CN1780230A - Method for managing a network of resource-constrained devices - Google Patents

Abstract

Translated fromChinese

一种由资源受限设备构成的网络的管理方法，包括步骤：a)确定所述网络是树形拓扑；b)在所述网络中，设备之间通过链路层协议报文交互来进行网络拓扑发现，以搜集网络拓扑信息，构造网络拓扑数据信息；以及c)基于所述网络拓扑数据信息，选择指定设备，通过链路层协议报文交互管理信息，来进行配置管理。该方法原理简单，开销极小，额外改动较少，系统代价低，远程管理功能获得了有效提高。

A method for managing a network composed of resource-limited devices, comprising the steps of: a) determining that the network is a tree topology; b) in the network, exchanging link layer protocol messages between devices to implement network Topology discovery, to collect network topology information and construct network topology data information; and c) select a designated device based on the network topology data information, and perform configuration management by exchanging management information through link layer protocol messages. The principle of the method is simple, the overhead is very small, the extra modification is less, the system cost is low, and the remote management function is effectively improved.

Description

Translated fromChinese

由资源受限设备构成的网络的管理方法Method for managing a network of resource-constrained devices

技术领域technical field

本发明涉及网络管理方法，具体地，涉及一种对由资源受限设备构成的网络进行管理的网络管理方法。The invention relates to a network management method, in particular to a network management method for managing a network composed of resource-limited devices.

背景技术Background technique

随着技术发展，数据通信设备在通讯运营、工业控制、家庭消费等领域的应用越来越广泛，并提出一些特殊需求。其中很重要的一个需求是：如何为资源受限设备构成的网络提供简单实用、成本低廉的设备管理方法。尤其在，用户驻地网、智能网络家电以及基于网络的分布式控制系统中，该需求尤为迫切。例如，用户驻地网中楼道交换机，分布式控制系统中智能传感器以及智能家居中智能家电，都需要通过网络进行集中管理，但是相关设备却都具有硬件资源严重受限的特点。由于大量采用低性能CPU或单片机，受硬件资源限制，传统网络管理方法难以应用。With the development of technology, data communication equipment is more and more widely used in communication operations, industrial control, household consumption and other fields, and some special requirements are put forward. One of the most important requirements is: how to provide a simple, practical and low-cost device management method for a network composed of resource-constrained devices. Especially in customer premises network, intelligent network appliances and network-based distributed control system, the demand is particularly urgent. For example, the corridor switches in the customer premises network, the smart sensors in the distributed control system, and the smart home appliances in the smart home all need to be managed centrally through the network, but the related devices all have the characteristics of severely limited hardware resources. Due to the extensive use of low-performance CPUs or single-chip microcomputers, limited by hardware resources, traditional network management methods are difficult to apply.

目前针对局域网络设备的管理方式主要有四种方式：控制台、简单网络管理协议(SNMP)、基于TELNET的字符界面和基于Web的图形管理界面。At present, there are mainly four management methods for LAN devices: console, Simple Network Management Protocol (SNMP), TELNET-based character interface and Web-based graphical management interface.

图1为控制台方式。该种方式将控制台主机(101)通过串口电缆(102)直接连接网络设备(103)的管理串口，通过串口电缆直接传输基于字符的控制命令，并返回基于字符的结果信息，从而完成配置管理。该种方法优点在于简单实用，系统要求低；缺点是串口电缆无法远距传输，工作人员只能亲临现场，无法进行远程管理。Figure 1 shows the console mode. In this way, the console host (101) is directly connected to the management serial port of the network device (103) through the serial port cable (102), the character-based control command is directly transmitted through the serial port cable, and the character-based result information is returned, thereby completing the configuration management . The advantage of this method is that it is simple and practical, and the system requirements are low; the disadvantage is that the serial cable cannot be used for long-distance transmission, and the staff can only visit the site in person, and cannot perform remote management.

图2为简单网络管理协议(SNMP)。该协议中SNMP Manager(201)和网络设备内嵌的SNMP Agent(204)通过Internet云(202)进行SNMP报文(203)传输来完成管理信息交互。SNMP报文(203)包含Set、Get、Get-Next和Trap等多种报文类型，SNMP Agent和SNMPManager各自构造、解析并执行相关配置、查询、报告等功能。SNMP协议基于IP(Internet Protocol)协议，因此设备必须配置IP地址。优点在于遵循IETF RFC标准，便于使用统一网管平台，可以远程管理；缺点为，设备必须支持SNMP协议，软件复杂。Figure 2 shows the Simple Network Management Protocol (SNMP). In this protocol, the SNMP Manager (201) and the SNMP Agent (204) embedded in the network equipment perform SNMP message (203) transmission through the Internet cloud (202) to complete management information interaction. SNMP message (203) includes multiple message types such as Set, Get, Get-Next and Trap, and SNMP Agent and SNMP Manager respectively construct, analyze and perform functions such as relevant configuration, query and report. The SNMP protocol is based on the IP (Internet Protocol) protocol, so the device must be configured with an IP address. The advantage is that it follows the IETF RFC standard, it is convenient to use a unified network management platform, and it can be managed remotely; the disadvantage is that the device must support the SNMP protocol, and the software is complicated.

