SESA

  • Currently, computer science is in the next period of abstraction.
  • A generation ago we learnt to abstract from hardware and currently we learn to abstract from software in terms of SERVICE oriented architectures (SOA).
  • It is the service that counts for a customer and not the specific software or hardware that is used to implement the service.
  • In a later stage, we may even talk in terms of problem-oriented architectures (or more positively expressed in terms of problem solving oriented architectures) because SOAs are biased towards the service provider and not towards the customer that has a problem that needs to be solved.
  • Service-oriented architectures will become quickly the leading software paradigm.
  • However, SOAs will not scale without signification mechanization of – service discovery, service adaptation, negotiation, service composition, service invocation, and service monitoring; and data and process mediation.
  • Therefore, machine processable semantics needs to be added to bring SOAs to their full potential.
  • Development of open standards (languages) and open source architectures and tools that add semantics to service descriptions

Three Layers of SESA

  • Problem-Solving Layer
    • Turns a service-oriented architecture into a domain specific problem-solving environment
  • Common Services Layer
    • The execution environment and the supporting infrastructure that maps the problem descriptions generated at the Problem Solving Layer to the services that can solve the problems
  • Resource Layer
    • Covers the deployment and provisioning of physical resource being responsible for the actual execution of the applications 

Problem solving layer

  • This layer turns a service-oriented architecture into a domain specific problem-solving environment. 
  • It represents the transparent interface to the user(s), where all computing resources are turned into or expressed as services
  • Supports the full set of operations from an e-commerce framework: information negotiation, etc.
  • Provides clear separation between business/process logic on one hand and the common service layer

Common Services Layer

 
  • Provides an adaptive execution environment and the supporting infrastructure that maps the problem descriptions generated at the Problem Solving Layer to the services that can solve the problems. 
  • Existing architectures and standards from Web service and Grid areas (e.g. OGSA, WSRF, WSDL) which operate only at a syntactic level are semantically enriched and integrated into this layer.
  • Semantically enrichment of SOAs that implement the Common Service Layer capabilities will help to automate: service discovery, service adaptation, negotiation, service composition, etc.
  • This layer could be implemented using the W technology

Resource Layer

  • Responsible for actual execution of applications. 
  • All tasks that involve resources such as resource discovery, selection and negotiation for advanced or on-the-fly reservation of resources are supported and implemented in this layer. 
  • Covers the deployment and provisioning of physical resource (e.g. computers, data servers, and networks, usually connected into a Grid) and logical resources (e.g. application components or common services). 
  • This layer may relay on two prominent and widely discussed areas that deal with distributed resources in the context of service oriented computing are Ubiquitous Computing and Grid Computing