Recent Results

A Model-Driven Approach to Develop Adaptive Firmwares — by Mark Hefke — last modified 2011-05-02 19:53

 

D6.3 Case Study Implementation and Validation — by Mark Hefke — last modified 2011-04-29 14:53

Thales and CAS provide two case studies for evaluating and validating the DiVA technology. The Thales case study is called "Crisis Management System for a civil airport" and, beside the base functionality of DiVA technology, mainly focuses on the support of distributed systems. The CAS case study is called "CRM: Service mash-ups in a hosted SaaS CRM application" and, beside the base functionality of DiVA technology, mainly focuses on the support of sessions. In deliverable D6.2, a first description of the case study implementation and evaluation in phase 1 of the DiVA project was given. This deliverable D6.3 has two purposes: The first one is to provide an updated description of the case study implementation and evaluation of the DiVA studio according to what has been extended in phase 2 of the DiVA project. The second one is to provide a detailed validation plan and validation results throughout the case studies. Open Document

D5.2 DiVA studio final version — by Mark Hefke — last modified 2011-04-29 15:13

This document describes the DiVA Studio tool deliverable which is downloadable from http://developer.berlios.de/projects/diva-unix/ . The DiVA Studio is a set of tools and case studies that realize the technology developed to support the DiVA methodology to design and build adaptive systems and integrates cutting edge tools that enable researchers and engineers to plan, design, and implement adaptive systems. The purpose of this studio is to coordinate and produce an integrated framework for adaptive system engineering, covering the different lifecycle phases of an adaptive system (from requirement to runtime). The main part of this deliverable is the Studio itself, which comprise an integrated toolset, execution platform, tutorials and methodology for construction and execution of adaptive systems. Each of the tools, the runtime platform and the methodology are described in detail in separate deliverables. Thus, this document only provides a brief overview of the studio and its components and explains how to install and get started with the studio. The document is structured as follows: After a brief introduction in Section 1, Section 2 describes how to install and get started with the software and the samples. The set of components which comprise the studio are briefly presented in Section 3. Open Document

D4.3 - Adaptation model and validation framework — by Mark Hefke — last modified 2011-04-29 14:39

This deliverable describes the Adaptation Reasoning and Validation Framework developed as part of the DiVA tools. The purpose of the framework is to provide services for efficient reasoning and validation throughout the specification, design and deployment of dynamic systems following the DiVA methodology and using the DiVA tools. By providing experimental results the documents helps users of the DiVA tool chain to understand and differentiate the capabilities of the available reasoners and reasoning approaches. The document shows also how a combination of random selection and standard reasoning approaches yields a scalable reasoning. It also illustrates the purpose of validation in the context of DiVA based development. The document is structured as follows: A brief explanation of the underlying DiVA Adaptation Model in chapter 2 is followed by a conceptual introduction in the framework in chapter 3. The following chapter describes the available reasoners and shows the results of experiments conducted with them. The validation concepts are explained in chapter 5. Open Document

D7.4 DiVA Whitepaper - A Model-based Approach for Construction and Run-time Management of Adaptive Systems: DiVA practices and Lessons Learned — by Mark Hefke — last modified 2011-04-29 13:38

Many of today’s software systems need to be available 24/7, be highly interactive, and continuously adapt according to varying environment conditions, user characteristics, and characteristics of other systems that interact with them. Such systems, called adaptive systems, are expected to be long-lived and able to undertake adaptations with little or no human intervention. Construction and run time management of such adaptive systems are complex tasks. Non-trivial challenges include provisioning of efficient design time and run time representations, system validation to ensure safe adaptation of interdependent components, and scalable solutions to cope with the possible combinatorial explosions of adaptive system artefacts such as configurations, variant dependencies and adaptation rules. Furthermore, in current approaches the adaptation logic is typically specified at the code level, tightly coupled with the main system functionality, making it hard to control and maintain. The DiVA approach provides a new tool-supported methodology with an integrated framework for specifying and managing dynamic variability in adaptive systems. The approach combines aspect-oriented and model-driven techniques in an innovative way to handle adaptive system complexities. The proposed approach has been implemented and validated through case studies. Open Document

