Catálogo de publicaciones - libros

This paper studies models of computation, software techniques, and analytical models for distributed timed systems. By “timed systems” we mean those where timeliness is an essential part of the behavior. By “distributed systems” we mean computational systems that are interconnected on a network. Applications of timed distributed systems include industrial automation, distributed immersive environments, advanced instrumentation systems, networked control systems, and many modern embedded software systems that integrate networking. The introduction of network time protocols such as NTP (at a coarse granularity) and IEEE 1588 (at a fine granularity) makes possible time coherence that has not traditionally been part of the computational models in networked systems. The main question we address in this paper is: Given time synchronization with some known precision, how does this change how distributed applications are designed and developed? A second question we address is: How can time synchronization help with realizing coordinated real-time events.

Complex distributed systems with control parts are difficult to develop and maintain. One reason of the complexity is the high degree of interaction and parallelism in these systems. Systematic, architecture-centric approaches are required to model, implement and verify such systems. To manage complexity, we apply a service-oriented development process, yielding manageable and flexible architecture specifications. We specify interaction patterns defining services using an extended Message Sequence Chart notation. We model a portion of the BART system as a case study, demonstrating the applicability of our methodology to this domain of complex, distributed, reactive systems. Our approach allows us to separate the problem of orchestrating the interactions between distributed components and developing the control algorithms for the various control tasks. We provide a brief overview of service-oriented development and service-oriented architectures, as well as a detailed description of our results for the BART case study.

Based on our investigations of a case study of controllers for train systems [6,7,13,14], we present a model of reactive systems which emphasizes partitioning of system behavior into and . The class of reactive systems considered are non-strict in the sense that their behavior is not entirely governed by past events; instead, future events must also be considered in the design of controllers for such systems.

The emergence of small, mobile, inexpensive computing platforms has made computation possible virtually anywhere, and has opened up countless opportunities for distributed and decentralized collaboration and information sharing among a wide range of actors. The software-intensive systems of today are increasingly shaped by their decentralized, resource-constrained, embedded, autonomic, and mobile (DREAM) computing environments. In this paper we present GridLite, a software architecture-based grid platform suitable for deployment in DREAM environments. Our prototype implementation of GridLite represents an effective and highly efficient marriage of our OODT data grid and Prism-MW architectural middleware solutions. The ultimate goal of GridLite is to extend the reach of the grid all the way to people’s “pockets”. Our initial experience suggests that this goal is achievable and worthy of further active pursuit.

This paper presents DARX, our framework for building failure- resilient applications through adaptive fault tolerance. It relies on the fact that multi-agent platforms constitute a very strong basis for decentralized software that is both flexible and scalable, and makes the assumption that the relative importance of each agent varies during the course of the computation. DARX regroups solutions which facilitate the creation of multi-agent applications in a large-scale context. Its most important feature is adaptive replication: replication strategies are applied on a per-agent basis with respect to transient environment characteristics such as the importance of the agent for the computation, the network load or the mean time between failures.

Firstly, the interwoven concerns of multi-agent systems and fault-tolerant solutions are put forward. An overview of the DARX architecture follows, as well as an evaluation of its performances. We conclude, after outlining the promising outcomes, by presenting prospective work.

The Carrier Strike Group (CSG) and the Expeditionary Strike Group (ESG) are two common types of US Naval units consisting of multiple ships traveling as a group. All vessels within the CSG/ESG transmit and receive data via satellite, even when those vessels are within radio frequency line of sight (RFLOS). Within the CSG/ESG, satellite communications (SATCOM) are clearly necessary for vessels well forward of the main body, but could be augmented by RFLOS wireless communications for some members of the CSG/ESG. The goal of this research is to identify software technology that minimizes the barriers to employing affordable, commercially available technology (i.e., 802.11x) for ship-to-ship communications at sea. Some of the existing barriers to 802.11x communications at sea result from communication protocols that do not support the varying topologies or human network intervention one would expect to encounter within the CSG/ESG. This paper advances the concept for a predictive routing protocol that proactively addresses the topological and human issues unique to the DANN. Proactive routing will re-route the transmissions prior to interruptions, thus preventing interruption of open communication sessions.

Heterogeneous non-functional requirements of Distributed Real-Time Embedded (DRE) system put a limit on middleware engineering: the middleware must reflect application requirements, with limited runtime impact. Thus, building an application-tailored middleware is both a requirement and a challenge.

In this paper, we provide an overview of our work on the construction of middleware. We focus on two complementary projects: the definition of middleware that provides strong support for both tailorability and verification of its internals; the definition of a methodology that enables the automatizing of key steps of middleware construction.

We illustrate how our current work on PolyORB, Ocarina and the use of Petri Nets allows designer to build the middleware that precisely matches its application requirements and comes with precise proof of its properties.

The prevailing paradigm in the regime of resource-constrained embedded devices is event-driven programming. It offers a lightweight yet powerful concurrency model without multiple stacks resulting in reduced memory usage compared to multi-threading. However, event-driven programs need to be implemented as explicit state machines, often with no or limited support from the development tools, resulting in ad-hoc and unstructured code that is error-prone and hard to debug. This paper presents TinyVT, an extension of the nesC language that provides a virtual threading abstraction on top of the event-driven execution model of TinyOS with minimal penalty in memory usage. TinyVT employs a simple continuation mechanism to permit blocking wait, thus allowing split-phase operations within C control structures without relying on multiple stacks. Furthermore, it provides fine-grained scoping of variables shared between event handlers resulting in safer code and allowing for optimizations in compile-time memory allocation. TinyVT source code is mapped to nesC with a source-to-source translator, using synchronous communicating state machines as an intermediate representation.

In this paper, we exemplify outdoor distributed computing and point out the key challenges. We present Split Smart Messages, a lightweight, portable, network failure resilient and relatively secure middleware that enables a large subset of outdoor distributed computing applications. We also present a Service Discovery, Interaction and Payment Protocol (SDIPP) tailored for mobile phones. We evaluate our middleware and protocol on Sony Ericsson P900 phones and present experimental results.

Current coordination models offer limited support for applications in which mobile hosts not only must coordinate their actions, but must also coordinate when those actions will be taken. This paper describes the design of TNM, a new coordination model based on — a novel extension to current coordination models through which mobile hosts can propose and negotiate which actions they will take and when. We discuss the use and advantages of this new coordination model in the context of the automatic motorway application challenge problem posed for the 2005 Monterey Workshop.