Catálogo de publicaciones - libros
Euro-Par 2007 Parallel Processing: 13th International Euro-Par Conference, Rennes ,France , August 28-31, 2007. Proceedings
Anne-Marie Kermarrec ; Luc Bougé ; Thierry Priol (eds.)
En conferencia: 13º European Conference on Parallel Processing (Euro-Par) . Rennes, France . August 28, 2007 - August 31, 2007
Resumen/Descripción – provisto por la editorial
No disponible.
Palabras clave – provistas por la editorial
Computer System Implementation; Computer Systems Organization and Communication Networks; Software Engineering/Programming and Operating Systems; Theory of Computation; Numeric Computing; Database Management
Disponibilidad
Institución detectada | Año de publicación | Navegá | Descargá | Solicitá |
---|---|---|---|---|
No detectada | 2007 | SpringerLink |
Información
Tipo de recurso:
libros
ISBN impreso
978-3-540-74465-8
ISBN electrónico
978-3-540-74466-5
Editor responsable
Springer Nature
País de edición
Reino Unido
Fecha de publicación
2007
Información sobre derechos de publicación
© Springer-Verlag Berlin Heidelberg 2007
Cobertura temática
Tabla de contenidos
Fast and Efficient Total Exchange on Two Clusters
Emmanuel Jeannot; Luiz Angelo Steffenel
Total Exchange is one of the most important collective communication patterns for scientific applications. In this paper we propose an algorithm called for the total exchange redistribution problem between two clusters. In our approach we perform communications in two different phases, aiming to minimize the number of communication steps through the wide-area network. Therefore, we are able to reduce the number of messages exchanged through the backbone to only 2× max ( ,) against 2× × messages with the traditional strategy (where and are the number of nodes of each clusters). Experimental results show that we reach over than 50% of performance improvement comparing to the traditional strategies.
- Topic 13: High-Performance Networks | Pp. 848-857
Performance Analysis of an Optical Circuit Switched Network for Peta-Scale Systems
Kevin J. Barker; Darren J. Kerbyson
Optical Circuit Switching (OCS) is a promising technology for future large-scale high performance computing networks. It currently widely used in telecommunication networks and offers all-optical data paths between nodes in a system. Traffic passing through these paths is subject only to the propagation delay through optical fibers and optical/electrical conversions on the sending and receiving ends. High communication bandwidths within these paths are possible when using multiple wavelengths multiplexed over the same fiber. The set-up time of an OCS circuit is non-negligible but can be amortized over the lifetime of communications between nodes or by the use of multi-hop routing mechanisms. In this work, we compare the expected performance of an OCS network to more traditional networks including meshes and fat-trees. The comparison considers several current large-scale applications. We show that the performance of an OCS network is comparable to the best of the network types examined.
- Topic 13: High-Performance Networks | Pp. 858-867
Network Fault Tolerance in Open MPI
Galen M. Shipman; Richard L. Graham; George Bosilca
High Performance Computing (HPC) systems are rapidly growing in size and complexity. As a result, transient and persistent network failures can occur on the time scale of application run times, reducing the productive utilization of these systems. The ubiquitous network protocol used to deal with such failures is TCP/IP, however, available implementations of this protocol provide unacceptable performance for HPC system users, and do not provide the high bandwidth, low latency communications of modern interconnects. This paper describes methods used to provide protection against several network errors such as dropped packets, corrupt packets, and loss of network interfaces while maintaining high-performance communications. Micro-benchmark experiments using vendor supplied TCP/IP and O/S bypass low-level communications stacks over InfiniBand and Myrinet are used to demonstrate the high-performance characteristics of our protocol. The NAS Parallel Benchmarks are used to demonstrate the scalability and the minimal performance impact of this protocol. Communication level micro-benchmarks show that providing higher data reliability decreases bandwidth by up to 30% relative to unprotected communications, but provides performance improvements of a factor of four over TCP/IP running over InfiniBand DDR. In addition, application level benchmarks (communication/computation) show virtually no impact of the data reliability protocol on overall run-time.
- Topic 13: High-Performance Networks | Pp. 868-878
Topic 14 Mobile and Ubiquitous Computing
Nuno Preguiça; Éric Fleury; Holger Karl; Gerd Kortuem
Mobile computing has evolved tremendously in the past few years with advances in wireless networks, mobile computing, sensor networks along with the rapid growth of small, portable and powerful computing devices. These advances offer opportunities for the development of new mobile/ubiquitous computing applications and services. Topic 14 of Europar 2007 covers all aspects related with the creation of such systems.
