upper protocol ——RSVP

本文详细介绍了资源预留协议(RSVP)的基本概念、关键属性、历史背景及相关标准。RSVP是一种用于在网络中为特定流量预留资源的传输层协议,支持简单流的服务质量请求,并通过路径消息和预留消息进行操作。

Resource reservation protocol

The Resource ReSerVation Protocol (RSVP ), described in RFC 2205 , is a Transport layer protocol designed to reserve resources across a network for an integrated services Internet . "RSVP does not transport application data but is rather an Internet control protocol, like ICMP , IGMP , or routing protocols" - RFC 2205 . RSVP provides receiver-initiated setup of resource reservations for multicast or unicast data flows with scaling and robustness.

RSVP can be used by either hosts or routers to request or deliver specific levels of quality of service (QoS ) for application data streams or flows. RSVP defines how applications place reservations and how they can relinquish the reserved resources once the need for them has ended. RSVP operation will generally result in resources being reserved in each node along a path.

RSVP is not itself a routing protocol and was designed to interoperate with current and future routing protocols.

RSVP by itself is rarely deployed in telecommunications networks today[citation needed ] but the traffic engineering extension of RSVP, or RSVP-TE , is becoming more widely accepted nowadays in many QoS-oriented networks.

 

Contents

    Main attributes

  1. RSVP requests resources for simplex flows: a traffic stream in only one direction from sender to one or more receivers.
  2. RSVP is not a routing protocol but works with current and future routing protocols.
  3. RSVP is receiver oriented: in that the receiver of a data flow initiates and maintains the resource reservation for that flow.
  4. RSVP maintains soft state (the reservation at each node needs a periodic refresh) of the host and routers' resource reservations, hence supporting dynamic automatic adaptation to network changes.
  5. RSVP provides several reservation styles (a set of reservation options) and allows for future styles to be added to protocol revisions to fit varied applications.
  6. RSVP transports and maintains traffic and policy control parameters that are opaque to RSVP.

   History and related standards

    RSVP is described in a series of RFC documents from the IETF:

  • RFC 2205 : The version 1 functional specification was described in RFC 2205 (Sept. 1997) by IETF . Version 1 describes the interface to admission (traffic) control that is based "only" on resource availability. Later RFC2750 extended the admission control support.
  • RFC 2210 defines the use of RSVP with controlled-load RFC 2211 and guaranteed RFC 2212 QoS control services. More details in Integrated Services . Also defines the usage and data format of the data objects (that carry resource reservation information) defined by RSVP in RFC 2205 .
  • RFC 2211 specifies the network element behavior required to deliver Controlled-Load services.
  • RFC 2212 specifies the network element behavior required to deliver guaranteed QoS services.
  • RFC 2750 describes a proposed extension for supporting generic policy based admission control in RSVP. The extension included a specification of policy objects and a description on handling policy events. (January 2000).
  • RFC 3209 , "RSVP-TE: Extensions to RSVP for LSP Tunnels" (December 2001).
  • RFC 3473 , "Generalized Multi-Protocol Label Switching (GMPLS) Signaling Resource ReserVation Protocol-Traffic Engineering (RSVP-TE) Extensions" (January 2003).
  • RFC 3936 , "Procedures for Modifying the Resource reSerVation Protocol (RSVP)" (October 2004), describes current best practices and specifies procedures for modifying RSVP.
  • RFC 4495 , "A Resource Reservation Protocol (RSVP) Extension for the Reduction of Bandwidth of a Reservation Flow" (May 2006), extends RSVP to enable the bandwidth of an existing reservation to be reduced instead of tearing down the reservation.
  • RFC 4558 , "Node-ID Based Resource Reservation Protocol (RSVP) Hello: A Clarification Statement" (June 2006).

    Key concepts

     The two key concepts of RSVP reservation model are flowspec and filterspec:

    Flowspec

RSVP reserves resources for a flow. A flow is identified by the destination address, the protocol identifier, and, optionally, the destination port. In MPLS a flow is defined as a LSP . For each flow RSVP also identifies the particular quality of service required by the flow although it does not understand the specific information of the flow QoS. This QoS specific information is called a flowspec and RSVP passes the flowspec from the application to the hosts and routers along the path. Those systems then analyse the flowspec to accept and reserve the resources. A flowspec consists of:

  1. Service class
  2. Reservation spec - defines the QoS
  3. Traffic spec - describes the data flow

   Filterspec

The filterspec defines the set of packets that shall be affected by a flowspec (i.e. the data packets to receive the QoS defined by the flowspec). A filterspec typically selects a subset of all the packets processed by a node. The selection can depend on any attribute of a packet (e.g. the sender IP address and port).

The currently defined RSVP reservation styles are:

  1. Fixed filter - reserves resources for a specific flow.
  2. Shared explicit - reserves resources for several flows and all share the resources
  3. Wildcard filter - reserves resources for a general type of flow without specifying the flow; all flows share the resources

An RSVP reservation request consists of a flowspec and a filterspec and the pair is called a flowdescriptor . The effects at the node of each spec are that while the flowspec sets the parameters of the packet scheduler at a node, the filterspec sets the parameters at the packet classifier.

