"Computer Network" Computer Network Architecture

1. Classification of computer networks

1.1 Classification by distribution range

• Wide Area Network (WAN)

  Long-distance communications, usually covering areas ranging from tens of kilometers to thousands of kilometers

• Metropolitan Area Network (MAN)

  The coverage can span several blocks or even the entire city, and the coverage is about 5~50 km

  Metropolitan area networks mostly use Ethernet technology

• Local Area Network (LAN)

  Generally, a microcomputer or workstation is connected by a high-speed line, and the coverage area is usually tens of meters to several kilometers.

  There are not too many restrictions on the number of computer configurations in the local area network

  Traditionally, local area networks use broadcast technology and wide area networks use switching technology

• Personal Area Network (PAN)

  Refers to a network that connects consumer electronic devices (such as tablets, smart phones, etc.) with wireless technology in the personal work place, covering a diameter of about 10 m

1.2 Classification by transmission technology

• Broadcast network

  All networked computers share a common communication channel.

  When a computer uses the shared communication channel to send a packet of messages, all other computers will "listen" to the packet. The computer that receives the packet will determine whether to receive the packet by checking the destination address.

  Basically, the local area network uses broadcast communication technology , and the wireless and satellite communication networks in the wide area network also use broadcast communication technology.

• Peer-to-peer network

  Each physical line is connected to a pair of computers. If there is no direct line, it will receive, store and forward through the nodes in between, until the destination node

1.3 Classification by topology

Network topology refers to the network structure represented by the geometric relationship (such as bus type, ring type) between nodes (routers, hosts, etc.) and communication lines (network lines) in the network, mainly refers to the topology of the communication subnet

• Bus network

Connect the computer with a single transmission line

The advantage is that it is easy to build a network, convenient to increase or decrease nodes, and save lines

The disadvantage is that the communication efficiency is not high when the load is heavy, and any part of the bus is sensitive to faults.

• Star network

  

  Each terminal or computer is connected to the central equipment by a separate line, generally the central equipment is a switch or router

  The advantage is that it is convenient for centralized control and management

  The disadvantage is that the cost is high and the central node is sensitive to faults

• Ring network

  

  All computer interface devices are connected into a ring, the most typical example is a token ring LAN

  The ring can be a single ring or a double ring, and the signal in the ring is transmitted in one direction

• Mesh type

  

  In general, each node has at least two paths connected to other nodes, which are mostly used in wide area networks. There are two types: regular and irregular.

  The advantage is high reliability

  The disadvantage is that the control is complicated and the line cost is high

1.4 Classification by exchange technology

The exchange technology refers to the data format and the way of the exchange device used to exchange information between each host, between each communication device, or between the host and the communication device.

• Circuit switched network

  A dedicated path is established between the source node and the destination node for data transmission, including three stages: connection establishment , data transmission and disconnection . The most typical circuit-switched network is the traditional telephone network.

  The main feature of this type of network is that the bit stream of the entire message continuously goes from the source point to the end point, as if it is transmitted in a pipe

  The advantage is that the data is directly transmitted and the delay is small

  The disadvantage is that the line utilization rate is low, the line capacity cannot be fully utilized, and the error control is inconvenient.

• Message switching network

  User data plus source address, destination address, check code and other auxiliary information, and then encapsulated into a message

  The entire message is transmitted to the neighboring node, and after all is stored, it is forwarded to the next node, and this process is repeated until it reaches the destination node.

  Each message can be individually selected to reach the destination 3. Node path

  The advantage is that line capacity can be fully utilized, different data rates can be converted between different links, format conversion can be realized, one-to-many and many-to-one access can be realized, and error control can be realized

  The disadvantage is that it increases resource overhead and buffer delay, and requires additional control mechanisms to ensure that the order of multiple messages is not out of order, and the buffer is difficult to manage

• Packet switched network

  Divide the data into shorter fixed-level data blocks, add auxiliary information such as destination address and source address to each data block to form packets (packets), and transmit them in a store-and-forward manner

  Its main feature is that a single packet (it is only a part of the entire message) is transmitted to the adjacent node, stored and then searched for the forwarding table, and forwarded to the next node .

  In addition to the advantages of a message switching network, it also has its own advantages: easy to manage the buffer; the average delay of packets is smaller, and the average buffer occupied by the network is less; easier to standardize; more suitable for applications

  Today's mainstream networks can basically be regarded as packet-switched networks

2. OSI Reference Model

• Physical layer

  The unit of transmission is bit , and the task is to transparently transmit the bit stream

  The physical layer mainly defines the physical and logical connection methods of data terminal equipment (DTE) and data communication equipment (DCE)

  The physical layer specifies the circuit interface (such as the size of the mechanical hard disk, the network cable interface, etc.) and the meaning and electrical characteristics of the signal transmitted on the communication link (for example, it is specified that the signal A represents the number 0, and the transmission signal A only needs to send the number 0)

• data link layer

  The transmission unit is a frame , and the task is to assemble the IP datagram from the network layer into a frame

  The functions of the data link layer can be summarized as framing, error control, flow control and transmission management, etc.

• Network layer

  The transmission unit is the datagram . It is concerned with the operation control of the communication subnet. The main task is to transfer the protocol data unit (packet) of the network layer from the source to the destination, and provide communication services for different hosts on the packet switching network.

  The key issue is to route packets

• Transport layer

  The unit of transmission is packet segment (TCP) or user datagram (UDP). The transport layer is responsible for the communication between the two processes in the host, and the function is to provide reliable transmission services for end-to-end connections.

• Session layer

  The session layer allows sessions between processes on different hosts

  The session layer uses the end-to-end services provided by the transport layer to provide its value-added services to the presentation layer. This service is mainly used to establish connections for presentation layer entities or user processes and to transfer data in an orderly manner on the connection. This is the session

• Presentation layer

  The presentation layer mainly deals with the way of exchanging information in two communication systems

• Application layer

  The application layer provides a means for specific types of network applications to access the OSI environment

  The application layer uses different application protocols to solve different types of application requirements, such as HTTP, FTP, SMTP, etc.

3. TCP/IP model

The TCP/IP model simplifies the OSI reference model into four layers, and the main protocols of each layer are as follows:

• Transmission Control Protocol (TCP)

  It is connection-oriented, and the unit of data transmission is segment, which can provide reliable delivery

• User Datagram Protocol (UDP)

  It is connectionless, and the unit of data transmission is user datagrams. Reliable delivery is not guaranteed, only "best effort delivery" is provided.