Sites Inria

Version française



Logiciel © Inria / Photo C. Morel

Developing large-scale collaboration services that guarantee the confidentiality of data and proposing secure and effective methods for service composition represents a major challenge.

Software as a service(SaaS) constitutes a major delivery paradigm shift for software publishers and an undeniable revolution for users. It frees clients from costly tasks such as infrastructure management, deployment and configuration of applications and offers them the promise of ubiquitous availability of their data and applications.  Nevertheless it raises several upscaling challenges in terms of volumetry of data and users, composition of an increasing number of services, in particular the security and confidentiality of data and applications hosted by a third party.  

To address the issue of data confidentiality, the Inria Nancy - Grand Est centre teams work on collaboration services that do not require a central authority and which upscale in terms of the number of collaborators. As a result, they have proposed optimistic replication mechanisms capable of replicating and resynchronising shared peer-to-peer data between applications that are run in simple web browsers. With regard to service composition, the researchers work on formal models and tools to orchestrate and distribute the services in an effective and reliable way, whilst also ensuring the respect of security rules during their execution.  

Its applications

  • Algorithms to maintain data coherence that do not require any coordination (and therefore a central server) and which upscale in terms of the number of collaborators.
  • Optimisation mechanism for the running of business processes in the cloud that maximise the number of clients using the same facility, whilst minimising costs and guaranteeing the specified quality of services. 
  • Validation of trust and security aspects of service-oriented architectures.
  • Delegation mechanism of security authorisations in business social networks based on the event computing model. 

Keywords: Software Licence Patent Industrial partnerships