除此之外，还有基于TELNET的字符管理界面。该字符管理界面，包含主机侧TELNET Client和网络设备侧TELNET Server，二者通过Internet云进行TELNET报文传输来完成字符交互。一般TELNET和控制台结合完成远程管理。设备需要支持基于IP协议的TELNET协议，协议复杂，设备需要配置IP地址。In addition, there is a character management interface based on TELNET. The character management interface includes the TELNET Client on the host side and the TELNET Server on the network device side, both of which transmit TELNET messages through the Internet cloud to complete character interaction. Generally, TELNET and console are combined to complete remote management. The device needs to support the TELNET protocol based on the IP protocol. The protocol is complex, and the device needs to be configured with an IP address.

与SNMP、TELNET管理方式相比，基于WEB的图形管理界面人机交互更为友好。但是，网络设备必须内嵌Web服务器，必须支持基于IP协议的HTTP协议，实现复杂，需要配置IP地址。Compared with SNMP and TELNET management methods, the human-computer interaction based on the WEB graphical management interface is more friendly. However, the network device must be embedded with a Web server and must support the HTTP protocol based on the IP protocol, which is complicated to implement and needs to be configured with an IP address.

对于由资源受限设备构成的网络，要求“低成本有效实现设备远程管理”。而现有四种网络管理方式对此需求均存在局限性。For a network composed of resource-constrained devices, "low-cost and effective remote management of devices" is required. However, the existing four network management methods all have limitations on this requirement.

控制台方式，简单、有效、成本低廉，但不能实现有效远程网络管理。The console mode is simple, effective, and low-cost, but it cannot realize effective remote network management.

简单网络管理协议、基于TELNET的字符界面和基于Web的图形管理界面三种方法均基于TCP/IP(Transfer Control Protocol/Internet Protocol)协议栈，功能强大，却大大增加设备成本。目前IP地址稀缺，而基于TCP/IP协议栈的管理方式却要求为每个网络节点分配IP地址。实际网络部署中，此三种方法难以有效应用，更不能降低设备成本。图3描述了简单网络管理协议、基于TELNET的字符界面和基于Web的图形管理界面相关协议的网络分层示意图。如图所示，支持这三种方案，除应用层相关协议外，还必须实现传输层、网络层及链路层相关协议。Simple network management protocol, TELNET-based character interface and Web-based graphical management interface are all based on TCP/IP (Transfer Control Protocol/Internet Protocol) protocol stack, which are powerful, but greatly increase the cost of equipment. At present, IP addresses are scarce, but the management method based on the TCP/IP protocol stack requires that each network node be assigned an IP address. In actual network deployment, these three methods are difficult to apply effectively, let alone reduce equipment costs. Figure 3 depicts the network layered schematic diagram of the simple network management protocol, TELNET-based character interface and Web-based graphical management interface related protocols. As shown in the figure, to support these three solutions, in addition to the application layer related protocols, the transport layer, network layer and link layer related protocols must also be implemented.

此外，这四种方式都不能监控网络拓扑变化，实时绘制网络拓扑。实际应用中需要二次开发，这一点将进一步增加设备成本。In addition, none of these four methods can monitor network topology changes and draw network topology in real time. Secondary development is required in practical applications, which will further increase equipment costs.

发明内容Contents of the invention

本发明的目的是提供一种由资源受限设备构成的网络的管理方法。该方法能够以较低的成本，实现网络任意位置的远程设备管理。本发明方法尤其适于由资源受限设备构成的网络。The purpose of the present invention is to provide a management method for a network composed of resource-limited devices. The method can realize remote device management at any position of the network at a lower cost. The inventive method is particularly suitable for networks composed of resource-constrained devices.

为实现上述目的，一种由资源受限设备构成的网络的管理方法，包括步骤：a)确定所述网络是树形拓扑；b)在所述网络中，设备之间通过链路层协议报文交互来进行网络拓扑发现，以搜集网络拓扑信息，构造网络拓扑数据信息；以及c)基于所述网络拓扑数据信息，选择指定设备，通过链路层协议报文交互管理信息，来进行配置管理。In order to achieve the above object, a method for managing a network composed of resource-limited devices includes the steps of: a) determining that the network is a tree topology; b) in the network, reporting between devices through a link layer protocol Network topology discovery by text interaction to collect network topology information and construct network topology data information; and c) select a designated device based on the network topology data information, and perform configuration management by exchanging management information through link layer protocol messages .

