资源预订协议RSVP和ST-II的分析和比较

文章介绍了Internet网络中支持资源预订的协议RSVP和ST-II的基本原理和工作过程,并对这两种资源预订协议的特性作了分析和比较,阐述了这两种协议在多媒体通信中应用前景.     

UPnP和IGRS互操作性研究

UPnP和IGRS是数字家庭领域中实现设备无缝连接的两大主流协议,由于这两种协议对各自设备及其所提供服务的描述格式各不相同,因此基于不同协议的设备不能直接相互识别、连接.针对此情况,从IGRS和UPnP两大主流协议的内部机制层面分析了两种协议的差别,并阐述了基于这两种协议设备互联互通的可行性,从而实现基于不同协议的信息设备资源共享和协同服务,为IGRS和UPNP设备的互联互通提供理论支持.     

复杂网络环境下SNMP和CMIP集成管理技术

网络管理协议是网络管理系统重要组成部分,简单网络管理协议(SNMP)和公共信息管理协议(CMIP)是目前两种重要的网络管理协议,这两种管理协议在信息结构定义和通信机制上存在很大差异,论文对这两种管理协议的集成技术进行了研究,并重点讨论用基于SNMP协议的管理系统管理基于CMIP协议的网络设备的委托代理技术的一种实现.     

Epilogue和BSD协议栈的实现机制的比较

协议栈是网络通信中最重要的软件部分,在数据交换中起着重要作用.对其进行研究分析有助于进行特定情况下的协议栈实现.文中对两种典型协议栈的实现方式进行比较,分析了两者在各个层次上的协议的实现,并比较其协议栈所占用资源,从而对其可靠性和进行二次开发的可能性进行研究,最后分析了这两种协议栈的实现方式的优缺点.两种协议栈的比较可为实现某种协议栈时提高其可靠性、可读性、可移植性提供有益参考.     

Epilogue和BSD协议栈的实现机制的比较

协议栈是网络通信中最重要的软件部分,在数据交换中起着重要作用。对其进行研究分析有助于进行特定情况下的协议栈实现。文中对两种典型协议栈的实现方式进行比较,分析了两者在各个层次上的协议的实现,并比较其协议栈所占用资源,从而对其可靠性和进行二次开发的可能性进行研究,最后分析了这两种协议栈的实现方式的优缺点。两种协议栈的比较可为实现某种协议栈时提高其可靠性、可读性、可移植性提供有益参考。     

H.323与MGCP在VoIP应用中的分析与比较

H．323和MGCP是目前在VoIP技术中被广泛应用的两种信令控制协议。文中从网络结构、呼叫控制机制的实现两方面介绍了H．323和MGCP这两种协议，并从实现原理、可扩展性等几方面对二者作了比较。最后，针对两种协议在VoIP中的应用指出了两者的优缺点，并提出了在VoIP系统的建设中采用两种协议的应用范畴。     

远程接入虚拟专用网的两种增强型设计方案

目前远程用户接入虚拟专用网主要基于L2TP协议和IPsec协议，但是这两种协议都存在各自的缺点，文章对此作了详细的分析，同时在这两个协议的基础上提出了两种增强型的解决方案：L2TP/sed接入方案和基于PKI和混合认证IEK的接入方案。     

基于3DES和SM4的Modbus安全通信算法

Modbus协议广泛用于工业控制系统(ICS)中,Modbus TCP协议通常用于SCADA系统中的人机界面和可编程逻辑控制器(PLC)之间的通信.本文介绍了Modbus TCP协议的主要安全漏洞,并提出两种加密算法满足Modbus TCP协议的保密性要求,即3DES加密算法和国密SM4算法,这两种分组密码算法分别属于...     

