tejas Logo

Home | Contact | Sitemap

Corporate Technology Products solutions parners Career News & Events
 

Introduction to Ethernet

What is Ethernet?

Ethernet is perhaps the most popular technology for network connectivity in the Local Area Network (LAN). The technology, as standardized in the IEEE standard, 802.3, includes the cabling and signaling information as well as the Carrier Sense Multiple Access with Collision Detection (CSMA/CD) scheme. The technology also defines a standard addressing format, with mechanisms to enable higher layer functionalities like VLAN awareness and Link Aggregation.

Ethernet, like many ubiquitous technologies of today, was born at Xerox PARC in the early 1970s. The technology was intended to facilitate the communication of computers over a shared medium. The initial medium of choice was the co-axial cable but this has now been replaced with twisted pair copper or optical fiber in almost all applications.

Ethernet, today, is the technology of choice for connectivity in all LAN applications. Be it connecting the computer to a switch, the switch to a router or a wireless access point to a router, Ethernet is everywhere. The initial applications used a shared medium, half duplex approach but nowadays, with switches becoming cheap and hence accessible to even the simplest of networks, almost all Ethernet connectivity is dedicated and full duplex.

Functional Details

Now, let us take a look into the various functional nitty-gritty of Ethernet.

  1. Carrier Sense Multiple Access with Collision Detection (CSMA/CD): In the shared-medium scenario, Ethernet uses a simple scheme to interconnect multiple stations. When a station has a packet ready for transmission, it checks if the medium is idle. If so, the packet is transmitted immediately. If not, the station waits for the medium to become idle.

    In case a collision (station senses signal on the medium, while it is transmitting) occurs, the station stops transmitting and retransmits after a random back-off period. This approach to connectivity is fairly obsolete now, with network designers using a switch with dedicated full-duplex links to various stations.


  2. Auto-negotiation: This is a mechanism, which provides for two connected stations to figure out the best possible configuration (with regard to speed and duplex mode) to operate in. This removes the possibility of manual misconfiguration and hence sub-optimal network performance greatly.


  3. Framing Details: The figure below shows the structure of a typical MAC frame.



Preamble (7 Bytes)
Start of Frame Delimiter (1 Byte)
Destination Address (6 Bytes)
Source Address (6 Bytes)
Length Type = 802.1Q Tag Type (2 Bytes)
Tag Control Information (2 Bytes)
MAC Control Length/Type (2 Bytes)
MAC Client Data + Pad (42- 1500 Bytes)
Frame Check Sequence (4 Bytes)

MAC Frame

The MAC Control Length/Type indicates the type of the frame (Example: Ethernet II, SNAP etc.) The 802.1Q Tag Type and the Tag Control Information are applicable to a VLAN tagged frame only.

© 2007 Tejas Networks. All rights reserved. Home | Contact | Sitemap Designed by Greynium Logo