    Messages

There are two primary types of messages:

  • Path messages (path )
The path message is sent from the sender host along the data path and stores the path state in each node along the path.
The path state includes the IP address of the previous node, and some data objects:
  1. sender template to describe the format of the sender data
  2. sender tspec to describe the traffic characteristics of the data flow
  3. adspec that carries advertising data (see RFC 2210 for more details).
  • Reservation messages (resv )
The resv message is sent from the receiver to the sender host along the reverse data path. At each node the IP destination address of the resv message will change to the address of the next node on the reverse path and the IP source address to the address of the previous node address on the reverse path.
The resv message includes the flowspec data object that identifies the resources that the flow needs.

The data objects on RSVP messages can be transmitted in any order. For the complete list of RSVP messages and date objects see RFC 2205 .

    Operation

An RSVP host that needs to send a data flow with specific QoS will transmit an RSVP path message that will travel along the unicast or multicast routes pre-established by the working routing protocol. If the path message arrives at a router that does not understand RSVP, that router forwards the message without interpreting the contents of the message and will not reserve resources for the flow.

When the destination router receives the path message it will:

  1. Make a reservation based on the request parameters. For this the admission control and policy control process the request parameters and can either instruct the packet classifier to correctly handle the selected subset of data packets or negotiate with the upper layer how the packet handling should be performed.
  2. Forward the request upstream (in the direction of the sender). At each node the resv message flowspec can be modified by a forwarding node (e.g. in the case of a multicast flow reservation the reservations requests can be merged).

The Resv message also has FilterSpec object; it defines the packets that will receive the requested QoS defined in the flowspec. A simple filter spec could be just the sender’s IP address and optionally its UDP or TCP port.


Each node in the path can either accept or reject the request.

    Other features

  • Integrity - RSVP messages are appended with a message digest created by combining the message contents and a shared key using a message digest algorithm (commonly MD5 ). The key can be distributed and confirmed using 2 message types: integrity challenge request and integrity challenge response .
  • Error reporting - when a node detects an error, an error message is generated with an error code and is propagated upstream on the reverse path to the sender.
  • Information on RSVP flow - two types of diagnostic messages allow a network operator to request the RSVP state information on a specific flow.
  • Diagnostic facility - An extension to the standard which allows a user to collect information about the RSVP state along a path. RFC2745 - RSVP Diagnostic Messages

    References

  • "Deploying IP and MPLS QoS for Multiservice Networks: Theory and Practice" by John Evans, Clarence Filsfils (Morgan Kaufmann, 2007, ISBN 0-12-370549-5 )

    External links

    RFCs

训练数据保存为deep_convnet_params.pkl,UI使用wxPython编写。卷积神经网络(CNN)是一种专门针对图像、视频等结构化数据设计的深度学习模型,在计算机视觉、语音识别、自然语言处理等多个领域有广泛应用。其核心设计理念源于对生物视觉系统的模拟,主要特点包括局部感知、权重共享、多层级抽象以及空间不变性。 **1. 局部感知与卷积操作** 卷积层是CNN的基本构建块,使用一组可学习的滤波器对输入图像进行扫描。每个滤波器在图像上滑动,以局部区域内的像素值与滤波器权重进行逐元素乘法后求和,生成输出值。这一过程能够捕获图像中的边缘、纹理等局部特征。 **2. 权重共享** 同一滤波器在整个输入图像上保持相同的权重。这显著减少了模型参数数量,增强了泛化能力,并体现了对图像平移不变性的内在假设。 **3. 池化操作** 池化层通常紧随卷积层之后,用于降低数据维度并引入空间不变性。常见方法有最大池化和平均池化,它们可以减少模型对微小位置变化的敏感度,同时保留重要特征。 **4. 多层级抽象** CNN通常包含多个卷积和池化层堆叠在一起。随着网络深度增加,每一层逐渐提取更复杂、更抽象的特征,从底层识别边缘、角点,到高层识别整个对象或场景,使得CNN能够从原始像素数据中自动学习到丰富的表示。 **5. 激活函数与正则化** CNN中使用非线性激活函数来引入非线性表达能力。为防止过拟合,常采用正则化技术,如L2正则化和Dropout,以增强模型的泛化性能。 **6. 应用场景** CNN在诸多领域展现出强大应用价值,包括图像分类、目标检测、语义分割、人脸识别、图像生成、医学影像分析以及自然语言处理等任务。 **7. 发展与演变** CNN的概念起源于20世纪80年代,其影响力在硬件加速和大规模数据集出现后真正显现。经典模型如LeNet-5用于手写数字识别,而AlexNet、VGG、GoogLeNet、ResNet等现代架构推动了CNN技术的快速发展。如今,CNN已成为深度学习图像处理领域的基石,并持续创新。 资源来源于网络分享,仅用于学习交流使用,请勿用于商业,如有侵权请联系我删除!
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值