Ethernet is the most common communication protocol standard used in existing local area networks today.

This standard defines the cable types and signal processing methods used in local area networks (LANs).

Ethernet transmits information packets between interconnected devices at a rate of 10~100Mbps. Twisted pair cable 10 Base T Ethernet has become the most widely used Ethernet technology due to its low cost, high reliability and 10Mbps rate.

The direct expansion wireless Ethernet can reach 11Mbps, and the products provided by many manufacturing suppliers can use common software protocols for communication, with the best openness.

△ Ethernet connection topology: Bus type: requires fewer cables, is cheap, has high management costs, is not easy to isolate fault points, and uses a shared access mechanism, which can easily cause network congestion.

In the early days, Ethernet mostly used a bus topology, using coaxial cables as the transmission medium. The connections were simple and usually did not require dedicated network equipment in small-scale networks. However, due to its inherent flaws, it has gradually been replaced by hubs and hubs.

The switches are replaced by a core star network.

Star: easy to manage, easy to expand, requires dedicated network equipment as the core node of the network, requires more network cables, and has high requirements for the reliability of the core equipment.

A dedicated network device (such as a hub or switch) is used as the core node, and each host in the LAN is connected to the core node through twisted pairs, forming a star structure.

Although the star network requires more cables than the bus type, the wiring and connectors are cheaper than the bus type.

In addition, the star topology can easily expand the network to a large scale through cascading, so it has been widely used and is used by most Ethernet networks.

Transmission media: Ethernet can use a variety of connection media, including coaxial cables, twisted pairs, and optical fibers.

Among them, twisted pairs are mostly used for connections from hosts to hubs or switches, while optical fibers are mainly used for cascading between switches and point-to-point links between switches and routers.

Coaxial cable, as the main connection medium in the early days, has gradually been phased out.

Working mode of the interface: The Ethernet card can work in two modes: half-duplex and full-duplex.

Half-duplex: Half-duplex transmission mode implements Ethernet carrier monitoring multi-access conflict detection.

Traditional shared LAN works under half-duplex and can only transmit data in one direction at the same time.

When data in both directions is transmitted at the same time, conflicts occur, which reduces the efficiency of the Ethernet network.

Full-duplex: Full-duplex transmission uses point-to-point connections. There is no conflict in this arrangement because they use two independent lines in the twisted pair, which increases the bandwidth without installing new media.

For example, a parallel railway track is added between the stations in the above example, so that two trains can pass in both directions at the same time.

In Duplex mode, the collision detection circuit is unavailable, so only one port per Duplex connection is used for point-to-point connections.

The transmission efficiency of standard Ethernet can reach 50% to 60% of the bandwidth, and duplex provides 100% efficiency in both directions.

△ How Ethernet works Ethernet uses the Carrier Frame Listening Multiple Access (CSMA/CD) mechanism with collision detection.

All nodes in Ethernet can see all information sent in the network. Therefore, we say that Ethernet is a broadcast network.

The working process of Ethernet is as follows: When a host in the Ethernet wants to transmit data, it will proceed as follows: 1. Check whether there is a signal being transmitted on the frame listening channel.

If there is, it indicates that the channel is busy, and the frame listening will continue until the channel is idle.

2. If no frame hears any signal, transmit the data. 3. Continue to listen to the frame during transmission. If a conflict is found, perform the backoff algorithm. After a random wait for a period of time, re-execute step 1 (when a conflict occurs, the computer involved in the conflict

A congestion sequence will be sent to warn all nodes) 4. If no conflict is found, the transmission is successful, and the computer will return to the frame listening channel state.

Note: Each computer is only allowed to send one packet at a time, and all computers must wait 9.6 microseconds after the last transmission (running at 10Mbps) before attempting to send data again.

△ Frame structure Overview of Ethernet frames: Ethernet frames are encapsulated by the data link layer. Data packets at the network layer are added with a frame header and frame tail to become a data frame (framing) that can be recognized by the data link layer.

Although the number of bytes used in the frame header and frame trailer is fixed, the length of the Ethernet varies depending on the size of the encapsulated data packet, ranging from 64 to 1518 bytes (not counting 8 bytes).

the leading word).

△ Conflict/Collision Domain Collision: In Ethernet, when two data frames are sent to the physical transmission medium at the same time and completely or partially overlap, a data conflict occurs.

When a conflict occurs, the data on the physical network segment is no longer valid.