Skip to main content
European Commission logo print header

Proteus and object-oriented frameworks applied to network data acquisition evolutionary solutions

Ziel

Requirements in the telecommunications software sector change extremely rapidly due to the introduction of new technologies and increasing competition resulting from de-regulation. Interoperability of new applications with legacy systems is also an imperative in a market where companies still operate a vast amount of Operation Support Systems (OSS) that were developed several decades ago.
Software reuse has been recognised as a strategic imperative for these companies. In particular, the production of software in a domain engineering context is most appropriate to foster software reuse at company-level, and thus reduce time-to-market, increase software quality, facilitate inter-operability of applications, and capture domain knowledge. PROFANES aims to reach the following 3 objectives:

- Demonstrate that Domain Analysis techniques derived from the PROTEUS method are adequate to capture the knowledge about an evolutionary domain.
- Demonstrate that Object-Oriented Frameworks design techniques are adequate and efficient to capture in a stable model the generic features of a domain component.
- Demonstrate that Domain Analysis techniques combined with Object-Oriented Frameworks together enable software reuse within a domain, by allowing multiple instantiations of a domain component, thus reducing overall time-to-market, reducing costs, and increasing software quality.

THE EXPERIMENT

The starting point for PROFANES will be the Domain Analysis methodology proposed in the PROTEUS ESPRIT project #6086. This method will be extended and adapted to fit into Sodalia's engineering environment. The resulting Domain Analysis methodology will be applied experimentally to define a Domain Model for a family of network data collection applications. Following this activity, an Object-Oriented Framework will be designed to demonstrate the correctness and completeness of the Domain Model. This framework will be made of classes that capture the generic features of the domain, and its variability in the various applications that compose it. Finally, two implemented prototypes representing sample application components of the domain under study will be built in order to validate the reuse potential of the framework. Results of the experiment will be measured by assessing through a series of metrics, the quality of the domain model, the stability over time of the derived object-oriented framework, and its reusability. In addition, a questionnaire will be submitted to the entire project team in order to measure overall satisfaction relative to the use of the Domain Analysis and Object-Oriented Framework methods. Sub-contracts for training and consulting will be offered to INTECS Sistemi and the University of Trento.

EXPECTED IMPACT AND EXPERIENCE

At the end of the PIE, Sodalia aims to obtain an improved software development process which integrates domain engineering techniques into it, so as to have a method that helps managing the complexity and evolution of families of similar systems. We also aim that the PIE results will help to promote systematic domain-specific reuse practice (a Sodalia management imperative) allowing our engineers to develop software applications faster, at a lower development cost, and with increased quality. At European level, we have noticed that the software development community remains squeamish of using reuse-based techniques in a full scale, consistent way. Thus, one foreseen outcome of the project is also to provide convincing evidences that these techniques are worthwhile, and encourage the software industry throughout Europe to reap the advantages of using domain analysis and framework technologies as an integral part of their software development process.

Aufforderung zur Vorschlagseinreichung

Data not available

Koordinator

Sodalia Spa
EU-Beitrag
Keine Daten
Adresse
Via Valentina Zambra 1
38100 Trento
Italien

Auf der Karte ansehen

Gesamtkosten
Keine Daten