Catálogo de publicaciones - libros

Sensor network technology is pushing towards integration into the business world. By using sensor node hardware to augment real life business items it is possible to capture the world and support processes where they actually happen. Many problems of the business logic running our world can be efficiently implemented “on the item”. In order for these smart items to couple back to the virtualized world of business processes it necessary to design a uniform system abstraction for enterprise systems. Service oriented architectures are the tool to describe functionality apart from its concrete implementation. This paper describes a system and the experiences made integrating wirelessly networked smart business items into high-level business processes.

This paper describes a dynamic, scalable and flexible Policy-based Management Architecture (PbMA), which is characterized by a reliable and autonomous deployment, activation and management of Grid Services. This architecture follows the implied conditions by the Open Grid Services Architecture (OGSA) standard. Although applicable to any user profiles, our system is essentially intended for non-massive resource owners accessing large amounts of computing, software, memory and storage resources. Unlike similar architectures, it is able to manage service requirements demanded by users, providers and services themselves. This architecture is also able to manage computational resources in order to fulfill Quality of Service (QoS) requirements, based on a balanced scheduling of resources exploitation. Our approach is scalable and flexible by extending itself the management components and policies interpreters needed to control multiple infrastructures regardless network technology, operative platform or administrative domain. The management architecture shows its reliability through a Grid Service deployment example.

The initiatives and introduced challenging visions for future computer systems. They address the growing complexity of these systems that demands for new ways to control them. Future systems should be able to adapt dynamically to the current conditions of their environment. They should be characterised by so-called self-x properties like self-configuring, self-healing, self-optimising, self-protecting, and context-aware. For the incorporation of self-healing capabilities into distributed systems the detection of failures is a crucial part. Recently we proposed a new failure detector that can be described as an adaptive accrual algorithm. It has been designed for flexible generic usability as a basis to realise self-healing of distributed systems. This paper introduces variations of the proposed basic algorithm to improve its performance and provides an evaluation of all algorithms using message delay and loss models of the internet.

This paper presents a framework for deploying software components over a distributed system by using the notion of dynamics between components. It enables an application to be composed of one or more mobile components that can be deployed to different computers when the application is being executed. The key idea behind the framework is to provide components with deployment policies corresponding to gravitational and repulsive forces. The polices control the relocation relation between two components. As a result, a federation of distributed components can be moved and changed over a distributed system in a self-organizing manner. This paper also presents a prototype implementation of the approach and its applications.

Evolutionary hardware design reveals the potential to provide autonomous systems with self-adaptation properties. We first outline an architectural concept for an intrinsically evolvable embedded system that adapts to slow changes in the environment by simulated evolution, and to rapid changes in available resources by switching to preevolved alternative circuits. In the main part of the paper, we treat evolutionary circuit design as a multi-objective optimization problem and compare two multi-objective optimizers with a reference genetic algorithm. In our experiments, the best results were achieved with TSPEA2, an optimizer that prefers a single objective while trying to maintain diversity.

Today’s technical systems are becoming increasingly complex. Future systems will consist of a multitude of complex soft- and hardware components, which interact with each other to satisfy global system functional requirements. This trend bears the risk of more and more breakdowns and other unexpected behaviour. Organic Computing (OC) has the vision of addressing the challenges of complex distributed systems by making them more life-like (organic), i. e. endowing them with abilities such as self-organisation, self-configuration, self-repair, or adaptation. This can only be achieved by giving the system elements adequate degrees of freedom. This may result in an emergent behaviour, which can be positive as well as negative. Therefore, we need an observer/ controller architecture, which allows for self-organisation but at the same time enables adequate reactions to control the - sometimes completely unexpected - emerging global behaviour.

In this paper, we give an introduction to a generic observer/controller architecture, adapt this framework to a scenario of a self-organising robot swarm, and show how to control and prevent global, collective, unwanted behaviour based on observations of the local behaviour of the distributed agents.

Adaptive clustering aims at improving cluster utilization for varying load and traffic patterns. Locality-based least-connection with replication (LBLCR) scheduling that comes with Linux is designed to help improve cluster utilization through adaptive clustering. A key issue with LBLCR, however, is that cluster performance depends much on a single threshold value that is used to determine adaptation. Once set, the threshold remains fixed regardless of the load and traffic patterns. If a cluster of PCs is to adapt to different traffic patterns for high utilization, a good threshold has to be selected and used dynamically. We present in this report an adaptive clustering framework that autonomously learns and adapts to different load and traffic patterns at runtime with no administrator intervention. Cluster is configured once and for all. As the patterns change the cluster automatically expands/contracts to meet the changing demands. At the same time, the patterns are proactively learned that when similar patterns emerge in the future, the cluster knows what to do to improve utilization. We have implemented this autonomous learning method and compared with LBLCR using published Web traces. Experimental results indicate that our autonomous learning method shows high performance scalability and adaptability for different patterns. On the other hand LBLCR-based clustering suffers from performance scalability and adaptability for different traffic patterns since it is not designed to obtain good threshold values and use them at runtime.

This paper presents a new approach to the problem caused by the exploding needs of computing resources in function calculation. The proposal argues for increasing the computing power at the primitive processing level in order to reduce the number of computing levels required to carry out the calculations. This trade-off is developed within the limits of function evaluation by substituting the usual primitives, namely sum and multiplication, by a unique weighted primitive that can be tuned for different values of the weighting parameters. All function points are carried out by successive iterations of the primitive. A parametric architecture implements the design. The case of combined trigonometric functions involved in the calculation of the Hough transform (HT) is analyzed under this scope. It provides memory and hardware resource saving as well as speed improvements, according to the experiments carried out with the HT.

This paper presents a new environment for exploring and optimizing VLIW architectures for multimedia applications. The environment consists of a generic VLIW architecture, in which virtually all characteristics can be changed, and an assembler with the corresponding parameterized scheduler based on an enhanced version of the list scheduling algorithm. A novel partitioned register file architecture is proposed and analyzed with this environment. This is performed using a highly time consuming task of the H.264 video decoder application. Performance improvements of up to 67% can be achieved when running this application on different architecture configurations.

In this paper, we present a new concept for modeling of interconnection networks in the field of massively parallel processor embedded architectures. The main focus of the paper is on two interconnection concepts, namely, interconnect-wrapper and DyRIBox definitions of reconfigurable interconnection networks. We compare both interconnection concepts against each other and formally prove their equality. Both concepts allow to model many different reconfigurable inter-processor networks efficiently. Furthermore, we point out how to define the interconnect using an architecture description language for massively parallel processor architectures called MAML. Finally, we demonstrate the pertinence of our approach by modeling and evaluation of different reconfigurable interconnect topologies.