Ethernet & IEEE 802.3

Computer Network & Network Security System — Data Link Layer, NEC licence examination syllabus (Nepal Engineering Council).

Ethernet & IEEE 802.3

The wired standard that quietly runs most of the world's local networks — and whose frame layout is worth being able to draw from memory.

The frame, field by field

📨 Ethernet Frame Format

PreambleA synchronisation pattern that lets the receiver's clock lock onto the sender's before any real data arrives.

Destination MAC6 bytes. Read first, so a card can discard a frame that is not for it without processing the rest.

Source MAC6 bytes.

Type / Length2 bytes. What protocol is inside — IP, ARP and so on.

Data46 to 1500 bytes of payload.

CRC4-byte check trailer, computed over the frame and verified on arrival.

The destination address comes first for a practical reason: a card can compare six bytes, decide the frame is not addressed to it, and stop — rather than buffering the whole frame before finding out. Ordering the fields this way costs nothing and saves work on every frame a machine ignores.

Why the payload has a minimum as well as a maximum

The 1500-byte ceiling is the easy half: a very long frame would monopolise a shared cable and delay everyone else, and an error in it would waste more retransmission.

The 46-byte floor is the interesting one, and it exists because of collision detection. A sender must still be transmitting when a collision reaches it, or it will never notice. On the original shared-cable Ethernet, the minimum frame size was chosen so that a frame takes longer to transmit than a signal takes to travel to the far end of the maximum-length cable and back.

Why a minimum frame length exists: A sends a very short frame ──▶ B starts sending too, before A's signal arrives A has already FINISHED and stopped listening → the collision happens, and A never finds out Padding short payloads up to the minimum keeps A transmitting long enough to hear the collision.

💡 So a payload shorter than the minimum is padded, not rejected. That is why a tiny message still occupies a full minimum-size frame on the wire, and why the Length field matters — the receiver needs to know how much of the padded payload is real.

The speed generations

Fast Ethernet raised the rate to 100 Mbps and Gigabit Ethernet to 1000 Mbps, both keeping the same frame format. That compatibility is the reason a modern network can carry decades-old software unchanged — the frame a machine builds today is laid out the same way as one from the original standard.

What changed underneath is the physical layer and, more importantly, the topology: modern Ethernet is switched rather than shared. Each device has its own collision-free link to a switch, which is why collisions have effectively disappeared from wired networks even though the mechanism for handling them is still specified.

Commonly confused

Ethernet and IEEE 802.3. Ethernet is the original commercial standard; 802.3 is its formal IEEE specification. They differ in the Type/Length field's interpretation but are treated as the same technology in practice.

The frame's maximum and the payload's maximum. 1500 bytes is the payload limit. The frame is larger, because the header and trailer are added to it.

Shared and switched Ethernet. The frame format is identical. Whether collisions can occur at all is not.

Rapid revision

Preamble, destination, source, type/length, data, CRC. Destination first so a card can discard early.

Payload 46–1500 bytes. The ceiling stops one sender hogging the medium; the floor keeps a sender transmitting long enough to detect a collision.

Short payloads are padded, which is why the length field is needed.

Fast and Gigabit Ethernet kept the frame format — that is what preserved compatibility.

Syllabus points

Ethernet frame format
Standards; Fast & Gigabit Ethernet

Create a free account to tick topics off, take notes as you read, watch the video lessons and get a day-by-day study plan built around your exam date.