expectations and service providers' objectives.


In 1996, prepaid services emerged as a value-added service. Many people in the industry saw these as yet another "˜phase,' but in reality prepaid services helped revolutionize the way in which the industry did business.


Prepaid services have accelerated growth in the industry and currently provide the main source of new subscribers -- 75% of net subscriber additions. At present, 64% of all mobile users have prepaid subscriptions.


Today, prepaid subscriptions are no longer considered a value-added service. Instead, they are one of two payment options enabled by an underlying real-time charging environment.


This environment, however, has presented the industry with a number of challenges, mainly because no standards have been defined for it. Many different approaches have been tried, including SIM-based applications, call trunking nodes, and intelligent network implementations.


The problem is not only one of standards, however. There is also the issue of how charging is to be handled in real time.


Before the advent of prepaid services, only non-real-time mechanisms had been defined for charging in existing standards. An interim approach was needed.


In 2000, the intelligent network (IN) approach was adopted as the de facto standard for managing real-time, prepaid voice services. This solution fell short for non-voice services, however, and one additional approach emerged: service-charging nodes.


The use of service-charging nodes to handle charging on behalf of service nodes in the network introduced other complications.


Real-time charging


Real-time charging differs from non-real-time charging.


Among other things, it is an integral, ongoing part of the service, not an after-service activity. It does not handle the aggregation of information.


In converged networks, each user will have several different identities and it must accommodate different classes of service and content. It was not standardized until 2005 when the 3GPP IMS standard was defined.


It must accommodate different classes of service ranging from the bearer to service and content. Service providers must be able to price usage in ways that are acceptable and understandable for users and at the same time create healthy margins.


Considering these differences, a real-time charging mechanism was defined together with a process for assessing charges for specific services.


The real-time charging network contains two main components: serving elements, which deliver the service to the user; and charging control, which identifies the service and user, calculates the charge, and updates the subscriber account according to the service provider's business rules.


In this model, a serving element must be intelligent and able to interact directly with charging control over an efficient charging interface - Diameter.


A third element, business support, invoices postpaid subscribers and performs other enterprise-specific functions, such as managing the general ledger and accounts receivables. It also logs and stores a history that can be used for data mining and to provide detailed usage specifications for any user.


The diminishing role of IN


In the IN model, the service control point and service switching function work together to form a serving element. IN was designed for call-completion services, not charging.


Although it is the de facto charging standard for real-time voice in present-day networks, it is not the mechanism of choice for future networks, such as the IP multimedia subsystem and broadband access networks.


Instead, serving elements interact directly with charging control over the Diameter protocol. Based on the information they receive from charging control, serving elements can control the delivery of service, including voice.


The user communication function in charging control informs users immediately after service usage about the actual cost of the service; bonuses, if they have received any; and current account balance.


In the context of W-CDMA and GSM, this communication is typically handled via unstructured supplementary services data.


The central charging control functions are common for all access networks, resolving, for example, user identities provided by the access networks.


Therefore, the serving elements need only be able to charge in real time and communicate with charging control over Diameter.


Price flexibility


In modern networks, different services are stacked on top of each other to realize end-user services. Service providers should be able to price the delivered services flexibly and in ways that are acceptable for users while creating healthy margins.


Peer-to-peer MMS, for example, requires underlying GPRS. Music services also require underlying GPRS so that users can browse a catalog of available songs. The same GPRS session may be used for sending and receiving person-to-person MMS. In addition, MMS can be used to deliver a song when the user has made his selection.


The charging infrastructure and rating rules must know how to identify different services and charge, or not charge, for them. At this point, packet-inspection technology has an integral part in enabling network charging, making it possible to monitor the bearer and ensure that data to and from various destinations is, or is not, charged for.


In IMS, the role of the packet-inspection function has been extended to enforce service requirements. The function ensures that sufficient bandwidth is made available for a particular service, guaranteeing good user experience.


Also, because control and payload are separated in the IMS network, the function makes certain that the requested service and actual usage match.


Evolution towards real-time charging


A clear trend in the industry is to employ real-time charging for high-value services to limit operators' risk of financial exposure to third parties. More and more operators are migrating their charging infrastructures toward convergent, real-time charging.


IMS is the first solution to integrate real-time charging from the outset.



Many operators will probably continue to run existing networks with existing charging and billing systems, but will implement modern, convergent, real-time charging systems for their IMS networks.


Initially, they will invest in and deploy a scalable and flexible charging-control function, integrating it with business support systems. They will then evolve the serving elements in the service network and connect them to charging control.



Regardless of payment method, end-users will benefit from up-to-date information on charges and fees for communication and media services - giving them vastly improved control over spending.


At the same time, service providers can protect their bottom line by limiting service to users who can afford it. Operators thus have greater control over credit.


Real-time charging makes it possible to inform users of service costs and, more importantly, the bonuses they have earned. This kind of customer intimacy, which is unrivaled in other industries, is an extremely powerful asset that can be used for building trust and closing the gap between service providers and subscribers.


Jaco Fourie is director for business development and strategies for revenue management at Ericsson