Section: New Results
Contract-based and Aspect Oriented Design
Design of QoS-aware Service Oriented Architectures
We investigated several aspects of Service Oriented Architectures, from methodological issues of handling their developemnt  , to tool support and QoS evaluation.
In the context of the Faros French national project, which aims at building a whole design process for QoS-aware Service Oriented Architectures, we designed a transformation framework to handle QoS contract models in two stages: a first stage transform a business model into a central model, which is business and target platform independent.A second stage transform this central model into a platform model, which is platform dependant. The technical delivrable on business to central model transformation is available on the Faros web site.
Further, in highly dynamic environments, software systems requires a capacity of self-adaptation to fit the environment and the user needs evolution, which increases the software architecture complexity. Despite most current execution platforms include some facilities for handling dynamic adaptation, current design methodologies do not address this issue. One of the requirement for such a design process is to describe adaptation policies in a composable and qualitative fashion in order to cope with complexity. This work introduces an approach for describing adaptation policies in a qualitative way while keeping the compositionality of adaptation policies. This work uses the Fuzzy Logic Theory as a fundation for the adaptation policies language. This approach has been tooled as an extension of the Fractal component model  ,  ,  ,  .
Entimid: a Model Based SOA for Building Automation
In the personal or corporate spheres, the home/office of tomorrow is soon to be the home/office of today, with a plethora of networked devices embedded in appliances, such as mobile phones, televisions, thermostats, and lamps, making it possible to automate and remotely control many basic household functions with a high degree of accuracy. In this domain, technological standardization is still in its infancy, or remains fragmented. The different functionalities of the various appliances, as well as market factors, imply that the devices that control them communicate via a multitude of different protocols (KNX, LonWorks, InOne). Building a high level middleware to support all the appliances seems to be a reasonable approach. However, market factors has shown that the emergence of a unique and universal middleware is a dream. To solve this issue, we have built a new generation of schizophrenic middleware in which service access can be generated from an abstract services description. EntiMid, our implementation of schizophrenic middleware, supports various services access models (several personalities): SOAP (Simple Object Access Protocol), UPnP and DPWS (Device Profile for WebServices). These personalities are generated using a Model Driven Engineering approach and this middleware will be deployed in the context of new services definition at the Rennes city level  .
Managing Software Product Line
Product derivation, i.e. reusing core assets to build products, did not receive sufficient attention from the product-line community, yielding a frustrating situation. On the one hand, automated product derivation approaches are inflexible; they do not allow products meeting unforeseen, customer-specific, requirements. On the other hand, approaches that consider this issue do not provide adequate methodological guidelines nor automated support. We proposed an integrated product derivation approach reconciling the two views to offer both flexibility and automation. First, we perform a pre-configuration of the product by selecting desired features in a generic feature model and automatically composing their related product-line core assets. Then, we adapt the pre-configured product to its customer-specific requirements via derivation primitives combined by product engineers and controlled by constraints that flexibly set product line boundaries. Our process is supported by the Kermeta metamodeling environment  ,  ,  .