将步骤b)和步骤c)的机制分别称为“源路径回溯”和“虚拟管理通路”，采用这两种机制实现网络任意位置的远程设备管理，解决了在资源受限设备构成的网络中低成本有效实现设备远程管理这一实际问题。“源路径回溯”算法进行拓扑发现，可以绘制全网拓扑树，并且非管理节点设备只转发协议报文，无需维护信息表。“虚拟管理通路”通过重定向机制直接建立管理节点到被管理节点控制台之间通讯，在管理节点键入的管理命令被中继到被管理节点控制台，被管理节点控制台输出中继返回管理节点控制台。原理简单，开销极小。“虚拟管理通路”机制，使得本地控制台管理才能进行的工作，如“基于Xmodem的远程系统软件升级”，能够以远程方式进行。除“虚拟管理通路”实现外，不需要任何额外改动，系统代价低，远程管理功能获得有效提高。如此功能包括：命令行、Xmodem/Zmodem文件传输、基于Xmodem的远程系统软件升级、全网远程热启动、基于脚本的复杂系统配置等。由于在“虚拟管理通路”机制中，管理数据报文传输路径中，中继设备只需转发，不需要解析报文，因此可以基于高级语言(JAVA等)开发网络管理软件NMS(Network ManagementSystem)，而NMS和网络设备节点之间按照“虚拟管理通路”机制通讯。因此，网络管理软件中配置管理命令可以基于字符脚本文件进行实现。The mechanisms of step b) and step c) are called "source path backtracking" and "virtual management path" respectively. Using these two mechanisms to realize remote device management at any position in the network solves the problem in a network composed of resource-constrained devices. The practical problem of realizing remote management of equipment effectively at low cost. The "source path backtracking" algorithm performs topology discovery and can draw the topology tree of the entire network, and non-management node devices only forward protocol packets without maintaining information tables. "Virtual management channel" directly establishes communication between the management node and the console of the managed node through the redirection mechanism. node console. The principle is simple and the overhead is minimal. The "virtual management channel" mechanism enables the work that can only be done by local console management, such as "Xmodem-based remote system software upgrade", to be done remotely. Except for the implementation of "virtual management channel", no additional changes are required, the system cost is low, and the remote management function is effectively improved. Such functions include: command line, Xmodem/Zmodem file transfer, remote system software upgrade based on Xmodem, remote hot start of the whole network, complex system configuration based on script, etc. Because in the "virtual management path" mechanism, in the transmission path of management data packets, the relay device only needs to forward and does not need to parse the packets, so the network management software NMS (Network Management System) can be developed based on high-level languages (JAVA, etc.), The communication between the NMS and the network device nodes is based on the "virtual management channel" mechanism. Therefore, configuration management commands in network management software can be implemented based on character script files.

附图说明Description of drawings

图1示出了传统的控制台网络管理方式的示意图。FIG. 1 shows a schematic diagram of a traditional console network management method.

图2示出了传统的SNMP网络管理方式的示意图。FIG. 2 shows a schematic diagram of a traditional SNMP network management method.

图3示出了简单网络管理协议、基于TELNET的字符界面和基于Web的图形管理界面相关协议的网络分层示意图。Fig. 3 shows a network layered diagram of the simple network management protocol, the TELNET-based character interface and the web-based graphical management interface related protocols.

图4示出了一种包含闭环的网络拓扑的示例。Figure 4 shows an example of a network topology including closed loops.

图5示出了一种树形网络拓扑的示例。Fig. 5 shows an example of a tree network topology.

图6示出了以图5所示拓扑为例进行拓扑发现的流程图。FIG. 6 shows a flow chart of topology discovery by taking the topology shown in FIG. 5 as an example.

图7示出了根据图6所示方法获得的拓扑信息来利用“虚拟管理通路”进行配置管理的流程图。FIG. 7 shows a flow chart of using "virtual management path" to perform configuration management according to the topology information obtained by the method shown in FIG. 6 .

图8示出了应用本发明方法的用户驻地网的示例。Fig. 8 shows an example of a customer premises network to which the method of the present invention is applied.

图9示出了应用本发明方法的智能家电网络的示例。Fig. 9 shows an example of a network of smart home appliances applying the method of the present invention.

具体实施方式Detailed ways

现在参考附图，对本发明的实施例进行说明。Embodiments of the present invention will now be described with reference to the drawings.

“资源受限设备构成的网络”的拓扑结构可能为任意网状拓扑结构，如图4所示，该网络拓扑种含有闭环(401，402)。然而，在大多数情况下，“资源受限设备构成的网络”已为树状拓扑，如图5所示。如果“资源受限设备构成的网络”拓扑中含有闭环，可以采用如“生成树”(IEEE802.1d)等的公知方法形成无闭环的树状拓扑。The topology of the "network composed of resource-limited devices" may be any mesh topology. As shown in FIG. 4, this network topology contains closed loops (401, 402). However, in most cases, the "network of resource-constrained devices" has a tree topology, as shown in Figure 5. If the "network composed of resource-limited devices" topology contains closed loops, a known method such as "spanning tree" (IEEE802.1d) can be used to form a tree topology without closed loops.