两种网络存储协议Fibre channel与iSCSI比较分析

并行SCSI受到传输距离和连接设备数的限制，不能满足海量信息存储应用的需要。光纤通道和iSCSI作为两种重要的网络存储协议，已得到广泛的研究和应用，都可以用于高性能数据存储和容灾系统中传输SCSI协议数据。从协议模型、协议特点和I／O性能几个方面对这两种协议进行了详细的比较分析，指出了各自的特点和应用领域。     

一种调和两种观点的安全协议分析法

在安全协议的形式化分析中,有两种完全不同的观点：符号方法和计算方法,两者各有优缺点．通过对两种观点的研究,按照优势互补的原则将两者相调和,建立了一种新的形式化分析方法．该方法将协议的安全性分为完全性和正确性,通过分析这两种性质,不仅能考察协议在逻辑上能否达到预期目标,而且能考察敌手在计算上的攻击能力．第一次从逻辑可靠性和计算可靠性两方面对协议进行了综合的安全性分析．     

基于Linux平台的NSIS协议试验床搭建与端到端QoS保障机制研究

在网络通信中,服务质量QoS（Quality of Service）是衡量一个网络可靠性、稳定性的重要指标。本文阐述了服务质量QoS的性能指标、模型框架,并从QoS保障控制层面的资源预留角度介绍了资源预留常用的RSVP（Resource ReSerVation Protocol）与NSIS（Next Steps In Singnaling）协议,然后介绍了Linux环境下基于NSIS协议的试验床搭建过程,最后在试验床的基础上实现了端到端的资源预留并进行了验证。     

Performance evaluation framework for IP resource reservation signalling

This paper investigates the scalability limitations of signalling-based Quality of Service (QoS) provisioning using the RSVP and Boomerang protocols as examples. We establish a performance evaluation framework, which contains per hop metrics for describing the router’s signalling capability and per reservation metrics for characterising the overall performance of the reservation protocol. The per hop metrics are captured in a simple router model, using parameters that are based on actual measurements. This router model is used in a simulation scenario, which characterises the end-to-end impact and overall performance of the RSVP and Boomerang reservation protocols.

This paper summarises the results of router benchmarking measurements and simulations. The per session results highlight that, with thorough selection of the control protocol, signalling-based resource reservation is a realistic approach in the Internet. It is also shown that the selection of the signalling protocol can seriously influence the reservation performance of a router and thus the overall performance of the entire network as well. The presented performance evaluation framework can be used for network dimensioning and it can even aid the design of new resource reservation protocols and new router devices.     

一个自适应的支持资源预定的路由算法

新型的应用问题，如视屏会议、分布式模拟等需要保障质量，资源豫剧 定协议如ＲＳＶＰ等可以用来提供有保障的服务质量，介它需要有新的路由协议。为此，文中提供一种自适应的、支持资源预定的、多点通讯的路由算法，并描述了该算法的特性。     

基于效益函数与优先级的网格服务质量保证

将资源预约请求作为网格调度对象是提高网格整体服务质量的有效手段。针对传统预约方式成功率低的问题,提出将效益函数与优先级引入资源预约,采用可连续预约的方法,在减少预约双方交互负担的同时,有效提高资源预约成功率。通过模拟与比较,表明该算法在应对大量预约请求时均有较好的性能。     

协调路由和资源预留机制的实时多播协议

流媒体应用程序必然涉及网络上的实时任务处理。在一般情况下每个数据流都有很多接收端,因此确保网络传输的时间限制的多播资源预留协议非常必要。提出了一个针对一对多,即一个发送端、多个接收端的实时多播协议,因为未涉及到多对多的情况,故称为简单实时多播协议。协议通过将路由机制和资源预留机制用一个通用接口协调起来,提供了一个更为灵活的连接建立方式,并同其它已存在的流媒体协议如RSVP、ST-II进行比较和分析。     

