What Is Ethernet? IEEE 802.3 LAN Technology Explained
Learn what Ethernet is, how IEEE 802.3 defines LAN media, frames, speeds, and access methods, and how modern switched Ethernet differs from legacy shared networks.
Ethernet is a family of networking standards used to connect devices on a local area network (LAN). It is not the name of one specific cable. Ethernet standards define how devices send data, what physical media and signaling they use, how data is organized into frames, and which communication rules apply.
In a typical Ethernet LAN, endpoint devices such as computers, printers, servers, and access points connect to a switch. The switch forwards Ethernet frames between devices. Ethernet is widely used in homes, small offices, data centers, campuses, and enterprise access networks.
What Is a LAN?
A local area network (LAN) connects devices within a limited geographic area, such as a home, office, school, or building. A LAN usually provides high-speed communication between nearby devices under the control of one person or organization.
An Ethernet LAN is a LAN whose devices and infrastructure communicate using Ethernet standards. A simple wired Ethernet LAN might contain a desktop computer, a printer, a wireless router or switch, and several Ethernet cables.
Ethernet describes the communication technology used on the local links. Other technologies, such as Internet Protocol, can operate above Ethernet to provide addressing and end-to-end communication across larger networks. For background on networks and their components, see Computer Network Explained.
IEEE 802.3: The Ethernet Standard Family
The Institute of Electrical and Electronics Engineers (IEEE) is a professional organization that publishes many technical standards. The Ethernet family is associated with IEEE 802.3.
IEEE 802.3 includes specifications for many Ethernet variations. A variation may define a particular data rate, cable type, signaling method, connector arrangement, maximum link distance, or other physical requirements. This is why two Ethernet links can look different while still belonging to the Ethernet family.
Common standards are important because equipment from different vendors can interoperate. For example, a network interface in one vendor's computer can communicate with a switch from another vendor when both support compatible Ethernet specifications. Without shared standards, each manufacturer could require a separate, incompatible networking method.
What Ethernet Standards Define
Ethernet covers both physical and communication details. At a high level, an Ethernet implementation defines:
- Physical media: the cable or other medium that carries signals.
- Signaling: how electrical or optical signals represent transmitted bits.
- Connectors and wiring: how a device attaches to the medium and how conductors or fibers are arranged.
- Frame format: how Ethernet data is organized into a structured unit.
- Transmission rules: how devices share a medium and send or receive data.
- Speed and distance capabilities: the data rate and link length supported by a particular implementation.
The Ethernet frame is the structured unit of data transmitted on an Ethernet network. It carries information such as source and destination hardware addresses, a payload, and error-detection information. Devices on an Ethernet LAN use frames to exchange data across a local link. Ethernet frames and MAC addresses are important parts of the data-link operation described in the OSI reference model.
Ethernet Physical Media
Ethernet has been implemented over several types of media. The medium affects installation cost, supported distance, data rate, resistance to interference, and the connectors used.
Coaxial Cable
Coaxial cable contains a central conductor, insulation, shielding, and an outer covering. Older Ethernet implementations used coaxial cable to create shared network segments. Several devices could attach to the same physical medium.
Coaxial Ethernet is now uncommon in modern LAN installations. Switched Ethernet over twisted-pair copper or fiber is easier to manage, scales better, and avoids the collision behavior associated with a shared cable.
Twisted-Pair Copper
Twisted-pair cable contains insulated copper wires arranged in pairs. The twists help reduce interference between pairs and from nearby electrical sources. It is a common choice for connecting computers, phones, printers, access points, and other devices to access switches.
In everyday conversation, people sometimes call a twisted-pair Ethernet cable an “RJ-45 cable.” More precisely, the familiar modular connector is commonly identified as an 8P8C connector, although “RJ-45” remains widely used in networking discussions. Not every Ethernet implementation uses this connector.
Fiber-Optic Cable
Fiber-optic cable carries information as pulses of light through glass or plastic fibers. Fiber is often selected for high-speed links, long runs, building-to-building connections, and environments with substantial electromagnetic interference.
Fiber can cost more to install and may require compatible optical transceivers and careful handling. Its advantages include long distance, high bandwidth, and electrical isolation. A long link that is unreliable or exceeds the practical distance of copper may be a good candidate for fiber.
Signaling, Connectors, and Wiring Variants
Ethernet variants define specific signaling and cabling requirements. A copper variant may specify the number of wire pairs, the electrical signaling method, cable characteristics, and connector type. A fiber variant may specify the optical wavelength, fiber type, transceiver characteristics, and connector arrangement.
