Design and Test of Multibus Adapter System on a Chip for Fault Tolerant Computer Systems

In order to improve the system reliability and performance and to reduce the system cost, volume and weight, we have designed, fabricated and tested the multibus adapter system of a trimodular redundant fault tolerant computer system on a single chip of 5000 gate CMOS gate array. The design, fabrication and test of this single chip system will be discussed..

SPECTRAL TECHNIQUES FOR OFF-LINE TESTING AND DIAGNOSIS OF COMPUTER SYSTEMS

In this paper we present some of our recent results on applications of spectral techniques over finite fields to the problems of testing and diagnosis of computer systems.

Contribution to the Improvement of the Petroleum Products Delivery Policy by the Implementation of a Computer System Based on the Dijkstra Method

The distribution and marketing of petroleum products are the sources of energy for the various activities of human society and household supply. We spent our time researching the competitive factors to reduce the time of distribution of petroleum products, by setting up a computer system based on Dijkstra’s algorithm our contribution will solve a real problem of delivery with direct targeting of gas stations. We set up a tool to help to map based on the method of operational research starting with the identification of delivery point, mapping and then applying the algorithm. Mainly the graph theory and the optimal path in a network by a short path using the method of Dijkstra for the design we used the UML language, the programming we used the JAVA language with very appreciable results to finish with a conclusion.

Study on Method of Comprehensive Evaluating for Information System Projects

In this paper, a new method of evaluation for information system project is proposed on the basis of the meta-synthesis methodology from qualitative analysis to quantitative analysis. DHGF is an integrated method of improved Delphi, analytic hierarchy process, grey interconnect degree and fuzzy comprehensive evaluating. It gives full play to their advantages and controls theirs disadvantages. The feasibility and effectiveness of DHGF are shown in the practical example.

Analysis and Study of Parallel Processing Mode inVLDB Decision Support System

Nowadays, many kinds of computer network data management systems have been built widely in China. People have realized widely that management information system (MIS) has brought a revolution to the management mechanism. Moreover, the managers of company need wide-range and comprehensive decision information more and more urgently which is the character of information explosion era. The needs of users become harsher and harsher in the design of MIS, and these needs have brought new problems to the general designers of MIS. Furthermore, the current method of traditional database development can't solve so big and complex problems of wide-range and comprehensive information processing. This paper proposes the adoption of parallel processing mode, the built of new decision support system (DSS) is to discuss and analyze the problems of information collection, processing and the acquirement of full-merit information with cross-domain and cross-VLDB (very-large database).

Dependent-Chance Goal Programming and an Application

This paper presents a dependent-chance goal programming (DCGP) and gives a successive factoring method to solve DCGP. We also discuss the application of DCGP in Qinhuangdao region for water supply and allocation.

Study on Federated Architecture for GPS/INS/TRN Integrated Navigation System

Based on the information fusion theory, a kind of integrated navigation system integration for cruise missile is presented in this paper. Besides, the way with which the system is integrated and the related data fusion technique are discussed. Information-fusion-based hybrid navigation system integration can fully utilize information provided by all kinds of navigation sensor subsystem and can improve the precision of the system effectively. Simultaneously, the reconstructing ability ensures the system of great reliability.

IDSS: Designing to Extend the Cognitive Limits

The paper presents the conceptual and operational basis of the creation of IDSS based on our recent research experience. In this paper, an intelligent decision support system, IDSS is defined as: any interactive system that is specially designed to improve the decision making of its user by extending the user's cognitive decision making abilities. As a result, this view of man-machine joint cognitive system stresses the need to use computational technology to aid the user in the decision making process. And the human's role is to achieve total systems's objectives. The paper outlines the designing procedure in successive steps. First, the decision maker's cognitive needs for decision support are identified. Second, the computationally realizable support functions are defined that could be provided by IDSS. Then, the specific techniques that would best fill the decision needs are discussed. And finally, for system implementation the modern computational technology infrastructure is emphasized.

计算机系统导论课程改革探索(英文)

In the paper,we analyzed the teaching problems in the course of "Introduction to Computer Systems",proposed guidelines with teaching practice and implement bilingual teaching.Then we explored the reform ways in the course of "Introduction to Computer Systems",to enhance students' self-learning ability through the full use of networked multimedia tools;and also the improvement of the teaching level of teachers.

ESTIMATION OF ATTRACTION DOMAIN AND EXPONENTIAL CONVERGENCE RATE OF CONTINUOUS FEEDBACK ASSOCIATIVE MEMORY

Analytical techniques and Liapunov method were used for the estimation of the attraction domain of memory patterns and local exponential stability of neural networks. The results were used to design efficient continuous feedback associative memory neural networks. The neural network synthesis procedure ensured the gain of large exponential convergence rate without reduction of the attraction domain.

PARALLEL COMPUTATIONAL ALGORITHM OF SUBSTRUCTURE METHOD OF LARGE-SCALE STRUCTURE ANALYSIS

In this paper, according to the parallel environment of ELXSI computer, a parallel solving process of substructure method in static and dynamic analyses of large-scale and complex structure has been put forward, and the corresponding parallel computational program has been developed.