本发明在设备间只通过链路层协议报文交互来进行拓扑发现和配置管理。其中“源路径回溯”用于发现拓扑，而“虚拟管理通路”用于配置管理。两者的相关查询、应答报文均为链路层报文。如果物理链路为以太网，则链路层报文标志为以太网链路帧目的MAC域填充指定多播MAC地址，如0×0180cccccc。如果物理链路为并行口或者串行口，也应适当定义帧格式，如此类推。该指定多播MAC地址或者帧格式应被“资源受限设备构成的网络”内所有设备知晓。The present invention performs topology discovery and configuration management between devices only through link layer protocol message exchange. Among them, "source path backtracking" is used to discover topology, while "virtual management path" is used for configuration management. Both related query and response messages are link layer messages. If the physical link is Ethernet, the link layer packet mark is filled with the specified multicast MAC address for the destination MAC field of the Ethernet link frame, such as 0×0180cccccc. If the physical link is a parallel port or a serial port, the frame format should also be properly defined, and so on. The specified multicast MAC address or frame format should be known by all devices in the "network composed of resource-limited devices".

下面以图5和图6作为示例来对本发明的方法进行详细说明。实际网络拓扑和节点布置并不局限于示例所示。The method of the present invention will be described in detail below by taking FIG. 5 and FIG. 6 as examples. Actual network topologies and node arrangements are not limited to those shown in the examples.

图5示出了“单根生成树”拓扑结构网络示例。在本示例中，包含六台设备A-F。图6示出了以图5所示拓扑为例实现根据本发明的”源路径回溯”的流程图。FIG. 5 shows an example of a "single-root spanning tree" topology network. In this example, there are six devices A-F. FIG. 6 shows a flow chart of implementing "source path backtracking" according to the present invention by taking the topology shown in FIG. 5 as an example.

601  管理节点周期性(周期长度可变)通过所有端口发送拓扑查询报文，查询第一跳邻接节点。报文内容包括：管理节点，下一节点(可含多项)，跳数。在本示例中：管理节点＝A，下一节点＝广播(查询)，跳数＝1。601 The management node periodically (the cycle length is variable) sends topology query messages through all ports to query the first-hop adjacent nodes. The content of the message includes: the management node, the next node (can contain multiple items), and the number of hops. In this example: management node=A, next node=broadcast (query), hop number=1.

602  第1跳邻接节点应答，应答报文内容包括：管理节点，下一节点(可含多项)，跳数。在本示例中：管理节点＝A，下一节点＝B，跳数＝1。602 The first hop adjacent node responds, and the content of the response message includes: the management node, the next node (can contain multiple items), and the number of hops. In this example: management node=A, next node=B, hop number=1.

603  第1跳邻接节点应答，报文内容包括：管理节点，下一节点(可含多项)，跳数。在本示例中：管理节点＝A，下一节点＝C，跳数＝1。603 The first hop adjacent node responds, and the content of the message includes: the management node, the next node (can contain multiple items), and the number of hops. In this example: management node=A, next node=C, hop number=1.

604  管理节点得到跳数＝1时网络拓扑结构数据库。604 The management node obtains the network topology database when the number of hops=1.

605  管理节点向跳数＝1时网络拓扑中所有设备分别发送拓扑查询报文，报文内容包括：管理节点，下一节点(可含多项)，跳数。在本示例中：管理节点＝A，下一节点＝B，下一节点＝广播(查询)，跳数＝2。605 The management node sends a topology query message to all devices in the network topology when the number of hops = 1. The content of the message includes: management node, next node (can contain multiple items), and hop number. In this example: management node=A, next node=B, next node=broadcast (query), hop number=2.

606  管理节点向跳数＝1时网络拓扑中所有设备分别发送拓扑查询报文，报文内容包括：管理节点，下一节点(可含多项)，跳数。在本示例中：管理节点＝A，下一节点＝C，下一节点＝广播(查询)，跳数＝2。606 The management node sends a topology query message to all devices in the network topology when the number of hops = 1. The content of the message includes: management node, next node (can contain multiple items), and hop number. In this example: management node=A, next node=C, next node=broadcast (query), hop number=2.

607  中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至除接收报文外所有端口。报文内容不变。在本示例中：管理节点＝A，下一节点＝B，下一节点＝广播(查询)，跳数＝2。607 The relay node forwards the message according to the information of the next node in the message content, and forwards the message to all ports except the received message. The content of the message remains unchanged. In this example: management node=A, next node=B, next node=broadcast (query), hop number=2.

