基于域的分布式系统管理体系结构模型

实现高效、稳定的分布式网络系统管理的核心内容是系统管理体系结构的设计。提出了分布式系统／网络管理体系结构设计的原则,并在分析常用网络管理体系结构的基础上,引入了管理器、管理者的管理者、综合管理器、域等概念,提出了基于域的分层树型体系结构（ＤＢＨＡ）模型。     

What planner for ambient intelligence applications?

The development of ambient intelligence (AmI) applications that effectively adapt to the needs of the users and environments requires, among other things, the presence of planning mechanisms for goal-oriented behavior. Planning is intended as the ability of an AmI system to build a course of actions that, when carried out by the devices in the environment, achieve a given goal. The problem of planning in AmI has not yet been adequately explored in literature. We propose a planning system for AmI applications, based on the hierarchical task network (HTN) approach and called distributed hierarchical task network (D- HTN), able to find courses of actions to address given goals. The plans produced by D-HTN are flexibly tailored to exploit the capabilities of the devices currently available in the environment in the best way. We discuss both the architecture and the implementation of D-HTN. Moreover, we present some of the experimental results that validated the proposed planner in a realistic application scenario in which an AmI system monitors and answers the needs of a diabetic patient.     

Concurrency control in hierarchical multidatabase systems

Over the past decade, significant research has been done towards developing transaction management algorithms for multidatabase systems. Most of this work assumes a monolithic architecture of the multidatabase system with a single software module that follows a single transaction management algorithm to ensure the consistency of data stored in the local databases. This monolithic architecture is not appropriate in a multidatabase environment where the system spans multiple different organizations that are distributed over various geographically distant locations. In this paper, we propose an alternative multidatabase transaction management architecture, where the system is hierarchical in nature. Hierarchical architecture has consequences on the design of transaction management algorithms. An implication of the architecture is that the transaction management algorithms followed by a multidatabase system must be composable– that is, it must be possible to incorporate individual multidatabase systems as elements in a larger multidatabase system. We present a hierarchical architecture for a multidatabase environment and develop techniques for concurrency control in such systems. Edited by R. Sacks-Davis. Received June 27, 1994 / Accepted September 26, 1995     

基于LSF集群系统的分布式并行计算

介绍了LSF集群系统和分布式并行计算的基本概念,基于LSF集群系统提出了一个分布式并行计算的基本结构和分级任务调度策略,另外在此结构上实现了矩阵相乘的分布式并行计算,并对此结构进行了性能评测,最后分析了基于LSF集群系统的分布式并行计算具有的优点和存在的不足.     

多媒体会议系统分层通信结构及共算法研究

目前,多媒体会议系统的通信结构主要分为两种：集中式通信结构和全分布式通信结构。在远程多媒体会议系统中,这两种通信结构都很难满足多媒体会议系统的实时性要求。针对这两种通信结构的缺点与不足,该文提出了一种新的通信结构：分层通信结构。这种通信结构可有效减少远程多媒体会议系统的时间延迟。对于不同的媒体混合技术,我们分别给出了构造其最佳分层通信结构的方法及相应算法。实验数据表明,：相对于集中式通信结构和全分布式通信结构而言,分层通信结构可以有效减少远程多媒体会议系统的时间延尺,从而更好的满足远程多媒体会议系统的实时性要求。     

基于聚合体的联盟管理服务的实现技术

与传统的分布式仿真运行支撑平台不同,StarLink＋采用层次式体系结构的实现方法。由于层次式体系结构中需要多个RTI服务器在网络上协调一致地为大规模仿真提供服务,因而其实现较其它结构的RTI软件要复杂得多。为降低StarLink＋系统管理的复杂性,支持大规模仿真盟员的加入,论文提出了基于”聚合体”的联盟管理服务的实现思想,并基于该方法具体地描述了各个服务的实现算法。该技术能够大大地降低系统管理的复杂性,从而相应地提高仿真系统的执行效率。     

A flexible communications protocol for a distributed surveillance system

On-line surveillance for safety and security is a major requirement of public transport and other public places to address the modern demands of mobility in major urban areas and to effect improvements in quality of life and environment protection. The surveillance task is a complex one involving technology, management procedures and people. Visual surveillance based on Closed Circuit Television system is an important part of such systems, but visual processing is not sufficient and the geographical distribution of devices and management has to be taken into account. In this paper we present a surveillance architecture that reflects the distributed nature of the monitoring task and allows for distributed detection processes, not only dealing with visual processing but also with devices such as acoustic signature detection and mobile smart cards, actuators and a range of other possible sensors. The design uses ideas from control engineering and distributed communications networks resulting in a communications architecture based on CORBA and XML messaging. We have shown how to define a generic device/sensor model appropriate for the surveillance task and sufficiently flexible so as to allow for scalability, expansion and customisation of a practical surveillance task. The paper gives sufficient details on the protocols to show how intelligent detection modules can be integrated as part of this kind of system. The system components have been implemented and integrated in two major successful trials in metropolitan railway stations in London and in Paris, as part of a major EU-funded project (PRISMATICA).     