- Topic 14: Mobile and Ubiquitous Computing | Pp. 879-880
Securing Sensor Reports in Wireless Sensor Networks*
Al-Sakib Khan Pathan; Choong Seon Hong
The sensor reports from a wireless sensor network are often used extensively in the decision making process in many systems and applications. Hence, classifying real and false sensor reports is necessary to avoid unwanted results. In this paper, we propose a scheme for securing the sensor reports in a wireless sensor network. We use one-way hash chain and pre-stored shared secret keys to provide data transmission security for the reports that travel from any source node to the base station. To introduce data freshness, our scheme includes an optional key refreshment mechanism which could be applied depending upon the requirement or the application at hand. We present an analysis along with the detailed description of our scheme.
- Topic 14: Mobile and Ubiquitous Computing | Pp. 881-890
An Algorithm for Dissemination and Retrieval of Information in Wireless Ad Hoc Networks
Hugo Miranda; Simone Leggio; Luís Rodrigues; Kimmo Raatikainen
Replication of data items among different nodes of a wireless infrastructureless network may be an efficient technique to increase data availability and improve data access latency. This paper proposes a novel algorithm to distribute data items among nodes in these networks. The goal of the algorithm is to deploy the replicas of the data items in such a way that they are sufficiently distant from each other to prevent excessive redundancy but, simultaneously, they remain close enough to each participant, such that data retrieval can be achieved using a small number of messages. The paper describes the algorithm and provides its performance evaluation for several different network configurations.
- Topic 14: Mobile and Ubiquitous Computing | Pp. 891-900
Surrendering Autonomy: Can Cooperative Mobility Help?
Ghassen Ben Brahim; Bilal Khan; Ala Al-Fuqaha; Mohsen Guizani; Dionysios Kountanis
In this paper, we develop a Cooperative Mobility Model that captures new salient features of collaborative and mission-oriented MANETs. In particular, the cost-benefit framework of our model is a significant advance in modelling heterogenous networks whose nodes exhibit the complete range of autonomy with respect to mobility. We then describe the design of CoopSim, a platform for conducting simulation experiments to evaluate the impact of parameter, policy and algorithm choices on any system based on the proposed Cooperative Mobility Model. We present a small but illustrative case study and use the experimental evidence derived from it to give an initial evaluation of the merits of the proposed model and the efficacy of the CoopSim software. In our case study, we propose studying the impact of the proposed model on improving the end-to-end communication based on the QoS parameter, namely BER.
- Topic 14: Mobile and Ubiquitous Computing | Pp. 901-910
A Context-Dependent XML Compression Approach to Enable Business Applications on Mobile Devices
Yuri Natchetoi; Huaigu Wu; Gilbert Babin
As the number of mobile device users increases, the need for mobile business applications development increases as well. However, such development is impeded by the limited resources available on typical mobile phones. This paper presents a context-dependent XML compression approach that enables the deployment of business applications on mobile devices. That is, the compressed XML document is not self-contained and cannot be de-compressed without using information shared between the sender and the recipient. By relying on shared information, we obtain a better compression ratio than existing context-free compression algorithms.
- Topic 14: Mobile and Ubiquitous Computing | Pp. 911-920
A Distributed, Leaderless Algorithm for Logical Location Discovery in Specknets
Ryan McNally; Damal K. Arvind
A speck is intended to be a miniature (5x5x5mm) device that combines sensing, processing, wireless communication and energy storage capabilities [1]. A specknet is an ad-hoc mobile wireless network of specks. The logical location of specks in the network is useful, for reasons ranging from routing data to giving the data sensed a spatial context. This paper presents a novel algorithm for discovering the logical location of specks and updating that information in the face of movement, without recourse to infrastructure support. The proposed algorithm exploits the location constraints implied by the neighbourhood links in order to compute a likely location: one hop neighbours must lie within radio range, two-hop neighbours probably lie outwith radio range. An iterative approach is used to converge on a location estimate that satisfies all constraints. The performance of the location discovery algorithm is evaluated in the SpeckSim simulator for a number of metrics, including location error. The results demonstrate that the quality of the computed locations is within 90% of optimal when used in routing calculations.
- Topic 14: Mobile and Ubiquitous Computing | Pp. 921-930
Analysis of a Kalman Approach for a Pedestrian Positioning System in Indoor Environments
Edith Pulido Herrera; Ricardo Quirós; Hannes Kaufmann
In this work we present the design principles of a wearable positioning system for users in unprepared indoor environments. We describe the most suitable technology for our application and we model the dynamics of a walking user. The system uses inertial sensors and a location system based on ultrawideband (UWB). Data fusion is carried out with a Kalman filter. The user position is estimated from data provided by the UWB location system. To update the position and direction of the user we use a dead reckoning algorithm. The use of redundant sensors and the data fusion technique minimises the presence of shadow zones in the environment. We show the advantages of combining different sensors systems.
- Topic 14: Mobile and Ubiquitous Computing | Pp. 931-940