608  中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至除接收报文外所有端口。报文内容不变。在本示例中：管理节点＝A，下一节点＝C，下一节点＝广播(查询)，跳数＝2。608 The relay node forwards the message according to the information of the next node in the message content, and forwards the message to all ports except the received message. The content of the message remains unchanged. In this example: management node=A, next node=C, next node=broadcast (query), hop number=2.

609  第2跳邻接节点应答，报文内容包括：管理节点，下一节点(可含多项)，跳数。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至查询报文接收端口。在本示例中：管理节点＝A，下一节点＝B，下一节点＝D，跳数＝2。609 The second-hop adjacent node responds, and the message content includes: management node, next node (can contain multiple items), and hop count. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to the receiving port of the query message. In this example: management node=A, next node=B, next node=D, hop number=2.

610  第2跳邻接节点应答，报文内容包括：管理节点，下一节点(可含多项)，跳数。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至查询报文接收端口。在本示例中：管理节点＝A，下一节点＝B，下一节点＝E，跳数＝2。610 The second hop adjacent node responds, and the message content includes: management node, next node (can contain multiple items), and hop count. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to the receiving port of the query message. In this example: management node=A, next node=B, next node=E, hop number=2.

611  第2跳邻接节点应答，报文内容包括：管理节点，下一节点(可含多项)，跳数。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至查询报文接收端口。在本示例中：管理节点＝A，下一节点＝C，下一节点＝F，跳数＝2。611 The second-hop adjacent node responds, and the message content includes: management node, next node (can contain multiple items), and hop count. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to the receiving port of the query message. In this example: management node=A, next node=C, next node=F, number of hops=2.

612  管理节点得到跳数＝2时网络拓扑结构数据库。612 The management node obtains the network topology database when the number of hops=2.

613  类推，管理节点向跳数＝2时网络拓扑中所有设备分别发送拓扑查询报文。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至除接收报文外所有端口。在本示例中：管理节点＝A，下一节点＝B，下一节点＝D，下一节点＝广播(查询)，跳数＝3。613 By analogy, the management node sends topology query messages to all devices in the network topology when the number of hops = 2. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to all ports except the receiving message. In this example: management node=A, next node=B, next node=D, next node=broadcast (query), hop number=3.

614  类推，管理节点向跳数＝2时网络拓扑中所有设备分别发送拓扑查询报文。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至除接收报文外所有端口。在本示例中：管理节点＝A，下一节点＝B，下一节点＝E，下一节点＝广播(查询)，跳数＝3。614 By analogy, the management node sends topology query messages to all devices in the network topology when the number of hops = 2. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to all ports except the receiving message. In this example: management node=A, next node=B, next node=E, next node=broadcast (query), hop number=3.

615  类推，管理节点向跳数＝2时网络拓扑中所有设备分别发送拓扑查询报文。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至除接收报文外所有端口。在本示例中：管理节点＝A，下一节点＝C，下一节点＝F，下一节点＝广播(查询)，跳数＝3。6.15 By analogy, the management node sends topology query messages to all devices in the network topology when the number of hops = 2. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to all ports except the receiving message. In this example: management node=A, next node=C, next node=F, next node=broadcast (query), hop number=3.

616  第2跳邻接节点应答，无下一节点，报文内容包括：管理节点，下一节点(可含多项)，跳数。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至查询报文接收端口。在本示例中：管理节点＝A，下一节点＝B，下一节点＝D，跳数＝2。616 The second hop adjacent node responds, but there is no next node, and the message content includes: management node, next node (can contain multiple items), and hop count. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to the receiving port of the query message. In this example: management node=A, next node=B, next node=D, hop number=2.

617  第2跳邻接节点应答，无下一节点，报文内容包括：管理节点，下一节点(可含多项)，跳数。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至查询报文接收端口。在本示例中：管理节点＝A，下一节点＝B，下一节点＝E，跳数＝2。617 The second hop adjacent node responds, but there is no next node, and the message content includes: management node, next node (can contain multiple items), and hop count. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to the receiving port of the query message. In this example: management node=A, next node=B, next node=E, hop number=2.

618  第2跳邻接节点应答，无下一节点，报文内容包括：管理节点，下一节点(可含多项)，跳数。中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至查询报文接收端口。在本示例中：管理节点＝A，下一节点＝C，下一节点＝F，跳数＝2。618 The 2nd hop adjacent node responds, there is no next node, and the content of the message includes: management node, next node (can contain multiple items), and hop count. The relay node forwards the message according to the information of the next node in the message content, and forwards the message to the receiving port of the query message. In this example: management node=A, next node=C, next node=F, number of hops=2.

619  管理节点得到最大跳数＝2信息，网络拓扑结构数据库构造完毕，停止发送拓扑查询报文。619 The management node gets the information that the maximum number of hops = 2, the network topology database is constructed, and stops sending topology query messages.

