Catálogo de publicaciones - libros

Mobile ad-hoc networks (MANETs) are dynamic computing environments where it is hard to make predictions about service provision. To ensure a level of predictability —and thus make the services more dependable—, it has been argued that the hosts must exchange information that allows guessing how the network is set up at a given moment, and how it will be in the near future. This paper introduces an approach to handling that information, which has been explicitly devised to deal with incomplete and changeable knowledge. As a contribution to the current state of the art, this approach enables a practical scheme where the different hosts in a MANET can collaborate to make up the network that best satisfies their service requirements.

We present a model for choreography à la WS–CDL and formalize it in ΔDSTL(x), a spatio–temporal logic for the specification and verification of global computing systems. The approach builds on the formalization of an atomic interaction and defines composition rules to describe complex choreographies. The logic permits to reason on the choreography formalization and to derive the properties of interest. A pleasant characteristics of the proposed approach is that the composition of formulae, corresponding to a choreography, results in a formula shaping as an atomic interaction formula. Therefore, the properties of complex choreographies can be uniformly described as interactions. We demonstrate the approach using a business scenario already tackled in the literature.

Mobile ad hoc networks (MANETs) define a challenging computing scenario where access to resources is restrained by connectivity among hosts. Replication offers an opportunity to increase data availability beyond the span of transient connections. Unfortunately, standard replication techniques for wired environments mostly target improvements to fault-tolerance and access time, and in general are not well-suited to the dynamic environment defined by MANETs. In this paper we explore replication for mobility in the context of a veneer for lime , a Linda-based middleware for MANETs. This veneer puts into the hands of the application programmer control over what to replicate as well as a set of novel replication and consistency modes meaningful in mobile ad hoc networks. The entire replication veneer is built on top of the existing lime model and implementation, confirming their versatility.

While in the sequential world the programmer can concentrate on the algorithmic solution to his given problem, in parallel and distributed systems he also has to consider aspects of communication, synchronization and data movement. In this paper we describe a prototypical middleware solution that enables the clear separation of these aspects. We combine algorithmic skeletons describing the computational aspects with overlapping data distributions describing the communication and synchronization. Both are expressed in a high-level manner. The system automatically coordinates the different activities and allows the programmer to easily change the underlying communication topology.

Coordination infrastructures can be used for the general-purpose support of WfMSs (workflow management systems). Suitably-expressive coordination artifacts can be specialised as workflow engines, encapsulating workflow rules expressed in terms of coordination laws. In this paper, we focus on the issue of inter-organisational workflow (IOW), and show how the issue of multiple, interdependent, distributed workflows requires coordination artifacts to be linkable , so as to create a network of inter-connected coordination flows. After discussing a model of workflow engine based on ReSpecT tuple centres, we introduce a distributed workflow architecture based on TuCSoN , exploiting a logic-based workflow language. In particular, we focus on the definition of a scoping mechanism, and show how this enable workflows to be dynamically governed and distributed upon a coordination infrastructure based on artifact linkability. An example of a VE (virtual enterprise) workflow is finally discussed.

This paper presents a coordination model, the Actor, Role and Coordinator (ARC) model, to address three main concerns inherent in a pervasive Open Distributed and Embedded (ODE) system: dynamicity, scalability, and stringent QoS requirements. The model treats a pervasive ODE system as a composition of concurrent computation and coerced coordination. In particular, concurrent computation is modeled as Actors, while coerced coordination specifies the system’s QoS requirements by mapping them to coordination constraints. The coordination constraints are transparently imposed on actors through message manipulations, which are carried out by the roles and coordinators. The coordinators are responsible for the coordination among roles, while the roles in our model provide abstractions for coordinated behaviors that may be shared by multiple actors and further assume local coordination responsibilities for the actors playing the roles. The role’s behavior abstraction decouples the syntactic dependencies between the coordinators and the actors, thus shielding the coordinator layer from the dynamicity of underlying actors inherent in ODE systems. This paper also formally defines the role and coordinator behaviors and the composition of the actor computation model with the proposed coerced coordination model. Our formal study has shown that the ARC system is closed under composition and recursion.

CAST is a coordination model designed to support interactions among agents executing on hosts that make up a mobile ad hoc network (MANET). From an application programmer’s point of view, CAST makes it possible for operations to be executed at arbitrary locations in space, at prescribed times which may be in the future, and on remote hosts even when no end-to-end connected route exists between the initiator and target(s) of the operation. To accomplish this, CAST assumes that each host moves in space in accordance with a motion profile which is accurate but which at any given time extends into the future for a limited duration. These motion profiles are freely exchanged among hosts in the network through a gossiping protocol. Knowledge about the motion profiles of the other hosts in the network allows for source routing of operation requests and replies over disconnected routes. In this paper, we present the CAST model and its formalization. We also discuss the feasibility of realizing this model.

Rebeca is an actor-based language which has been successfully applied to model concurrent and distributed systems. The semantics of Rebeca in labeled transition system is not compositional. In this paper, we investigate the possibility of mapping Rebeca models into a coordination language, Reo, and present a natural mapping that provides a compositional semantics of Rebeca. To this end, we consider reactive objects in Rebeca as components in Reo, and specify their behavior using constraint automata as black-box components within Reo circuits. Modeling coordination and communication among reactive objects as Reo circuits, and the behavior of reactive objects as constraint automata, provides a compositional semantics for Rebeca. Although the result is a compositional model, its visual representation in Reo shows very well that it still reflects the tight coupling inherent in the communication mechanism of object-based paradigms, whereby the real control and coordination is built into the code of the reactive objects themselves. We describe an alternative design that overcomes this deficiency. This illustrates the differences between objects and components, and the challenges in moving from object-based to component-based designs.