D7.5 DiVA Roadmap — by Mark Hefke — last modified 2011-04-29 13:34

Deliverable D7.5 explores and encourages future research issues beyond achieved technical and methodological results within the DiVA project, which nevertheless meet demands of future emerging markets. In this context, two different perspectives have been taken into account, the research perspective and the market-oriented perspective. Regarding the research perspective, we identified future research demands based on Technical Results and Methodologies achieved in the DiVA project, which in turn are based on the two scenarios for possible future applications concerning Dynamic Variability in complex adaptive systems (Crisis Management and Next Generation CRM Solutions, cf. Deliverable 6.1) as well as on SOTA Research background. With respect to the market oriented-perspective, we identified and categorized future research challenges based on (future) market-specific requirements and research challenges. Based on both perspectives, two technology roadmaps have been worked out, providing an expert-based consensus view of the future science and technology landscape that at the same time considers future market needs. Open Document

Inferring Test Results for Dynamic Software Product Lines — by Phil Greenwood — last modified 2011-03-08 18:32

 

Investigating Testing Approaches for Dynamically Adaptive Systems — by Phil Greenwood — last modified 2011-03-08 18:29

 

Tracing Requirements for Adaptive Systems using Claims — by Phil Greenwood — last modified 2011-03-08 18:27

 

4th International Workshop on Models@run.time — by Nelly Bencomo — last modified 2010-09-13 20:14

 

How dynamic is your Dynamic Software Product Line? — by Nelly Bencomo — last modified 2010-09-13 20:08

Recently, there have been increasing demands for the postponement of decisions on software adaptations and product variations to provide the flexibility required by dynamic environments and users. The goal is that software adaptations and product variations can be chosen even at runtime. As such, a research theme that addresses development issues for reusable and dynamically reconfigurable core assets has emerged and it is called dynamic software product lines (DSPLs) with its consequential need to manage runtime variability. Research on the use of runtime variability, however, is still heavily based on the specification of decisions during design time. That is, a system simply postpones “when to adapt” to runtime but “how to adapt” is already decided at design time. In this paper, we present a brief assessment of the current research in the area and discuss some research issues related to the feasibility of DSPL oriented approaches to build self-adaptive systems.

Requirements Reflection: Requirements as Runtime Entities — by Nelly Bencomo — last modified 2010-09-13 19:57

 

Modelling Service Requirements Variability: The DiVA Way — by Phil Greenwood — last modified 2010-09-09 15:35

This chapter tackles the challenges of variability identification, modelling and implementation for service-based systems. The DiVA methodology is applied to the Mobile Phone Service Portability case-study to demonstrate its solutions to these challenges. The DiVA methodology utilises concepts of Aspect-Oriented Software Development to encapsulate service variants in distinct modules and uses Model- Driven Development techniques to analyse and transform conceptual designs into executable services. The DiVA approach provides a tool-supported methodology for managing dynamic variability in adaptive systems and taming system complexity.

D1.4 Framework for evaluation of configuration alternatives and trade-offs — by Phil Greenwood — last modified 2010-09-13 13:37

In the previous DiVA deliverables D1.2 and D1.3 we have presented the DiVA RE framework for identification and modelling of dynamic variability in user requirements and composition of system of systems using these models. In the present deliverable we continue to the refine the DiVA RE framework by: a) developing the previously outlined traceability support into a Framework for Evaluation of Configuration Alternatives and Trade-offs using the simulation results form the design/run time (as defined in the DiVA workplan); b) improving the DiVA RE framework constituents to address the shortcomings revealed as a result of their evaluation. The Evaluation Framework is based on propagating design refinements and model simulation results back to the requirements level to inform and support a scoped configuration evaluation and trade-off analysis from the stakeholder’s perspective. While DiVA RE framework evaluation provides useful input for its evolution. FMP Models: http://www.ict-diva.eu/DiVA/developer-zone/downloads/FMP.zip

D3.3 Reference architecture - final version — by Brice Morin — last modified 2011-04-29 15:02

This deliverable presents the final version of the reference architecture to support dynamic variability using model-driven engineering techniques and aspect models. This reference architecture leverages the design-models of WP2, as well as the reasoning techniques of WP4, at runtime. The purpose of this document is to provide an overview of the reference architecture, to detail the important parts of this reference architecture and give some implementation details. This document is associated with a software system, demonstrating the reference architecture, which is integrated into DiVA Studio (WP5). This document (D3.3) is an extension of D3.2. The main changes are: • The reference architecture is now based on OSGi, the former one was based on Fractal. We thus give some implementation details for OSGi (Section 6).. • Integration of the new version of SmartAdapters (weaver). In this new version, aspect models are now compiled into Java code, which makes it possible to weave aspect model at runtime, in an efficient way. We detail the compilation process of SmartAdapters as well as the new features of this weaver (Section 5). • The reference architecture now provides support for distribution and describes the architectural extensions to support distribution as well as co-existence and co-dependency. Furthermore it provides a distributed consistency framework to support distributed aspect configuration and reconfiguration (Section 8). Open Document

