Ethernet

Ethernet is a widely used communication technology that connects devices such as computers, switches, and routers into a local area network (LAN). It enables data transmission between connected devices and forms the foundation of office, home, and even industrial networks. The connection between devices can be wired or wireless (WiFi). It is defined by the IEEE 802.3 standard. Ethernet itself is easy to understand, implement, and maintain. It offers flexibility in choosing the network topology. The topology of Ethernet is generally bus-based, but with the use of switches and routers, it transforms into a star topology. The IEEE 802.3 standard defines the physical and data link layers of the OSI model.

Physical layer of Ethernet
For fixed device connections, copper or fiber optic cables can be used. With physical cables, the maximum distance between devices is 100 meters, but when using fiber optics, the distance increases significantly. Wireless data transmission is possible via WiFi.
Copper cables (twisted pair, RJ45 connectors)

  • Cat5e: up to 1 Gbps over 100 m
  • Cat6/Cat6a: up to 10 Gbps over 100 m
  • Cat7/Cat8: higher speeds over shorter distances

Optical cables

  • Multimode fiber: shorter distances (up to 550 m)
  • Singlemode fiber: long distances (around 40 km)

Wireless transmissions

  • WiFi – IEEE 802.11 standard

The typical encoding used in Ethernet is Manchester encoding. This encoding is based on a change in signal level, where the logical value is determined by the rising or falling edge of the signal. Ethernet allows the creation of a star topology network using switches. An overview of Ethernet standards can be seen in the table.

Speed Common Name Informal IEEE Standard name  Formal IEEE Standard name
10 Mbps Ethernet 10BASE-T 802.3
100 Mbps FastEthernet 100BASE-T 802.3u
1 Gbps GigabitEthernet 1000BASE-T 802.3ab
10 Gbps 10 GE 10GBASE-T 802.3an
40 Gbps 40 GE 40GBASE-T -

Data link layer of Ethernet
This layer of the model defines data transmission in the network. It specifies the frame structure, addressing, and access methods. The image shows the structure of an Ethernet frame.

Where is: 
Preamble: 7 bytes with an alternating bit pattern of ones and zeros, used for synchronization
SFD (Start Frame Delimiter): 1 byte with the bit pattern 10101011, marking the start of the frame
Destination Address: 6 bytes, the MAC address of the device to which the message is being sent
Source Address: 6 bytes, the MAC address of the device sending the message
Length: 2 bytes that indicate the total length of the Ethernet frame. The maximum frame length is limited to 1500 bytes
Data: 46 to 1500 bytes containing the transmitted data
CRC (Cyclic Redundancy Check): 4 bytes used to detect transmission errors

Arbitration in an Ethernet network
This process determines which device is allowed to transmit at a given moment. Ethernet uses the CSMA/CD mechanism (Carrier Sense Multiple Access with Collision Detection), which minimizes collisions. This method is primarily used in half-duplex Ethernet communication.
1. CS (Carrier Sense)
• Before starting transmission, the device monitors the transmission medium to determine whether it is free.
• If the medium is busy, it waits until it becomes available and then begins transmitting.
2.MA (Multiple Access)
• Multiple stations have the ability to access the same medium and can begin transmitting when the medium is free.
3.CD (Collision Detection)
• A collision is detected by comparing the transmitted and received signal.
• If the received signal differs from the transmitted one, a collision has occurred.
When a collision is detected, the device immediately stops transmitting and sends a special jam signal (a 32-bit pattern) to inform other devices on the network about the collision. After sending the jam signal, the station waits for a random amount of time before attempting to retransmit. The length of this waiting period is determined by the Backoff algorithm, and this interval increases with the number of detected collisions to reduce the likelihood of repeated collisions.

Advantages of CSMA/CD Disadvantages of CSMA/CD
Simple Implementation Less efficient under high network traffic load
Does not require complex hardware or software With multiple collisions, latency increases
Allows efficient utilization of the shared medium Limited to half-duplex mode
Automatic resolution of collisions between multiple devices Unsuitable for modern high-speed networks
Suitable for small and medium-sized networks with lower traffic load