Database interoperability issues in intelligent networks

Intelligent Networks are evolving as data-intensive multi-layer information systems. Databases can be found within or associated with various functional entities of the IN reference model. The aspects of database distribution and heterogeneity are present both within the functional entities and between them, in a real implementation. In particular, the Service Data Function (SDF) has to be implemented in such a way that two different physical entities: the Service Control Point and the Service Management Point have access to the SDF. What makes an IN system a different kind of a distributed and heterogeneous information system is that correctness and reliability requirements vary vastly among different physical entities and yet transactions are supposed to span many physical entities (real systems). A reference model of IN data management is proposed whereby the full service lifecycle is addressed, including service creation, deployment, provisioning and execution. Essential data management requirements are presented. The needs of transaction processing is presented from the point of view of service provisioning and execution. A classification of transactions interacting with the SDF is proposed, and a corresponding correctness criterion is suggested.