Catálogo de publicaciones - libros

Implementing context-dependent behaviour of pervasive computing applications puts a great burden on programmers: Devices need to continuously adapt not only to their own context, but also to the context of other devices they interact with. We present an approach that modularises behavioural adaptations into roles. Role selection takes the context of all the devices involved in an interaction into account, ensures an unambiguous scope of adaptation even in the presence of concurrency, and protects the privacy of the devices. Thus, our context-dependent role (CDR) model facilitates expressing interactions between applications in different, possibly conflicting contexts.

This paper presents a twofold context modelling approach that integrates beliefs (uncertain knowledge) and facts to reason about various everyday situations. Awareness of everyday situations enables mobile devices to adapt to the social and conceptual settings in which they operate; it also enables resources which share a similar context to cooperate in order to carry out a distributed task on behalf of their user. Our context modelling process involves the identification of the context of interest, the determination of those aspects of a context which can be captured by employing sensors, the determination of contextual states for each aspect, and finally, the determination of logical and probabilistic relationships between the contextual aspects and the context they represent. We demonstrate our approach by modelling physical places. Data from various heterogeneous sensors build our system’s belief, while containment relationships build its factual knowledge regarding places. The system utilises its belief and factual knowledge to reason about the whereabouts of a mobile user.

Context-aware applications use and manipulate context information to detect high-level , which are used to adapt application behavior. This paper discusses the specification of situations in context-aware applications and introduces a rule-based approach to detect situations. Situations are specified using a combination of UML class diagrams and OCL constraints. We support a wide range of situations, which can be composed of more elementary kinds of context. We discuss how to cope with distribution and to exploit it beneficially for context manipulation and situation detection. We employ a generic rule-based platform (DJess [2]) to support the derivation of situations in a distributed fashion.

Software components embedded in wireless devices are subject to behavior which cannot be fully and realistically predicted. This calls for a runtime management infrastructure that is able to observe and control the components’ states and to make their behaviors explicit, tangible and understandable, in any case and at any time. In this paper, we propose a framework for remotely administrating the functional behavior of software components deployed on wireless nodes. This framework is based on components which are locally managed by internal managers on the wireless side. The controllable nature of components relies on executable UML models that persist at runtime. On the administration side, models are replicated and synchronized with the models that constitute the inner workings of the wireless components.

The present emergence of loosely-coupled, inter-enterprise collaboration, i.e., virtual organizations calls for new kind of middleware: generic, common facilities for managing contract-governed collaborations and the autonomous business services between which those collaborations are formed. While further work is still needed on the functional governance of the collaborations and services, even more work is awaiting on the management of non-functional aspects of the virtual enterprises and their members. In this paper, languages and architectures for service level agreement between Web Services are discussed and the maturity of the service level management solutions is reflected against the needs of federated virtual organizations.

Constructing and executing distributed applications that can adapt to their current operating context, in order to maintain or enhance Quality of Service (QoS) attribute levels, are complex tasks. Managing multiple, interacting QoS features is particularly difficult since these features tend to be distributed across the system and tangled with other features. The crosscutting nature of QoS features can make them difficult to evolve, and it can make it complicated to dynamically optimize with respect to provided QoS during execution. Furthermore, it complicates efficient construction of application variants that differ in their QoS characteristics to suit various execution contexts. This paper presents an aspect-oriented and model driven approach for constructing and a QoS-aware middleware for execution of QoS-sensitive applications. Aspect-oriented modeling techniques are used to separate QoS features from primary application logic, and for efficient specification of alternative application variants. Model driven engineering techniques are used to derive run-time representations of application variants from platform independent models. The developed middleware chooses the best variant according to the current operating context and the available resources.

Because they are required to support component deployment and composition, modern execution environments embody a number of common features such as dynamic linking and support for multiple component versions. These features help to overcome some classical maintenance problems focused largely on component evolution, where successive generations of collaborating components need to be kept collaborating. What has been less studied has been component adaptation, whereby a component developed in an environment consisting of one set of service components is required to operate in one or several other environments containing qualitatively different components. In this paper we examine the needs developers and deployers have arising out of component adaptation and explore the concept of Flexible Dynamic Linking as a means of satisfying them. We describe a suite of tools developed to demonstrate this approach to component adaptation support within the .NET Common Language Runtime.

Mobile computing is characterised by variations in user needs and in the computing and communication resources. We have developed a middleware centric approach for the development of software capable of dynamically adapting to such variations. The middleware leverages models of needs and resources and the adaptation capabilities of the software and performs context monitoring, adaptation planning and dynamic reconfiguration at runtime. In this paper we focus on the modelling of resources of a distributed mobile computing infrastructure and how the resource model is used in adaptation planning. We present a distributed resource management framework and mechanisms necessary to maintain an up to date resource model at runtime. The challenge is to balance the level of abstraction so as to hide some of the heterogeneity of the actual infrastructure while retaining sufficient detail to serve the needs of distributed and centralized adaptation planning. The proposed framework is illustrated through a running example.

Contemporary object-oriented programming seeks to enable distributed computing by accessing remote objects using blocking remote procedure calls. This technique, however, suffers from several drawbacks because it relies on the assumption of stable network connections and synchronous method invocations. In this paper we present an approach to support distributed programming, which rely on local object replicas keeping themselves synchronized using an underlying peer-to-peer infrastructure. We have termed our approach (PDSO). This PDSO approach has been implemented in the DOLCLAN framework. An evaluation demonstrates that DOLCLAN can be utilized to create a real distributed collaborative system for ad-hoc collaboration in hospitals, which demonstrates that the approach can support the creation of non-trivial distributed applications for pervasive computing.

Web service QoS (Quality of Service) is a key factor for users to evaluate and select services. Traditionally, run-time QoS of web services stores in centralized QoS registry, which may have performance and availability problems. In this paper, we propose a P2P (Peer-to-Peer) QoS registry architecture for web services, named Q-Peer. Q-Peer is an unstructured P2P system. Query of QoS is naturally achieved by getting QoS address from corresponding service description. Q-Peer has a replication based mechanism to ensure load-balance of the whole architecture. The architecture takes advantage of P2P systems to ensure its availability, performance and autonomy. We are currently implementing Q-Peer and planning to test it on Planet-Lab.