Chapter 4: MAC Layer Flashcards
Medium Access control (MAC)
Within the data-link layer is communication between 3 or more hosts, and this communication is contained in the MAC layer.
Most important applications
Effective functioning of wired LANs.
Most important problem with communication between more than 2 hosts.
How the bandwidth is shared between hosts.
Why won’t multiplexing work
If a host goes offline, then bandwidth will be wasted. If a channel is formed using multiplexing, then you cannot apply other techniques atop of it.
TO dynamically allocate bandwidth, the MAC protocol needs 5 assumptions about the network and the hosts of the network
- Each station/host has its own independent traffic
- There is a single channel that is being shared.
- Collisions (signals collide) are observable
- The usage time for the channel can be either slotted or continuous
- The hosts may or may not have carrier sense, i.e. the ability to determine if the channel is already being used by another host.
MACA
Multiple access with collision avoidance
- each hosts declares its intention to send by transmitting a request-to-send frame and then waiting for the receiver to respond with a clear-to-send frame.
Multiple access protocols
These protocols acknowledge the possibility of collisions and try to work around them. Of these, the most elementary protocol is the Aloha protocol.
Aloha protocol
Each host transmits its data whenever it wants, without regard for the state of the channel. If collisions happen, they happen.
How Aloha identifies collisions
Once a host receives a frame, it retransmits it on the network. Once the retransmitted frame is received by the sender, it is compared to the frame that was originally sent. If the frames differ, a collision has occurred, and the frame is resent after a random amount of time.
Problem with Aloha
The more hosts that are part of the network, the more collisions occur. It counters this by making channel time slots where hosts have to wait for their slot to send data.
Carrier sense multiple access (CSMA) protocols
Implements technologies and techniques that allow a host to sense if a channel is busy.
3 main CSMA protocols
All 3 of them are CSMA/CD (collision detection)
1. 1-persistent
2. 0-persistent
3. p-persistent
1-persistent
Very greedy protocol and tries to use the channel as soon as it becomes available. If it detects that the channel is not busy, then it sends its data immediately. If it is not busy, then it will wait, check again and see if it is busy or not.
0-presistent/non-persistent
When a host has data that it wants to transmit, it senses the channel; if the channel is idle, it sends it immediately.
If the channel is busy, it waits a random amount of time and senses again. If it is idle, then it sends it
P-persistent
The host will sense the channel when it has data to send. However, this will only be done at the start of a time slot; if the channel is busy at the start of the time slot, the host will wait till the next time slot to sense again.
If the channel is idle, it will do a probability calculation to check the chances of collisions occurring. If the chances of collision is smaller than the chance of success, then it sends the frame.
Collision free protocols
Attempt to eradicate collisions by establishing a sharing scheme amongst the hosts.
The three protocols are bitmap, token passing and binary countdown.
Bitmap
Is a reservation protocol.
First step (reservation): each host is given a short period of time to declare if they have anything to send. If yes, they broadcast a 1 to the network within their time slot. If not, they dont transmit anything.
All chances that did declare can transmit data in order. The process then repeats.
Token passing
With token passing, the hosts are arranged into a ring order and a token is passed from host to host. Only the host with a token can transmit.
Once the host has finished transmitting, it transfers the token to the next host in the ring.
Binary countdown
Allocates a binary identifier to each host. When a channel wants to send, it transmits its identifier to the other hosts. The identifiers are then OR-ed by all the channels that want to transmit. The results of the OR-ed identifiers are then used to determine a priority, which is followed when the hosts transmit their data.
Another function of the MAC layer
It allocates addresses to each host. They are known as MAC addresses and uniquely identifies devices on a data-link layer.
MAC addresses contain…
Information that can be used to identify a device’s type and manufacturer, which can be used to optimise both bandwidth allocation and security.
Control devices used on networks
Repeaters, hubs, switches/bridges, routers and gateways
Repeaters
Repeaters take a signal, strengthen it, and then retransmits it on the same channel. These devices do not analyse data in any way, and are not responsible for flow control. As such, they are physical layer devices
Hubs
Rudimentary connection points that connect multiple hosts to one another. Like repeaters, they do not control the transmission in any way; if a hub receives a transmission from one host, it forwards that transmission to every other connected host. They are also physical layer devices.
Switches
A.K.A bridges, and they have a similar function as hubs. Switches read the data received, extract a destination address for each transmission, and then send the transmission to exactly one recipient. They use the MAC address, so it is a data-link layer device.
Routers
Routes find the optimal path through the network from one point to another. The internet is highly dependent on them.
This device requires routing algorithms and protocols found on the network layer.
Gateways
Gateways act as single access points to other networks. They also monitor and restrict data in the interest of security, targeted network use and other stuff. Because they monitor the data, they form part of the transport or application layer.
