Ethernet, one of the most popular LAN technologies, refers to a number of related LAN protocols that use different media and have different capacities. On the low end are 10BASE-T Ethernet, which uses unshielded twisted pair media, and 10BASE-2 Ethernet, which uses coaxial cable. Both support data rates of 10 megabits per second (that's what the 10 in the names refers to), which is a little more than 1 megabyte per second. To put that in perspective, this level of network could transmit the contents of a full CD-ROM in about ten minutes.
On the upper end is Gigabit Ethernet, which, as the name implies, can transmit a gigabit, or 1,000 megabits, per second. That means it could transmit an entire CD-ROM in about six seconds. There is even a 10 Gigabit Ethernet that would reduce that to half a second! Gigabit Ethernet can use multiple copper wires but most often uses optical fiber.
In Ethernet, nodes compete for the use of the network. As you've seen in other types of networks, each node shares a data link with adjacent nodes, and some nodes act as a message forwarding service. In Ethernet, all the nodes on the network share a common link. All the nodes share the medium, are always listening, and can send messages at any time.
You might wonder how the nodes schedule the use of the network. The answer is that they don't. Any node can use the network as long as it does not detect another message on the network. A collision could occur when two nodes try to send at the same time, the result being garbled data. Because it takes time for a message to propagate down the line, if node A starts to send a message, there is a short delay before node B senses this message on the line. If node B sends its own message during this delay, the collision occurs.
Fortunately, nodes can sense when a collision has occurred. The nodes that tried to send will wait a random amount of time and then resend the message if the network is clear. Because the wait is random, usually one node starts retransmitting first and then the other detects it and waits. In this way, Ethernet allows the nodes to share the network without directly communicating with each other about the details.
If a node were to send out its entire message before a collision occurred, even though it could sense the collision, it would not realize that its own message was affected and would instead assume a message from other nodes caused the collision. Because of this erroneous assumption, it would not resend the message. If the wires connecting the nodes are short, this situation cannot happen because the first part of the message reaches every node before the sender sends the last part of the message. Any collision must occur within that time because nodes only send when they haven't sensed another message on the wire.
The collision domain is the maximum distance of wire between two nodes so that the beginning of a message from one node reaches the other before the end of the message is sent. When Ethernets are designed, they must ensure that the wiring used doesn't exceed this length from one of the nodes to the other.
Unfortunately, if faster Ethernet versions are used, messages travel faster, and the collision domain is smaller. A smaller domain means a shorter network, probably with fewer nodes, which sounds like a bad deal because with more nodes, one wants more speed. The solution is bridges, which we discussed previously. By placing bridges strategically throughout the Ethernet, the overall network becomes a set of mini-Ethernets, each one well within the confines of its collision domain.
Frame Relay is a high-speed network protocol used for connecting LANs to make a WAN. WANs were originally created by leasing dedicated backbone lines from long distance service providers. If a company has an office in New York and another in Albany, it would lease the exclusive use of either a bundle of wires or an optical fiber that ran between the two points.
This sort of lease was wasteful, though, because the company would have to lease based on its maximum usage of the line. In other words, if it occasionally needed to send a gigabyte of information in a few seconds, the company would have to buy a line that could support that even if most of the time it didn't use the line much at all. When a sender transmits a lot more data than its average rate, on the other hand, it is known as bursty data.
Another problem is that if the company had multiple offices, it would have to lease lines to connect each office to every other office. While a company with two offices needs one line, a company with five offices needs ten.
What was needed was a WAN that could handle bursty data without requiring excess bandwidth and that could simplify connections between multiple offices. Frame Relay does both. A long distance service provider that offers Frame Relay is essentially allowing all the WANs it supports to share the use of all of its cables or optical fibers. There's plenty of bandwidth for bursty data as long as all the WANs are not bursty at the same time.
A disadvantage of Frame Relay is that individual messages experience variable delays in reaching their destinations. This delay makes it problematic for applications like real-time video because the image data must be chopped up into small messages to be sent, and a delay in any of the messages can prevent the video from being displayed properly.
Our website is not responsible for the information contained by this article. Articleinput.com is a free articles resource thus practically any visitor can submit an article. However if you notice any copyrighted material, please contact us and we will remove the article(s) in discussion right away.
Note: This article was sent to us by: Raymond Harris at 02132011
1. Modeling humans and shapes
All articles are property of their respective authors. Please read our Privacy Policy!
© 2009 ArticleInput.com.
Partners: Damenmode