Many Telecommunication services involve a wide variety of service delivery components, both hardware and software, that are used together to create and implement the service. The problem of collecting, aggregating, and combining service records is often labeled Billing Mediation, and for many applications these tasks need to be done in a continuous, near real-time process.

Range of Services

The range of such service elements is broad:

The list goes on. In order to create a complete service record, data need to be collected, processed, and combined from many types of service elements and many vendors.

Issues and Timing

Differing aspects need to be addressed and combined:

For some services, it is sufficient to calculate these elements on a daily basis. Increasingly, however, people expect to get real-time or near real-time records of their service usage, and some applications fundamentally require it in order to operate effectively, such as pre-paid billing.

Relational Relevance

The relational model is an elegant and powerful way of aggregating and combining data coming from many sources to create complete records. Relational iteration, aggregation, filtering, and joining operators are well suited to the task.

However, when the volumes of data are great, using RDBMS for these naturally relational data massaging tasks encounters difficult issues of cost, efficiency, and effectiveness:

• Performing queries concurrently with high rates of inserting records into tables becomes expensive: When should the queries be run?

• Duplicate or incomplete answers are common in these circumstances. You only want to run the queries when the last element of the service record has arrived.

• De-duplicating and repeatedly processing incomplete data is very expensive and wasteful.

• Difficult issues arise in dealing with time-sensitive data, that is, data that must arrive within a given time window or constrained by other service-related data. AAA servers handling Authentication, Authorization and Accounting, such as Radius or Diameter, commonly encounter such constraints and their associated time-sensitive constraints and requirements.

RAMMS Advantages

SQLstream's RAMMS provides the perfect answer to the Real-time Billing Mediation problem because it can elegantly handle the following necessary components in the solution:

• Allowing SQL operations to be specified on the changing data

• Detecting when the last element of a service record arrives and then emits the complete record

• Processing the data continuously

• Performing simultaneous cross-checks and aggregations

• Pre-rating the service records, calculating the price or cost based on diverse tariff rules expressed in SQL

• Dealing with esoteric data types and service element protocols using adapters and plug-ins

• Scaling effortlessly to handle very large volumes of records

It will come as no surprise that such applications were some of the sources of inspiration for creating the SQLstream RAMMS.