Therefore, Ethernet ports and cables must be compatible. If a connection does not work after installing a cable, verify that both interfaces support the same Ethernet variant, that the connector fits correctly, and that the cable type and distance meet the specification. Ethernet does not mean that every cable, connector, or port is interchangeable.
Ethernet Frames and Communication Rules
When a device sends data over an Ethernet LAN, it places the data into an Ethernet frame. The frame identifies the local source and destination and provides a defined structure that receiving Ethernet devices understand.
A switch examines Ethernet frame information and forwards the frame toward the appropriate port when possible. The connected endpoint then processes the frame if it is the intended recipient. Ethernet standards also define rules for transmitting frames, detecting errors, and operating over particular physical links.
Ethernet is commonly associated with the lower layers of the networking model: physical signaling is a Layer 1 concern, while frame delivery and hardware addressing are primarily Layer 2 concerns. Ethernet does not by itself provide every function needed for communication across the Internet.
CSMA/CD and Ethernet Access
CSMA/CD stands for Carrier Sense Multiple Access with Collision Detection. It was the access method associated with shared, half-duplex Ethernet.
Its basic historical operation was:
- A device used carrier sensing to listen to the shared medium before transmitting.
- If the medium appeared idle, the device began sending.
- Multiple devices could still begin transmitting close enough together to create a collision, which is a transmission conflict on shared half-duplex media.
- The devices detected the collision, stopped transmitting, waited for different periods, and tried again.
This process allowed multiple devices to share one medium, but collisions reduced efficiency. Adding more devices or increasing traffic could increase contention.
Modern switched Ethernet normally uses separate point-to-point links between endpoints and switch ports. These links operate in full duplex, meaning a device can transmit and receive at the same time. Because the two ends are not competing for one shared collision-prone medium, collisions and CSMA/CD are generally not relevant on a correctly operating switched full-duplex link.
Ethernet Speed Evolution
Ethernet has evolved from early low-speed networks to links capable of hundreds of gigabits per second and beyond. A rate such as 1 Gbps means one gigabit per second, while Mbps means megabits per second.
These rates are representative generations, not a complete list of every Ethernet specification. The achievable rate depends on the Ethernet variant, interface hardware, cable or fiber type, link length, and configuration. A higher advertised port speed does not make an incompatible cable or transceiver suitable.
Practical Ethernet LAN Examples
Home or Small-Office Wired Network
A desktop computer and printer can connect to a small switch or wireless router using twisted-pair Ethernet cables. The devices use Ethernet for their wired local connections, while the router may provide connectivity to other networks.
Enterprise Access Link
A user's workstation commonly connects over twisted-pair copper to an access switch. The switch provides the local Ethernet connection and forwards frames to other ports or to an uplink. This design uses separate switched links rather than one shared cable for all users.
Building-to-Building Uplink
Two switches in different buildings may use fiber-optic Ethernet. Fiber is useful when the distance is too great for a typical copper run or when the link requires high bandwidth and resistance to electrical interference.
Legacy Shared Segment
Imagine several older devices attached to one shared coaxial Ethernet segment. Each device must listen before transmitting, but two devices can still transmit at nearly the same time. CSMA/CD handles the resulting collision. Switched full-duplex Ethernet replaced this approach in most current LANs because it provides more predictable performance and better scalability.
Troubleshooting Ethernet Concepts
- Connection fails with a new cable: Check that the ports and installed cable support the same Ethernet variant. Confirm connector compatibility, cable condition, and supported distance.
- Ethernet is described as only an RJ-45 copper cable: Remember that Ethernet is a standards family that also includes fiber and historically coaxial implementations.
- A long copper link is unreliable: Check the length and environment against the specification. Fiber may be more appropriate for a longer, high-speed, or electrically noisy link.
- CSMA/CD is assumed to apply everywhere: Determine whether the link is shared and half duplex. Modern switched full-duplex links generally do not use CSMA/CD.
Key Points to Remember
- Ethernet is a family of LAN standards associated with IEEE 802.3.
- Ethernet defines media, signaling, connectors and wiring variants, frame structure, and communication behavior.
- Coaxial cable was used by older shared Ethernet; twisted-pair copper is common for short access links; fiber supports longer distances and high bandwidth.
- An Ethernet frame is the structured data unit transmitted across an Ethernet LAN.
- CSMA/CD belongs to shared half-duplex Ethernet, not normally to modern switched full-duplex Ethernet.
- Ethernet speeds include 10 Mbps, 100 Mbps, 1 Gbps, 10 Gbps, 100 Gbps, and higher-speed families.
- Media selection depends on distance, cost, required speed, and the deployment environment.