Packet exchange

Introduction

Datagram packet switching technology means that there must be at least one data transmission path between the two communication parties. These paths may span multiple intermediate nodes, and the source node will transmit the data before communication. The data packet to be exchanged is prepared, and finally transmitted and exchanged in the form of packets. If the source and the sink are adjacent nodes, the source can directly deliver the data to the sink. If the source and sink are connected through a central node, the source will pass the packet to the appropriate intermediate node through a suitable routing mechanism, and the intermediate node will select the appropriate path to transfer the packet data to the sink after several routing selections.

Connectionless network

Data packet switching is a network architecture with fault tolerance. In order to solve the problem of fault tolerance, the network has developed into a packet-switched connectionless network. In a packet-switched network, a single message is divided into multiple data blocks. These data blocks are called packets, which contain the address information of the sender and receiver, are transmitted in one or more networks, and are recombined at the destination.

The transmission of these packets are independent of each other, do not affect each other, and usually follow different routes to reach the destination. The message is usually divided into thousands of packets, and usually some of them are lost in transmission. The protocol allows this to happen, and includes a method of requiring retransmission of data packets that were lost in transmission.

Packet switching technology is connectionless, because it does not need to establish a dynamic connection for the call. This is more efficient than a circuit-switched network because multiple users can use the network circuit. Packet switching technology is fault-tolerant because it avoids the danger of relying on a single circuit to provide service reliability. If one network path fails, the other lines can guarantee transmission, thus making the message complete.

Packet switching is the standard Internet, but circuit-switched networks still have a portion of the market share. Modern circuit networks allow circuit failures and session replies, and some consumers like the reliability and safety of modern dedicated circuits. But circuit-switched connections are more expensive than data packet-switched connections, but many organizations need this continuously effective and safe circuit and are willing to pay extra.

Main features

1. Different groups of the same message can pass through the communication subnet by different transmission paths, just like there are multiple traffic lines between two places, any one of them will be selected. Can successfully reach the destination.

2. When different packets of the same message arrive at the destination node, disorder, duplication and loss may occur.

3. Each packet must have a destination address and a source address during the transmission process for the routing of the intermediate node, that is, each packet is routed and forwarded at the intermediate node. The purpose of adding the destination address is to transfer the data to the destination, and adding the source address is to facilitate the return of the error content to the source after a packet error, so that it can be sent again.

4. The datagram packet mode has a large transmission delay, which is suitable for burst communication, but not suitable for long message and conversational communication.

Data packet switching in Ethernet

The Ethernet switch performs data packet switching based on the destination MAC address of the data packet. There is a table of MAC address and port correspondence in the switch, this table is the MAC address table.

The Ethernet switch is connected to the PC. When the port of the switch receives the first data packet of the data stream, it reads the source MAC address field of the data packet, associates the source MAC address with the receiving port, and records To the MAC address table. The MAC address table is stored in the memory of the switch, and the MAC address table will be cleared when the switch restarts.

Assuming that PC1, PC2, and PC3 are directly connected to the switch to form a simple network, the MAC addresses of the PCs are respectively,

PC1: 0011.ABCD.0001

PC2:0022.ABCD.0002

PC3:0033.ADCB.0003

If the switch receives data packets on all three ports, it will learn as follows The MAC address table shown in the table.

The processing procedure after the switch receives a data packet on a port is as follows:

The switch first determines the purpose of the data packet Whether the MAC address is a broadcast or multicast address, if it is, the flooding operation is performed.

If the destination MAC address is not a broadcast or multicast address but a unicast address to a device, the switch will look up this address in the MAC address table. If the address is unknown, it will also follow the flooding The way to forward.

If the destination address is a unicast address and it already exists in the MAC address table of the switch, the switch will forward the data packet to the port associated with the destination MAC address.