D7.3b Plan for using and disseminating the knowledge — by Dirk Balfanz — last modified 2010-09-09 16:42

This deliverable replaces Deliverable 7.3a by providing an overview of the current dissemination and exploitation status as well as of future plans of the DiVA consortium. Exploitation / use of knowledge: D7.3b describes the general exploitation approach in DiVA and roughly features the main exploitable assets, being current or prospective results of the DiVA research and development work. Addressing the different exploitable assets, the exploitation plans of all partners are given, i.e. plans for using the DiVA knowledge later on, commercially or in other ways. R&D work has already achieved a number of innovative results (cf. D5.1). Based on that, concrete exploitation steps have already been defined, extended and concretized. Dissemination of knowledge: D7.3b wraps up all dissemination activities (conference participation, workshops, participation in trade fairs etc.) of preceding and current period, and lists as well related publications (scientific, marketing or other).

D2.3 - DiVA methodology — by Vegard Dehlen — last modified 2010-09-06 17:57

This deliverable explains a model-driven, iterative and test-driven methodology for developing adaptive systems using the DiVA tools and technologies. It covers the lifecycle from requirements analysis until deployment and consists of four main phases: 1. Requirements engineering 2. Modelling of the adaptation concerns 3. Modelling of runtime architecture 4. Transformation and deployment All of the phases and involved tools connect to form a coherent process, often involving semi- or fully automatic tool support where possible in the spirit of model-driven engineering. The DiVA methodology is concerned with developing the adaptive features of a system and is intended to be combined with existing tried-and-tested methodologies for development of the functional aspects of the software.

D2.2 - Transformation framework, final version — by Vegard Dehlen — last modified 2010-09-06 17:54

Constructing and executing distributed systems that can automatically adapt to the dynamic changes of the environment are highly complex tasks. Non-trivial challenges include provisioning of efficient design time and run time representations, system validation to ensure safe adaptation of interdependent components, and handling of possible combinatorial explosions of adaptive system artefacts such as configurations, variant dependencies and adaptation rules. These are all challenges where current approaches offer only partial solutions. Furthermore, existing technologies are typically only provided at the implementation level which makes them complex to use. This deliverable describes the final version of the model transformation framework in Work Package 2. As stated in the DoW the objective of WP2 is to develop a domain-specific language for adaptive system specifications, and a composition and transformation framework. This deliverable is an updated version of deliverable 2.1 which reported on the status of the work half-way through the project. The main revisions and extensions to D2.1 are: • Adaptation DSL: updated to reflect the latest developments. • Aspect framework: new section on the tools developed for architecture and aspect modelling. • Case Studies: Updated CAS case study, new section on the Thales case study and 3 new case studies. The reader of this deliverable will find all the most up-to-date information on the work carried on the WP2 transformation framework in this deliverable. This deliverable consists of two parts. The main part is the tools and meta-models developed in the context of WP2. The second part is this document.

Flexible Model Element Introduction Policies for Aspect-Oriented Modeling — by Brice Morin — last modified 2010-07-09 15:28

Aspect-Oriented Modeling techniques make it possible to use model transformation to achieve advanced separation of concerns within models. Applying aspects that introduce model elements into a base model in the context of large, potentially composite models is nevertheless tricky: when a pointcut model matches several join points within the base model, it is not clear whether the introduced element should be instantiated once for each match, once within each composite, once for the whole model, or based on a more elaborate criteria. This paper argues that in order to enable a modeler to write semantically correct aspects for large, composite models, an aspect weaver must support a flexible instantiation policy for model element introduction. Example models highlighting the need for such a mechanism are shown, and details of how such policies can be implemented are presented.

Security-Driven Model-Based Dynamic Adaptation — by Vegard Dehlen — last modified 2010-07-09 15:25

Security is a key-challenge for software engineering, especially when considering access control and software evolutions. No satisfying solution exists for maintaining the alignment of access control policies with the business logic. Current implementations of access control rely on the separation between the policy and the application code. In practice, this separation is not so strict and some rules are hard-coded within the application, making the evolution of the policy difficult. We propose a new methodology for implementing security-driven applications. From a policy defined by a security expert, we generate an architectural model, reflecting the access control policy. We leverage the advances in the models@runtime domain to keep this model synchronized with the running system. When the policy is updated, the architectural model is updated, which in turn reconfigures the running system. As a proof of concept, we apply the approach to the development of a library management system.