Efficient Management and Packets Forwarding for Multimedia Flows

Well known requirements for handling multimedia flows in routers are resource reservation and fast packets forwarding. The former takes into account the quite stable and long lasting bandwidth occupation, whereas the latter takes into account the large number of packets routed along the same path. Many techniques have been proposed and standardized to face these requirements, but their application is often complex, expensive and sometimes limited by the need of agreements between network managers of the many networks and Autonomous Systems. This paper introduces IMFM (Integrated Multimedia Flows Management), an innovative, scalable, and extremely lightweight technique to provide routers deterministic and dynamic resource reservation and a fast forwarding table lookup. It is based on a distributed linked data structure that allows direct (searchless) access to entries in the routers’ tables, extending the resource reservation algorithm REBOOK. Unlike conventional virtual circuits, IMFM does not require any interaction with (nor change in) the underlying routing protocols and autonomously recovers from errors, faults and route changes. If information stored in its data structure becomes obsolete, packet handling is reverted to best-effort, lookup-driven forwarding, so that packets are never dropped nor misrouted. IMFM can be gradually deployed, providing a framework for congestion avoidance solutions and increasing the forwarding speed in IMFM-aware router even along partially IMFM-unaware paths. IMFM has been fully implemented. Experiments have been designed to demonstrate its feasibility and the measured performance is reported and compared with existing techniques.     

支持综合业务的Internet-ATM网络及互联

文中阐述了Ｉｎｔｅｒｎｅｔ上综合业务的模式,介绍了资源预订协议ＲＳＶＰ、Ｉｎｔｅｒｎｅｔ流协议ＳＴ－ＩＩ、资源预先预订协议ＲｅＲＡ和几种主要的Ｉｎｔｅｒｎｅｔ与ＡＴＭ主干网络的互联协议以及传统的ＴＣＰ／ＩＰ协议族的报文在ＡＴＭ网络中传送的协议ＩＰ ｏｖｅｒＡＴＭ;阐述了局域网模拟ＬＡＮＥ的体系结构及其实现的一个实例和ＡＴＭ上支持多种网络层议（如ＩＰ,ＩＰＸ／ＳＰＸ,ＡｐｐｌｅＴａｌｋ等）的规范ＭＰ     

Extensions to P2MP RSVP-TE for VPN-specific state provisioning with fair resource sharing

Among the resource provisioning algorithms for the hose-based Virtual Private Network (VPN) Quality of Service (QoS), VPN-specific state provisioning allows the service provider to obtain highest resource multiplexing gains. In this paper, we show that the existing resource reservation protocols proposed for the Internet are not appropriate for the VPN-specific state provisioning. Furthermore, since the VPN-specific state provisioning makes the reserved resources to be randomly shared by the sites belonging to the same VPN, a site generating heavy traffic may unfairly dominate the resources reserved for the VPN. We propose extensions to an existing resource reservation protocol proposed for the Internet, i.e., P2MP RSVP-TE (Point-to-Multipoint Resource Reservation Protocol-Traffic Engineering), for the resource reservation of VPN-specific state provisioning. The proposed extensions also enable the fair usage of reserved resources among the users of a VPN that is provisioned by the VPN-specific state. Through simulation experiments, the effects of deploying a fair usage mechanism into the resource reservation of VPN-specific state provisioning is presented.     

全IP移动网络中的动态资源预留机制与QoS研究

全IP移动网络中的QoS是最近的研究热点之一,其中资源预留协议(Resource ReserVation Protocol,RSVP)是一个关键技术。为了保证切换时的QoS,需要进行资源预留。针对RSVP不适应移动环境的缺点,文中在全IP移动网络的接入网的微移动域中,提出一种动态资源预留的QoS解决方案,实现微移动协议和RSVP协议之间的互操作。当有多个移动主机需要进行预留时,需要对它们进行优先级排序。文中提出了一种在全IP移动网络中基于信令预测优先级排队的动态排序算法,以有效地对多个移动主机进行资源预留。仿真结果表明文中提出的算法的有效性。     

A comparison of wavelength reservation protocols for WDM optical networks

In WDM optical networks, wavelength is the critical resource for efficient communication of traffic. Hence, suitable protocols are required to allocate and use the wavelengths efficiently. We have studied the wavelength reservation protocols of such networks, which are already developed and reported in literature. Following their historical development, we have outlined some generalized classification for them and provided comparison of their performances. Finally, on the basis of the previous works, we have discussed some future scopes, which may be explored for further improvement in performance of the protocols.