在上述步骤中，由资源受限设备构成的网络中任一设备均可以作为管理节点，发起查询，实时地获得网络拓扑结构。该拓扑结构可以按照图、表等形式显示。In the above steps, any device in the network composed of resource-limited devices can act as a management node, initiate a query, and obtain the network topology in real time. The topology structure may be displayed in the form of a graph, a table, or the like.

报文内容中的“下一节点”项应该是能够唯一标识网络中特定节点的标识符，节点标识符可以是MAC地址，或产品编号等。当查询跳数是M时，M是大于等于1的正整数，将处于层次M的所有节点称作M跳节点，由于第M跳的设备未知，查询报文中的路径信息中可以包括第1到第M-1跳的信息，也可以用约定的特殊标识符来代替第M跳的节点标识符，如0×0，但该特殊标识符必须被“资源受限设备构成的网络”内所有设备知晓，这两种方式实质上是相同的。The "next node" item in the message content should be an identifier that can uniquely identify a specific node in the network, and the node identifier can be a MAC address or a product number. When the number of query hops is M, M is a positive integer greater than or equal to 1, and all nodes at level M are called M-hop nodes. Since the device at the M-th hop is unknown, the path information in the query message may include the first For the information to the M-1th hop, the agreed special identifier can also be used to replace the node identifier of the Mth hop, such as 0×0, but this special identifier must be all The device is aware that the two approaches are essentially the same.

中继节点通过公知方法得到查询报文接收端口号、下一节点所在端口号，例如硬件转发表、软件缓存信息表等。The relay node obtains the receiving port number of the query message and the port number of the next node through known methods, such as a hardware forwarding table, a software cache information table, and the like.

此外，被查询节点可以在查询应答报文中承载相邻节点的节点标识符。据此信息，管理节点可以在树状拓扑的基础上进一步优化，得到实际网络拓扑。In addition, the queried node may carry the node identifier of the neighboring node in the query response message. Based on this information, the management node can be further optimized on the basis of the tree topology to obtain the actual network topology.

图7示出了根据图6所示方法获得的拓扑信息来利用“虚拟管理通路”进行配置管理的流程图。FIG. 7 shows a flow chart of using "virtual management path" to perform configuration management according to the topology information obtained by the method shown in FIG. 6 .

701管理节点配置被管理节点E。查询网络拓扑结构数据库，根据拓扑路径构造管理命令报文。报文内容包括：管理节点，下一节点(可含多项)，跳数，管理命令。701 The management node configures the managed node E. Query the network topology database, and construct management command messages according to the topology path. The content of the message includes: management node, next node (may include multiple items), hop count, and management command.

702  中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至下一节点。依次类推，逐跳转发。报文内容包括：管理节点，下一节点(可含多项)，跳数，管理命令。702 The relay node forwards the message according to the next node information in the message content, and forwards the message to the next node. And so on, hop-by-hop forwarding. The content of the message includes: management node, next node (may include multiple items), hop count, and management command.

703  被管理节点收到管理命令报文，读出命令字符串，传递给控制台模块直接执行。得到执行结果。将执行结果字符串流封装后发送管理应答报文。报文内容包括：管理节点，下一节点(可含多项)，跳数，命令执行结果。703 The managed node receives the management command message, reads out the command string, and passes it to the console module for direct execution. Get the execution result. Encapsulate the execution result string stream and send the management response message. The content of the message includes: the management node, the next node (can contain multiple items), the number of hops, and the command execution result.

704  中继节点按照报文内容中下一节点信息进行报文转发，将报文转发至下一节点。依次类推，逐跳转发。管理节点收到报文后，读出字符串，直接显示在本地控制台上。报文内容包括：管理节点，下一节点(可含多项)，跳数，命令执行结果。704 The relay node forwards the message according to the next node information in the message content, and forwards the message to the next node. And so on, hop-by-hop forwarding. After the management node receives the message, it reads out the character string and displays it directly on the local console. The content of the message includes: the management node, the next node (can contain multiple items), the number of hops, and the command execution result.

同样，在上述步骤中，在“资源受限设备构成的网络”中任一设备均可以作为管理节点，通过指定路径，向指定节点，发送配置管理命令。配置结果将通过指定路径，反向回溯至管理节点的控制台显示，由此实现了远程配置管理。管理命令和命令执行结果可以基于字符串进行设计。Similarly, in the above steps, any device in the "network composed of resource-limited devices" can serve as a management node, and send configuration management commands to a designated node through a designated path. The configuration result will be traced back to the console display of the management node through the specified path, thus realizing remote configuration management. Management commands and command execution results can be designed based on character strings.

