The proceedings contain 50 papers. The special focus in this conference is on Design Environments and Tools. The topics include: A system for designing parallel processor arrays;on the role of architectures in systems...
ISBN:
(纸本)3540638113
The proceedings contain 50 papers. The special focus in this conference is on Design Environments and Tools. The topics include: A system for designing parallel processor arrays;on the role of architectures in systems engineering;an approach to safety critical systems design;extensions and semantics of controller state diagrams for practical applications;design and implementation of an energy management system for an integrated iron and steel plant;representation and management of memory and decision in evolving software systems;a generic engineering framework for mechanical engineering based upon meta models;formal representation and proof of the interpreted sequential machine model;modelling of concurrent and embeddedsystems;a dense-time temporal logic with nice compositionality properties;introducing planning in discrete event systems;multirate control of a class of nonlinear systems;global structural approximate controllability of polynomial nonlinear systems;vectorized simulations of normal processes for first crossing-time problems;an architecture for complex systems;design of nonlinear lag-lead compensators by extended linearization using symbolic computation techniques;a numerical computation for declarative expressions;an automatic calculator with penrose diagrams;cognitive modeling for navigation of mobile robots using the sensory gradient concept;fuzzy traffic police for autonomous vehicles;object-oriented model of an open communication architecture for flexible manufacturing control;a formal model for integrated system development;moving towards an agent-based engineering environment;experiencing modeling and development of an intelligent autonomous robot;modular control system for robotized cells and lines and latest developments in robotics.
The view of processing as an integral part of function suggests that information systems must be embedded into the structure and dynamics of the functions they support. This requires that the architecture of the compu...
详细信息
The view of processing as an integral part of function suggests that information systems must be embedded into the structure and dynamics of the functions they support. This requires that the architecture of the computer-based information system be compatible with the structure and dynamics of the organization it supports. In this paper we propose a model-based approach to the development of information systems that uses models of the structure and dynamics of organizations in order to develop effective function support. The proposed approach uses a conceptual tool to model key features of the structure and dynamics of the organization and a methodology that maps such features into the architecture of its supporting information system. A modeling and simulation environment is suggested that includes an extensive model base that can be used for the simulation and analysis of organizational function in addition to its use in systems development.
This paper presents algorithms for the mean response time of generalized stochastic Petri nets (GSPN) models based on a methodology for performance modeling and prediction of real time distributed applications running...
详细信息
This paper presents algorithms for the mean response time of generalized stochastic Petri nets (GSPN) models based on a methodology for performance modeling and prediction of real time distributed applications running in parallel architectures. For architectures with monoprogrammable processors, the mean response time is obtained by successive aggregation without the use of state space, which prevents state explosion. For multiprogrammable processor architectures, reductions in immediate transitions are applied. The methodology described herein aims to help the designer to predict application execution times early in their developmental phase.
The proceedings contain 37 papers. The special focus in this conference is on Evolutionary Methods for modeling, Training and Alternative Frameworks for the Computational Study of Evolutionary Social systems. The topi...
ISBN:
(纸本)9783540627883
The proceedings contain 37 papers. The special focus in this conference is on Evolutionary Methods for modeling, Training and Alternative Frameworks for the Computational Study of Evolutionary Social systems. The topics include: Complexity formalisms, order and disorder in the structure of art;the application of evolutionary computation to selected problems in molecular biology;parallel evolutionary programming for constructing artificial neural networks;scaling behavior of the evolution strategy when evolving neuronal control architectures for autonomous agents;an object oriented simulation platform applied to markets and organizations;an agent-based computational model for the evolution of trade networks;performance-enhanced genetic programming;comparing subtree crossover with macromutation;composing 16th-century counterpoint with genetic programming and symbiosis;design of a high-gain operational amplifier and other circuits by means of genetic programming;modeling speculators with genetic programming;fast evolution strategies;airspace congestion smoothing by stochastic optimization;evolutionary optimization based on lagrangian with constraint scaling;solving static and dynamic fuzzy constraint networks using evolutionary hill-climbing;applying family competition to evolution strategies for constrained optimization;supporting polyploidy in genetic algorithms using dominance vectors;an individually variable mutation-rate strategy for genetic algorithms;inductive learning of mutation step-size in evolutionary parameter optimization;a note on the escape probabilities for two alternative methods of selection under gaussian mutation;raising theoretical questions about the utility of genetic algorithms;some geometric and algebraic results on crossover and an analysis of evolutionary algorithms based on neighborhood and step sizes.
The proceedings contain 37 papers. The special focus in this conference is on High Performance Computing. The topics include: The generation of optimized codes using nonzero structure analysis;on the importance of an ...
