BackMediationZone® is mediation system developed by DigitalRoute. Mediation is used for collecting, processing and distributing data files in telco IT. MediationZone® is designed for every kind of fixed, mobile or IP networks providers. The application is written in Java language, so it’s platform and architecture independent. It is high scalable system, it can run in cluster of servers which work parallel. This way it’s easy to improve the performance in case of increasing load.
Workflows are the cornerstones of the MediationZone® system. A workflow is identified as a set of nodes connected to each other to represent a data flow. Each node operates on the data flow in different ways, depending on what type of agent it represents and how it is configured. Having identified this, the workflow can be said to contain three types of agents – collection, processing and forwarding agents. This way we can also divide every workflow into three general areas: Input, Data processing and Output.
Input
MediationZone® collects data files from many kind of external systems – typically telephone switches and similar subsystems of mobile operators (e.g. IN, MMSC, SMSC). Following transfer protocols are supported: FTP, FTP/DX200, SFTP, it is also possible to retrieve data directly from relation database (e.g. Oracle)or directly from disk space. The input files needs to be decoded before processing. The decoders are defined in UDL – Ultra Definition Language, the variety of data formats is supported: text (csv, xml, …), binary(with fixed length of field, TLV, …). Supported are also complex record structures including ASN.1 and TAP-3. After decoding the data are referred to as UDRs – Usage Data Records.
Processing
MediationZone® processes the collected usage records. The UDRs flows through nodes of workflow. Every node is driven by processing agent. Business logic of these agents is defined in APL – Application Definition Language, which is based on Java language. Following the list of operations which is delivery by process agents:
- Check for duplicate records/files
- Validation of records
- Selection and filtering of records
- Aggregate – assembling of partial records
- Create new records
- Gather statistics about data
Following the workflow diagram example:
Output
The UDRs are coded to appropriate format before sending to output. The same could be applied for coders as for decoders – defined in UDL, the variety of data format is supported. The output files are then distributed to downstream system by some transfer protocol. The downstream systems are computer systems that perform functions like billing, fraud detection, archive, operational support, or marketing analysis. It is also possible inserting data directly to relation database (e.g. Oracle) it’s mean also in DWH – Data Warehouse.
Workflow management of MediationZone® cares about consistency of data. In case of occurrence of some output portal error or error during processing of data then the automatic recovering of all changes is performed – the processing is atomic and there can’t by any data corruptions. It’s one of the main benefit of mediation systems. There is also possible to send wrong UDR to ECS – Error Correction System. This way the processing isn’t interrupted and there is still possibility to reprocess the data again. Of course there is also log system in which you can find the errors of different severity.
MediationZone® is scalable, configurable, reliable and robust mediation system. ARTIN has many experiences with delivery this product from many projects for telco providers in Algeria, Austria, Czech Republic and Japan: T-Mobile, HP Algeria – Watania, Nokia Siemens Networks – Vertu, Nokia Siemens Networks – Blyk NL.
