Ethernet – Understanding Interpacket Gap and Idle Detection

The reason for 46 byte payloads:

According to paragraph 6.3.2.3 of the Ethernet V2 Spec, the minimum ethernet frame is based on the Ethernet Slot Time, which is 512 bit lengths (64 Bytes) for 10M ethernet. Slot time governs both maximum cable length and minimum frame size.

After subtracting 18 bytes for the ethernet header and CRC, you get 46 bytes of Ethernet payload as the minimum payload size.

Slot time (and thus the minimum frame size) is also closely related to Ethernet collision detection. Quoting An Introduction to Computer Networks, paragraph 2.1.2:

... a collision can be received, in principle, at any point up until the end of the slot time. As a result, Ethernet has a minimum packet size, equal to the slot time, ie 64 bytes (or 46 bytes in the data portion). A station transmitting a packet this size is assured that if a collision were to occur, the sender would detect it (and be able to apply the retransmission algorithm, below). Smaller packets might collide and yet the sender not know it, ultimately leading to greatly reduced throughput

Ethernet Slot Time was specified so CSMA/CD would correctly function. The minimum size of a frame is defined to make sure that its transmission takes enough time so that even with a shortest valid frame, a possible collision can be reliably detected; if the frame size is too small (with respect to the maximum cable length), deterministic collision detection would be impossible.

1500 byte payloads:

We have already discussed the reason for 1500 byte MTUs; please refer to that question for specifics.

Note: ytti's comment about the FCS limitation of ethernet is not the reason that 1500 bytes was chosen. It was chosen due to issues with interpreting the Length field in 802.3 encap frames vs the Type field in Ethernet II frames.

802.3 (which uses 802.2 LLC format) has a Length field in the same place that Ethernet II has a Type field.

• IEEE 802.3 with 802.2 LLC (used by Spanning-Tree, ISIS) use the highlighted bytes for a Length field. 802.3 Upper-layer protocols are decoded via the 802.2 LLC Header / SNAP bytes. The SNAP bytes are used to decode protocols using traditional ethertype values; SNAP is only included when the 802.2 LLC DSAP / SSAP = 0xAAAA.

      +----+----+------+------+------+------+-----+
      | DA | SA | Len  | LLC  | SNAP | Data | FCS |
      +----+----+------+------+------+------+-----+
                ^^^^^^^^
  
      DA      Destination MAC Address (6 bytes)
      SA      Source MAC Address      (6 bytes)
      Len     Length of Data field    (2 bytes: <= 0x05DC or 1500 decimal)  <---
      LLC     802.2 LLC Header        (3 bytes)
      SNAP                            (5 bytes)
      Data    Protocol Data           (46 - 1500 bytes)
      FCS     Frame Checksum          (4 bytes)
  

• RFC 894 (commonly known as Ethernet II frames) use these bytes for Type. Upper-layer protocols are decoded via the Type field

      +----+----+------+------+-----+
      | DA | SA | Type | Data | FCS |
      +----+----+------+------+-----+
                ^^^^^^^^
  
      DA      Destination MAC Address (6 bytes)
      SA      Source MAC Address      (6 bytes)
      Type    Protocol Type           (2 bytes: >= 0x0600 or 1536 decimal)  <---
      Data    Protocol Data           (46 - 1500 bytes)
      FCS     Frame Checksum          (4 bytes)