如上所述，本发明的网络设备管理方法不依赖网络层及网络层以上协议，设备可以不实现网络层及网络层以上协议，上述用于拓扑查询、拓扑应答、管理命令或管理应答的报文均为链路层协议报文，设备间仅通过链路层协议报文交互就可实现上述功能，可以通过报文的类型字段来定义报文类型。由中继节点执行转发，不需要维护状态机、数据库等。由此，降低软件复杂度，减少硬件资源要求，有效降低设备成本。As mentioned above, the network device management method of the present invention does not depend on the network layer and above-network layer protocols, and the device may not implement the network layer and above-network layer protocols. The above-mentioned messages used for topology query, topology response, management command or management response All are link layer protocol messages, and the above functions can be realized only through the exchange of link layer protocol messages between devices, and the message type can be defined through the type field of the message. The forwarding is performed by the relay node, and there is no need to maintain state machines, databases, etc. Therefore, software complexity is reduced, hardware resource requirements are reduced, and equipment costs are effectively reduced.

管理节点可以开发基于本发明的网络管理软件。该软件通过”源路径回溯”发现并显示拓扑结构。通过“虚拟管理通路”实现配置管理。网络管理软件中每一个GUI配置动作，对应一个或多个基于字符的控制台命令。因此，网络管理软件升级，不需要修改网络设备内嵌软件模块。The management node can develop the network management software based on the present invention. The software discovers and displays topologies through "source path backtracking". Configuration management is implemented through a "virtual management channel". Each GUI configuration action in the network management software corresponds to one or more character-based console commands. Therefore, the network management software upgrade does not need to modify the embedded software modules of the network equipment.

本发明可以应用于任意类型“资源受限设备构成的网络”中。下面将参考图8和图9，对本发明方法的具体应用进行说明，其中图8和图9分别示出了应用本发明方法的用户驻地网和智能家电网络的示例。The present invention can be applied to any type of "network of resource-constrained devices". The specific application of the method of the present invention will be described below with reference to Fig. 8 and Fig. 9, wherein Fig. 8 and Fig. 9 respectively show examples of a customer premises network and a smart home appliance network applying the method of the present invention.

用户驻地网由汇聚交换机(804)、楼道交换机(808)、以太网络(807)构成。根据本发明，需要楼道交换机和汇聚交换机均支持本发明，如此控制台(801)可以直接通过和汇聚交换机(804)的串口电缆连接，直接针对任意一台楼道交换机(808)、汇聚交换机(804)，进行控制台配置管理。如果控制台(801)能够接入Internet，且支持Telnet或基于Web及本发明的设备管理程序等，则控制台(806)可以通过远程网管，查看网络拓扑及配置管理任意一台楼道交换机(808)、汇聚交换机(804)。楼道交换机(808)、汇聚交换机(804)不需支持TCP/IP等高层协议。The customer premises network is composed of a convergence switch (804), a corridor switch (808), and an Ethernet network (807). According to the present invention, both the corridor switch and the aggregation switch need to support the present invention, so that the console (801) can be directly connected with the serial port cable of the aggregation switch (804), directly for any corridor switch (808), aggregation switch (804) ) for console configuration management. If the console (801) can access the Internet, and supports Telnet or based on the Web and the equipment management program of the present invention, etc., then the console (806) can check the network topology and configure and manage any corridor switch (808) by remote network management. ), aggregation switch (804). Corridor switches (808) and aggregation switches (804) do not need to support high-level protocols such as TCP/IP.

智能家电网络由智能家电(901)、家用电脑(902)、家庭网络(905)构成。根据本发明，需要智能家电(901)、家用电脑(902)均支持本发明，由此家用电脑(902)可以直接通过本地控制台或者基于本发明方法的图形管理程序，针对任意一台智能家电(901)，进行配置管理。如果家用电脑(902)能够接入Internet，且支持Telnet或基于Web及本发明的设备管理程序等，则外部主机(904)可以通过远程网管，查看网络拓扑及配置管理任意一台智能家电(901)。智能家电(901)不需支持TCP/IP等高层协议。The smart home appliance network consists of smart home appliances (901), home computers (902), and home networks (905). According to the present invention, both the smart home appliance (901) and the home computer (902) need to support the present invention, so the home computer (902) can directly use the local console or the graphics management program based on the method of the present invention to target any smart home appliance (901), performing configuration management. If the home computer (902) can access the Internet, and supports Telnet or the device management program based on the Web and the present invention, etc., then the external host (904) can check the network topology and configure and manage any smart home appliance (901) through the remote network management. ). Smart home appliances (901) do not need to support high-level protocols such as TCP/IP.