Hybrid Nanoelectronics： Future of Computer Technology

Nanotechnology may well prove to be the 21st century＇s new wave of scientific knowledge that transforms people＇s lives. Nanotechnology research activities are booming around the globe. This article reviews the recent progresses made on nanoelectronic research in US and China, and introduces several novel hybrid solutions specifically useful for future computer technology. These exciting new directions will lead to many future inventions, and have a huge impact to research communities and industries.

Communication Analysis and Granularity Assessment for a Transputer-Based System

This paper exposes the concurrent characteristic in the communication between processors in a transputer based system, and proposes an approximate formula for assessing the granularity of distributing the parallel processes among processors. The really measured results for a pipeline sorting program running on the T-rack confirm these deduced results.

A new family of two-dimensional triple-codeweight asymmetric optical orthogonal code for OCDMA networks

A new generation algorithm of two-dimensional triple-codeweight asymmetric optical orthogonal codes for optical code division multiple access （OCDMA） networks is proposed. The code cardinality is obtained and the error-probability performance for corresponding OCDMA system is analyzed. The codes with two constraints （i.e., auto- and cross-correlation properties） being unequal are taken into account. On the premise of fixed system resources, the code of the error-probability performance, it is shown cardinality can be significantly improved. By analysis that the codes with different parameters have different performances. Therefore, this type of codes can be applied to support diverse quality of service （QoS） and satisfy the quality requirement of different multimedia or distinct users, and simultaneously make the better use of bandwidth resources in oDtical networks.

Optimal Partitioning and Granularity of Uniform Task Graphs

Task partitioning is an important technique in parallel processing.In this paper,we investigate the optimal partitioning strategies and granularities of tasks with communications based on several models of parallel computer systems.Different from the usual approach,we study the optimal partitioning strate- gies and granularities from the viewpoint of minimizing T as well as minimizing NT^2,where N is the number of processors used and T is the program execution time using N processors.Our results show that the optimal partitioning strategies for all cases discussed in this paper are the same——either to as- sign all tasks to one processor or to distribute them among the processors as equally as possible de- pending only on the functions of ratio of running time to communication time R/C.

Optical attenuator based on phase modulation of a spatial light modulator

In this Letter, we propose an optical attenuator based on the phase modulation of a spatial light modulator （SLM）. In this system, we use two polarized beam splitters （PBSs） to control the polarized light and one SLM to modulate the phase of the polarized light. In the initial state, the light beam is divided into p-light and s-light when it passes through the first PBS. When the light passes through the second PBS, s-light is reflected and p-light is detected by the CCD camera. By loading different grayscales on the SLM, p-light changes its polarized state to s-light. The light power can be attenuated during the loading process. Our experiment shows that the system can obtain a wide optical attenuation from 1-27.2 dB. When loading two grayscales, the SLM has a fast switching time of 25 ms under a low actuated voltage of 5.5 V. The response time of the optical attenuator depends on the switching time of the SLM. Therefore, the system can also have a fast response time. By using the method of spatial multiplexing and adding two mirrors in the system, it can also be extended into a 1 × 2 optical switch. The results verify its feasibility. The optical attenuator has wide applications in photonic signal processing and fiber-optic communication.

Macro-Dataflow Computational Model and Its Simulation

This paper discusses the relationship between parallelism granularity and system overhead of dataflow computer systems,and indicates that a trade-off between them should be determined to obtain optimal efficiency of the overall system.On the basis of this discussion,a macro-dataflow computational model is established to exploit the task-level parallelism.Working as a macro-dataflow computer,an Ex- perimental Distributed Dataflow Simulation System (EDDSS) is developed to examine the effectiveness of the macro-dataflow computational model.

TAUS: A File-Based Software Understanding Tool

A program called TAUS, a Tool for Analyzing and Understanding Software, was developed. It is de- signed to help the programmer analyze and understand the software interactively. Its aim is to reduce the depen- dence on human intelligence in software understanding and improve the programmer's understanding productivity. The design and implementation of TAUS and its application are described.

On the Execution Mechanisms of Parallel Graph Reduction

Parallel graph reduction is a promising model for new generation computer because of its amenability to both programming and parallel computing. In this paper, an initial design for a parallel graph reduction model, PGR model, is presented, which employs eager evaluation strategy to exploit conservative parallelism and provides with primitives and associated tags of nodes to synchronize concurrent tasks .Moreover, a direct operational description of graph reduction in terms of high level instructions (primitives) is given to obtain a virtual machine, called PGRVM.

TST——An Algorithm for Global Microcode Compaction with Timing Constraints

Existing global microcode compaction approaches have all assumed a target architecture that has microoperation conflicts and data dependencies as the two fundamental compaction constraints.However, new practical micromachine features demand that the timing constraint be introduced into the traditional compaction model to guarantee compaction correctness.This paper starts by an analysis on the nature of timing constraints,then modifies the roles for microoperation motions,presents an algorithm,TST,based on Trace Scheduling,for global compaction under timing constraints,and finally shows results of experiments.