Service vulnerability scanning based on service-oriented architecture in Web service environments

Web services are becoming the critical components of business application, but they are often invoked with critical software and application bugs that can be explored by malicious users. Because the existing centralized vulnerability scanning systems often face performance bottleneck because of huge amount of tasks, a novel service vulnerability scanning scheme is high desirable. In this paper, we propose a service vulnerability scanning scheme based on service-oriented architecture (SoA) in Web service environments. The scanning scheme contains three components, i.e., domain-oriented distributed architecture, service providing mode based on SoA and hierarchical strategy scheduling model. The hierarchical strategy scheduling model is the key of the scanning scheme, which is used to solve the problems of distributed scheduling management in vulnerability scanning process for Web service environments. We conduct a centralized scanner to compare our scheme with other schemes by the implement of prototype system. Experimental results show that our proposed scheme outperforms other schemes with respect to time cost, accuracy and load.     

基于Agent的网上高校招生管理系统模型

文章提出了基于Agent的分布式网上高校招生管理系统模型。该模型由用户接口Agent、任务Agent和资源Agent三部分组成。文中阐述了各Agent的功能,建立了任务Agent的动态选择机制,描述了Agent间通讯方法。基于Agent的分布式管理系统模型已用于高校招生管理系统中,它提高了网上高考管理系统的性能,解决了分布式环境下从不同资源中获取所需信息的能力。     

Pilot performance and preference for short cycles of automation in adaptive function allocation

The present experiment examined pilot response to the rapid cycling of automation. The experiment was conducted using a multi-task simulation environment consisting of tracking, fuel management, and system monitoring sub-tasks. Monitoring and fuel management sub-tasks were performed manually in all conditions. The tracking sub-task cycled between manual and automated control at fixed intervals of either 15, 30 or 60 sec. These cycle times were completely crossed with three levels of tracking difficulty giving nine within-subject conditions which lasted 5 min each. Performance was measured on each of the sub-tasks, as was pilot fatigue level and subjective workload for the respective conditions. Results indicated that both difficulty and cycle duration significantly affected tracking performance which was degraded with task difficulty and longer cycle times. Fuel management and system monitoring performance were unaffected by tracking difficulty and automation duration. However, a subsequent analysis was conducted using the 15 sec period immediately following each automation episode as a ‘window’ of performance. A different pattern of results was observed. Tracking performance was similarly affected by difficulty, but was no longer affected by cycle duration. Furthermore, fuel management error indicated a trend toward better performance in low difficulty conditions. Results illustrate micro trade-offs within sub-tasks and macro trade-offs between sub-tasks. Overall, the results support the contention that excessively short cycles of automation prove disruptive to performance in multi-task conditions.     

A network based mechanism for managing decomposable tasks via crowdsourcing

If a task is decomposable in a competitive crowdsourcing environment, thereby allowing collaboration, rational workers may choose to divide it into multiple sub-tasks among themselves. But as the winners are selected independently, there is no benefit out of this decomposition. We show that by the appropriate combination of such decomposed solutions, obtained from multiple workers, we can achieve a better solution for a given task. We present a network based mechanism to choose the best mixture of sub-tasks in a competitive environment for selecting collaborating winners.     

分布式结构在IP网络管理中的应用

计算机网络被日益广泛地应用,越来越深入地影响着人类的生活。随着网络规模的不断扩大,管理维护网络的任务变得越来越困难。传统的网络管理系统采用集中式结构,已经不能适应现在普遍存在的大型网络的管理需要。当前存在三种典型网络管理体系结构:集中式、分布式、分层式,通过分析比较三种结构的优缺点,可以看出分布式结构拥有最多的忧点,非常适应当前大型网络管理的需要。本文介绍了一种分布式结构的实现方式。通过选举机制,通过拓扑发现节点之间维持一种动态的主从关系,并且系统对自身节点的故障有检测手段和一定的处理措施,从而大大增强了系统的健壮性。     

基于多主体技术的分布式入侵检测系统的研究与设计

本文提出了一种基于多主体技术的分布式入侵检测系统。该系统将整个网络的入侵检测工作分配到各台主机上进行,克服亍中心化和层次化入侵检测系统所存在的单点失效和处理能力瓶颈问题;将面向整个网络的管理任务和分布式入侵检测任务交给网络安全管理主体NSMA来处理,用NSMA之间的协作代替各个底层检测节点间的协作,提高了系统的的检测和管理能力;在系统设计上引入了功能冗余的思想,在有效克服单点失效问题的同时提高了系统对入侵的实时枪测能力．太文详细论述了系统设计思想和实现方案．并给出了相关技术问题的解决方法。     

基于Web服务发展分布式网络监测

传统的区域网络监测系统只能监测以监测服务器为中心的局部区域。文中提出了虚拟监测子域的概念,研究了基于Web服务技术在区域网络监测系统的集合上构建分布式网络监测系统的技术。各区域网络监测系统的监测功能被定义为Web服务并发布,其管理者通过调用Web服务制定虚拟监测子域,实现分布式网络监测功能。一个分布式网络监测原型系统的试验结果表明,这种分布式网络监测系统能够很好地满足监测大型网络区域和网络管理多样性的需求。     