Claims (6)

Translated fromChinese

1.一种由资源受限设备构成的网络的管理方法，包括步骤：1. A method for managing a network composed of resource-constrained devices, comprising the steps of:a)确定所述网络是树形拓扑；a) determining that the network is a tree topology;b)在所述网络中，设备之间通过链路层协议报文交互来进行网络拓扑发现，以搜集网络拓扑信息，构造网络拓扑数据信息；以及b) In the network, network topology discovery is performed between devices through link layer protocol message exchange to collect network topology information and construct network topology data information; andc)基于所述网络拓扑数据信息，选择指定设备，通过链路层协议报文交互管理信息，来进行配置管理。c) Select a designated device based on the network topology data information, and perform configuration management by exchanging management information through link layer protocol packets.2.根据权利要求1所述的方法，其特征在于所述链路层协议报文包括以下内容：查询跳数M，其中M是大于等于1的正整数、管理节点标识符、源路径第1到第M跳节点标识符；2. The method according to claim 1, wherein the link layer protocol message includes the following content: query hop count M, wherein M is a positive integer greater than or equal to 1, management node identifier, source path No. 1 to the Mth hop node identifier;报文类型标志位，包括类型：拓扑查询、拓扑应答、管理命令或管理应答；Message type flag, including type: topology query, topology response, management command or management response;数据载荷长度和数据载荷。Data payload length and data payload.3.根据权利要求1所述的方法，其特征在于：3. The method according to claim 1, characterized in that:在所述网络中使用能够被网络内所有设备知晓的节点标识符来唯一地标识一台设备。A device is uniquely identified in the network using a node identifier that can be known by all devices in the network.4.根据权利要求3所述的方法，其特征在于：4. The method according to claim 3, characterized in that:所述节点标识符是MAC地址、产品编号或约定的其他标识符。The node identifier is a MAC address, a product number or other agreed identifiers.5.根据权利要求1或2所述的方法，其特征在于所述步骤b)包括以下步骤：5. The method according to claim 1 or 2, characterized in that said step b) comprises the following steps:管理节点以预定周期发起拓扑查询；The management node initiates a topology query at a predetermined period;将所述管理节点作为树形拓扑的根，树的深度是N，按照树的层次，Taking the management node as the root of the tree topology, the depth of the tree is N, according to the hierarchy of the tree,从1开始递增进行分别对应于针对第1到第N跳节点的拓扑查询；Incrementing from 1 respectively corresponds to the topology query for the 1st to Nth hop nodes;对于层次M的查询，1≤M≤N，管理节点构造针对M跳节点的、报文类型标志位是拓扑查询的链路层协议报文，该报文包括从管理节点到达各个节点路径的已知路径信息，管理节点根据所述路径信息选择发送端口，非管理节点收到报文后，进行判断：For the query of level M, 1≤M≤N, the management node constructs a link layer protocol message aimed at the M-hop node, and the message type flag bit is the topology query. Knowing the path information, the management node selects the sending port according to the path information, and the non-management node makes a judgment after receiving the message:如果所述非管理节点不是M跳节点，则If the non-management node is not an M-hop node, then当其不是第M-1跳时，将报文转发报文至下一跳，When it is not the M-1th hop, forward the message to the next hop,当其是第M-1跳时，将查询报文发送至除查询报文接收端口之外的所有端口，如果没有应答，则构造报文类型是查询应答的链路层协议报文，回答管理节点，When it is the M-1th hop, send the query message to all ports except the receiving port of the query message, if there is no response, construct a link layer protocol message whose message type is a query response, and answer the management node,如果所述非管理节点是M跳节点时，则构造应答报文，回答管理节点。If the non-management node is an M-hop node, construct a response message to reply to the management node.6.根据权利要求1或2所述的方法，其特征在于：所述步骤c)包括以下步骤：6. The method according to claim 1 or 2, characterized in that: said step c) comprises the following steps:管理节点根据利用步骤b)获得的网络拓扑数据，得到到达被管理节点的路径信息，构造标志位是“管理命令”的链路层协议报文，并根据路径信息发送所述报文，报文内容包括到达被管理节点的路径信息，报文数据载荷域承载管理节点发送往被管理节点的管理命令；The management node obtains the path information to the managed node according to the network topology data obtained in step b), constructs a link layer protocol message whose flag is a "management command", and sends the message according to the path information, the message The content includes the path information to the managed node, and the message data payload field carries the management command sent by the management node to the managed node;收到报文的节点根据报文中的路径信息进行判断：The node receiving the message judges according to the path information in the message:如果其不是报文中指定的被管理节点、管理节点，则将报文转发给报文中指定的下一节点，If it is not the managed node or management node specified in the message, the message will be forwarded to the next node specified in the message,如果其是报文中指定的被管理节点，则所述被管理节点解析报文数据载荷，执行所述管理命令，并根据执行结果构造标志位是“管理应答”的链路层协议报文进行应答，报文内容包括到达管理节点路径等信息，报文数据载荷域承载发送往管理节点的被管理节点管理命令执行结果。If it is the managed node specified in the message, the managed node parses the message data load, executes the management command, and constructs a link layer protocol message whose flag is "management response" according to the execution result. Response, the content of the message includes information such as the path to the management node, and the data payload field of the message carries the execution result of the managed node management command sent to the management node.

Images