Description of the data link layer

1. The Media Access control layer (MAC)
2. The Logical Link Control layer (LLC)

1. Creating, transmitting, and receiving packets (described below),
2. Handling physical (MAC) addressing,
3. logical link control proccessing,
4. creating logical topologies, and
5. controlling media access.

Packets

The data link layer takes data that it receives from the upper layer protocols and divides them into logical chunks called packets. A packet is a unit of data transmission and varies in size and format depending upon the transmission technology.

Physical (MAC) Addressing

All Network Interface Cards (NICs) have a protocol independent address assigned to them at the factory. This address can be called the hardware address, physical address, or the Ethernet address. The most correct term, however, is MAC (Media Access Control) Address. It is a 12-digit hexadecimal number and there is cooperation between NIC manufacturers to avoid duplication of addresses. As the number of NICs produced grows some manufacturers are having to reuse their MAC addresses. If an administrator has a problem on his network with duplicate addresses he can use a program supplied by the manufacturer to resolve this conflict.

Logical Topology

In an earlier section we discussed physical topologies, the manner in which cables and computers are laid out. The data link layer dictates the Logical Topology of a network, that is, the manner in which information flows across a network. The type of logical topologies are the same as physical topologies, except that the data flow specifies the type of topology to use. The logical topology also describes the method of media access. The three main types of media access are:

1. Contention: All stations on a network compete for the opportunity to transmit. If two stations transmit at the same time a collision occurs and the stations try again.
2. Polling: A controller polls each station, in turn, asking if it has data to transmit. This method was developed to help eliminate collisions.
3. Token Passing: A special data packet called a token is passed from station to station. A station must be in possession of this token in order to transmit. This type of media access technology is also used to help eliminate collisions.

Media Access Technologies

There are, at least, two media access technologies that are instrumental in the implementation of the above methodologies. They are:

1. CSMA/CD (Carrier Sense Multiple Access/Collision Detection): This is a type of media access technology used mostly by Ethernet networks. It works by first detecting to see if a signal is already on the wire. If no one else is transmitting, it sends its data. It then checks to see if another station sent its own data at the same moment. If this is the case and a collision has occurred both stations wait for a period of time and try again.
2. CSMA/CA (Carrier Sense/Multiple Access With Collision Avoidance): This is similar to the above with the following exception; instead of sending the entire chunk of data and then checking for a collision, it sends a signal called RTS (request to send). It then waits for a CTS (clear to send) response. Once it receives this signal it sends the data chunk.

Data Link Layer Devices

The two devices used to manipulate data at the data link layer are:

1. Bridges: A bridge is a network device that logically separates a network into two segments, but lets each segment appear as one to higher layer protocols. This is usually used to separate two parts of a network from one another when one part experiences higher traffic that causes the other part of the network to slow down. Bridges can also be used to connect dissimilar network types (Ethernet and Token Ring are an example of dissimilar networks) as long as the bridge operates at the LLC (Logical Link Control) sub-layer of the data link layer. If a bridge operates at the lower MAC (Machine Addressing Control) sublayer it can only connect similar networks.
2. The Switching Hub: A switching hub is more intelligent then a standard hub. It has the ability to understand some of the traffic that passes through it. A switch operates at the data link layer and is also reffered to as a layer 2 switch. Layer 2 switches build a table of the MAC addresses for all stations that it connects to. The main advantage of a switching hub over a standard bun is that it is able to analyze outgoing data from one station and determine the MAC address of the recieving station. It does this by looking at the data link header and, because of this, is able to open a "virtual pipe" between the sending and receiving station that utilizes the full bandwidth of the topology. This greatly reduces the incidence of data collision.

Back to the Computer Area
Back to The Nest
On to the next lesson