基于雾计算的计算资源分配方案

雾计算可以为用户提供近距离的数据存储、计算和其他服务，因此雾计算中的任务调度和资源分配已经成为一个新的研究热点。考虑终端用户和雾设备通常处于一种相对开放的状态，扩展了雾计算的体系结构，提出一种开放式雾计算环境中基于稳定匹配的计算资源分配方案，利用雾网络中动态的计算资源协同为用户提供计算服务并收取计算收益，同时终端用户向雾服务器提交任务请求并支付一定的费用。基于稳定匹配的思想，利用子任务的优先级列表、子任务和计算服务设备的偏好列表解决子任务与计算服务设备的分配问题，保证任务的完成时间和计算服务设备的收益。通过实验对方案性能进行了分析，实验结果表明该方案的资源分配时间相对稳定，且在执行雾计算任务时延以及任务违规率上都优于SGA算法和ACOSA算法。     

一个面向复杂任务的多机器人分布式协调系统

基于多智能体系统理论, 研究在非结构、不确定环境下面向复杂任务的多机器人分布式协调系统的实现原理、方法和技术. 提出的递阶混合式协调结构、基于网络的通讯模式和基于有限状态机的规划与控制集成方法, 充分考虑了复杂任务和真实自然环境的特点. 通过构建一个全实物的多移动机器人实验平台, 对规划、控制、传感、通讯、协调与合作的各关键技术进行了开发和集成, 使多机器人分布式协调技术的研究直接面向实际应用, 编队和物料搬运的演示实验结果展示了多机器人协调技术的广阔应用前景.     

基于JMX/JDMK技术的SNMP分级代理

JMX是新的Java平台标准扩展，它支持现有的SNMP网络管理协议。Java DMK是建立在JMX架构之上Java系列的一个部分，并与该架构兼容。在探讨JMX／JDMK技术提供的分级代理方法的基础上，设计出SNMP分级代理体系，以构建灵活、有效及分布的仍以SNMP为基础的网络管理系统。     

面向Storm的数据流编程模型与编译优化方法研究

数据流编程模型将程序的计算与通信分离,暴露了应用程序潜在的并行性并简化了编程难度。分布式计算框架利用廉价PC构建多核集群解决了大规模并行计算问题,但多核集群层次性存储结构和处理单元对数据流程序的性能提出了新的挑战。针对数据流程序在分布式架构下所面临的问题,设计并实现了数据流编程模型和分布式计算框架的结合——在COStream的基础上提出了面向Storm的编译优化框架。框架包括两个模块:面向Storm的层次性任务划分与调度,以及面向Storm的层次性软件流水与代码生成。层次性任务划分利用Storm的任务调度机制将程序所有子任务分配到Storm集群节点内的多核上。层次性软件流水与代码生成将子任务构造成集群节点间的软件流水和节点内多核间的软件流水,并生成相应的目标代码。实验以多核集群为目标平台,在集群上搭建Storm分布式架构,选取数字媒体处理领域典型程序作为测试程序,对面向Storm的编译优化后的程序进行实验分析。实验结果表明了结合方法的有效性。     

A hierarchical task queue organization for shared-memorymultiprocessor systems

There are two basic ways in which waiting ready tasks can be organized: centralized organization or distributed organization. In the centralized organization, a single central task queue is maintained. In the latter case, private ready queue of tasks. Ideally, a central ready queue global to all processors is desired over the distributed organization because the centralized organization provides perfect load sharing. However, the centralized organization is not suitable for large parallel systems because the central task queue could become a system bottleneck. The distributed organization, on the other hand, creates the load imbalance problem, which results in performance deterioration, while techniques have been proposed to reduce the ill-effects of task queue contention in the centralized organization and load imbalance in the distributed organization these techniques introduce problems of their own. We propose the use of a hierarchical task queue organization to incorporate the best features of these two organizations. Our study into the performance of this hierarchical organization shows that a properly designed hierarchical organization provides performance very close to that of the centralized organization while eliminating the ready queue contention problem, we also provide an analysis that identifies and provides guidance for designing the hierarchical task queue organization that avoids ready queue access contention. A brief discussion of task scheduling policies is also included     

智能建筑室内环境分布式可计算WSN任务调度研究

针对智能建筑室内环境下并行计算的动态任务调度问题,构建了基于分布式CPS思想的无线传感器网络(WSN)模型,并分别设计了基于可计算复杂性的任务分配策略和基于动态调度算法的任务调度策略。通过先将任务分配成若干个子任务,采用多带图灵机输入任务,由合适的计算节点进行计算,形成有向无环图,再按调度优先级排列任务,形成任务调度序列表,依序处理任务,从而达到了将任务分配、调度和执行相结合的目的。实验结果表明该策略可有效减少智能建筑室内环境分布式可计算WSN分布运行时任务之间的通讯时间和等待时间,同时提高了任务调度的成功率,最终优化系统的运行效率。