ISBN:
(纸本)3540637664
The proceedings contain 37 papers. The special focus in this conference is on High Performance Computing. The topics include: The generation of optimized codes using nonzero structure analysis;on the importance of an end-to-end view of memory consistency in future computer sysmtems;high performance distributed object systems;instruction cache prefetching using multilevel branch prediction;high performance wireless computing;high-performance computing and applications in image processing and computer vision;evaluation of multithreaded processors and thread-switch policies;a building block for high performance computers;data dependence path reduction with tunneling load instructions;performance estimation of embedded software with pipeline and cache hazard modeling;an implementation and evaluation of a distributed shared-memory system on workstation clusters using fast serial links;designing and optimizing 3-connectivity communication networks using a distributed genetic algorithm;adaptive routing on the recursive diagonal torus;an algorithm for automatic detection of loop indices for communication overlapping;hybrid approach for non-strict dataflow program on commodity machine;scenario-based hypersequential programming;parallelization of space plasma particle simulation;implementing iterative solvers for irregular sparse matrix problems in high performance fortran;cost estimation of coherence protocols of software managed cache on distributed shared memory system;a portable distributed shared memory system on the cluster environment;a method for runtime recognition of collective communication on distributed-memory multiprocessors;efficiency of parallel machine for large-scale simulation in computational physics and a parallelization method for neural networks with weak connection design.
This paper describes a simulator for the Shiva multiprocessor system and the simulator construction methodology (SCM) used in its creation. The SCM, based on the active functional unit (AFU) construct, is a modern SCM...
详细信息
This paper describes a simulator for the Shiva multiprocessor system and the simulator construction methodology (SCM) used in its creation. The SCM, based on the active functional unit (AFU) construct, is a modern SCM which is flexible, accurate, fast, easy to use, capable of dynamic reconfigurability at run-time, and most of all simple and capable of quick simulator construction. The AFU SCM is capable of all these things through the use of object-oriented software techniques. The Shiva simulator constructed using the AFU SCM is program-driven and capable of micro and macro architectural simulation.
Fault tolerance is a survival attribute of complex computersystems and software in their ability to deliver continuous service to their users in the presence of faults. Formulating an analytic model for dependability...
详细信息
Fault tolerance is a survival attribute of complex computersystems and software in their ability to deliver continuous service to their users in the presence of faults. Formulating an analytic model for dependability and performance evaluation of hardware/software fault tolerant architectures can be quite cumbersome. Also, in practice, isolating the effect of various parameters on a system, while holding the others constant requires exploring a variety of scenarios. It is economically infeasible to build several such systems. simulation offers an attractive mechanism for dependability evaluation and the study of the influence of various parameters on the failure behavior of the system. In this paper, we develop algorithms to simulate the failure behavior of three commonly used fault tolerant architectures, viz., Distributed Recovery Block (DRB), N-Version Programming (NVP) and N-Self Checking Programming (NSCP). We demonstrate the ability of the approach to simulate complex failure scenarios with various dependencies using some illustrative numerical examples.
This report discusses two related activities. The first is the development of a software simulation that predicts the performance of a computer network under different loading conditions. The second is the validation ...
详细信息
This report discusses two related activities. The first is the development of a software simulation that predicts the performance of a computer network under different loading conditions. The second is the validation of that simulation. The simulation model is configured to match a laboratory computer network, and each is subjected to similar traffic. The goal of these activities is to establish the credibility of the simulation, which will be used to predict message latency performance, host-to-host, in more complex tactical system architectures planned for Navy use on AEGIS class ships.
We introduce a general purpose object oriented simulation environment which enables the study of different architectural configurations for servers. The simulation environment has been built on the top of AWESIME whic...
详细信息
We introduce a general purpose object oriented simulation environment which enables the study of different architectural configurations for servers. The simulation environment has been built on the top of AWESIME which provides a process oriented discrete event simulation environment. On the top of AWESIME, the simulator provides the base class definitions for major components required to build large scale high-performance computersystems, and a more advanced model of any particular component can be derived from these base classes and easily replaced with the old one. This feature enables rapid evaluation of alternatives in simulating different architectures. In addition this design allows studying a component of interest in various degrees of detail while keeping the other components at a coarse-level for a faster simulation time.
In this paper we present an approach for quantitative analysis of application-specific dataflow architectures. The approach allows the designer to rate design alternatives in a quantitative way and therefore supports ...
详细信息
In this paper we present an approach for quantitative analysis of application-specific dataflow architectures. The approach allows the designer to rate design alternatives in a quantitative way and therefore supports him in the design process to find better performing architectures. The context of our work is video signal processing algorithms which are mapped onto weakly-programmable, coarse-grain dataflow architectures. The algorithms are represented as Kahn graphs with the functionality of the nodes being coarse-grain functions. We have implemented an architecture simulation environment that permits the definition of dataflow architectures as a composition of architecture elements, such as functional units, buffer elements and communication structures. The abstract, clock-cycle accurate simulator has been built using a multi-threading package and employs object oriented principles. This results in a configurable and efficient simulator. Algorithms can subsequently be executed on the architecture model producing quantitative information for selected performance metrics. Results are presented for the simulation of a realistic application on several dataflow architecture alternatives, showing that many different architectures can be simulated in modest time on a modern workstation.
暂无评论