|
A frame is the lowest level of organizing bits for transmission on a circuit. It corresponds to Layer 2 of the OSI 7-layer reference model.
Layer 1 of the OSI model specifies the physical circuit: what kind of physical medium is used, how 1s and 0s are represented on the physical medium (e.g. light on / light off on fiber), how often that can be done and so forth. The result is the ability to transmit bits from one box to another.
Layer 2 deals with the fact that there is no such thing as an error-free, unlimited capacity physical circuit; and that there may be multiple devices connected on a single circuit. The result is the ability to transmit bits with some notion of reliability to the correct device on the same circuit .
Today, this is implemented with LAN technology. Ethernet was a brand name for the original LAN product, and is now used to refer to a set of standards published by study group 802 of the Institute of Electrical and Electronic Engineers (IEEE).
To deal with errors, data is transmitted in blocks of about 1500 bytes - this is the payload or information field that makes up the heart of a frame – and an error detection mechanism called a Frame Check Sequence (FCS) is calculated by the transmitter and appended to the block. Then with a mathematical operation, the receiver can detect to better than 99.99% accuracy if any of the received bits are in error. If an error is detected, the whole thing is discarded and retransmitted. This entire process is known as Cyclic Redundancy Checking (CRC).
Two mechanisms are required to deal with multiple devices on the same circuit: access control and addressing. Access control is controlling which device is permitted to transmit at a given time, and ensuring that only one device does so. Addressing is dealing with the fact that on a circuit with multiple devices physically connected, all devices will receive all the bits that are transmitted – so it is necessary to indicate for whom a transmission is intended; which device should react to a transmission.
Other functions include framing, marking the beginning and end of the transmission, and control information such as sequence numbers or an indication of the size of the payload.
The 802 family of standards specify, amongst many other things, the access control mechanism, the addressing scheme and the frame format illustrated. One of the standards describes the Media Access Control (MAC) sublayer, and so the addresses are often called MAC addresses .
MAC addresses are permanently assigned by manufacturers to LAN interfaces, i.e. the electronics sitting behind a jack that a LAN cable is plugged into. Currently, MAC addresses are six bytes long: the first three bytes identify the manufacturer, and the last three bytes identify the device.
To communicate data from one station to another on the same physical connection, the transmitting station's LAN driver software takes a block of data and puts the MAC address of the desired destination device on the front of the block, along with the source MAC address, control information, framing and the FCS at the end of the block, creating the frame.
When allowed to do so by the access control mechanism, the frame is then transmitted on the physical circuit one bit at a time. All devices physically connected on the circuit receive the frame, verify if there are any errors, then compare the destination MAC address on the frame to the MAC address on their LAN interface. If they are the same, then the device knows it should react to the frame.
The result is to move a block of data from a source device to a specified destination device on the same physical circuit and know that there are no errors in the block.
The next set of questions: what happens if we want to transmit data to a device on a different circuit?
That is next month's tutorial: Layer 3, IP addresses, packets and routing.
|
