- Collision:
On Ethernet, only one device may transmit at a time. Any device trying to transmit will check whether the medium is busy by using carrier sense. If carrier is not detected and medium has not been busy for a specified period, then the device begins transmitting its data. While it is transmitting a frame, a device listens to ensure that no other station starts transmitting.
However, if any other station begins transmitting at the same time, a collision occurs. In other words, a collision is a condition where two devices detect that the network is idle and try to send packets at approximately the same time and end up trying to send packets at exactly the same time (within 1 round-trip delay). Since only one device can transmit at a time, both devices must back off and attempt to retransmit again.
The retransmission algorithm requires each device to wait a random amount of time, so the two stations are very likely to retry at different times; thus, the second station will sense that the network is busy and wait until the frame is finished being transmitted. If the two devices retry at the same time again, they will collidge once more. This process repeats until either the packet finally makes it onto the network without collisions, or 16 consecutive collisions occur and the packet is aborted.
Collisions that occur due to two devices transmitting at the same time are normal events on Ethernet. As the amount of traffic on a network increases, the number of collisions on the network also increases. However, collisions due to two stations transmitting at different times are not considered normal.
Unusual number of collisions on a network indicates a problem and you should examine the network. Oftentimes, these collisions are the result of either a misconfigured network or a device transmitting when it should not be and stepping on other packets. Most likely, the collisions are the result of reflections or other problems with the physical plant, such as a bad transceiver. A ground loop is another possible cause of collisions if the cable is grounded at two points. Collisions can also happen if there is a bad cable or an incorrectly connected station on the network. - Collision Rate:
The collision rate is calculated by dividing the total number of packets by the total number of collisions and multiplying the result by 100. What is an acceptable level of collisions? This depends on your application and protocol along with the physical design of your network. In some cases, collision rates of 50% will not cause a large decrease in perceived throughput. If your network is slowing down and you notice the percentage of collisions is consistently above 10%, you may want your network examined. - Cyclic Redundancy Check:
The Cyclic Redundancy Check (CRC) is a an error-checking algorithm used to check for the validity of a frame. A transmitting station uses the algorithm to compute a CRC value based on the bits in the frame and sets the Frame Check Sequence field of the frame to the calculated CRC value. A station receiving the frame can then use the same algorithm to determine whether or not the frame has been corrupted. If the computed CRC value of a frame received by a station does not match the value in its Frame Check Sequence field, then the frame is bad.
C
Submitted by nestor on Sat, 2005-04-30 01:03.
»
Wed, 2010-09-01 12:57
- Login or register to post comments
»