Full duplex is the base of the internet as we know it nowadays. Connections on networks can operate using half duplex, where data is sent in one direction at a time and full duplex is the operation of sending data simultaneously between nodes. In the context of Ethernet, full duplex is the way of allowing a simultaneous communication between two nodes and the data sent is sent without interfering with the transmissions of each other.
The link between these nodes must be point-to-point and cables like twisted pair or fiber optic can be used in full duplex operation since they provide paths for transmit and receive. In full duplex operation, the capacity of the link is doubled since data between the nodes is sent simultaneously.
In order to have full duplex operation in Ethernet, several requirements are to be met. One of the main hardware required is the media link between the nodes as stated above. The link must provide paths for transmit and receive in order for full duplex to be operational. Another hardware requirement for full duplex in Ethernet to be operational is the fact that the Ethernet interface (controller) of the sender and the receiver to be configured to be in full duplex mode of operation.
Since Ethernet uses full duplex operation and collisions on a network setup occurs when one or more nodes attempts to send data at the same time, then a collision occurs, in this context, each node on the network will have its own collision domain meaning that collisions on a full duplex network setup cannot happen since the data cannot collide because the transmit and receive of the data is being done simultaneously.
In the context of Ethernet, full duplex operation with the use of CSMA/CD algorithm is useless since the medium is not shared between the nodes, therefore the use of multiple access algorithms like CSMA/CD are unnecessary. When any two nodes are configured to use full duplex operation, the multiple access algorithm CSMA/CD is shut off in order to avoid problems since it is irrelevant to use. (Spurgeon, 2000)
The access method implemented on each of Ethernet networks as well as wireless networks are Carrier Sense Multiple Access Collision Detection (CSMA/CD)and Carrier Sense Multiple Access Collision Avoidance (CSMA/CA) respectively.
The CSMA/CD protocol is a protocol for carrier transmission access and it is used in Ethernet IEEE802.3 standard. This protocol deals with collisions on the network after it has occurred since only one frame can be sent at a time. (ethermanage, n.d.)
When a node on the network wants to send a frame it first must wait until the channel contains no signals, after it can start transmitting. Each node on the network must be separated by a time interval of 96 bit times. This can also be called interframe gap and when passing through a repeater, this gap can be reduced. If another node is transmitting on the same channel at the same time, a collision occurs and after detecting that collision, the sender will continue to send 32 additional jam bits. After the collision detection, the frames are discarded and the nodes are given a random delay period before they attempt to transmit again. After that random delay period expires, the node can attempt to transmit again. (Feit, 2000)
The CSMA/CA protocol is used in the IEEE802.11 standard. This protocol tries to avoid collision in order to provide better multiple access by nodes. The nodes transmitting must first listen to see if other nodes are transmitting, but if the medium is not available since another node is transmitting, it will have to wait until the medium becomes available. This waiting time is called backoff factor which is set randomly and when it reaches to zero, it can check the medium again in order to start transmitting. When the medium is available, it still doesn't send the data but instead sends a signal of intent, stating that this particular node is about to transmit data onto the cable so that other nodes doesn't send data at that time to avoid collisions. After that, the data is transmitted from the sender to the receiver. (thought1, n.d.) (searchnetworking.techtarget, n.d.)
The main reason CSMA/CD protocol cannot be used in the IEEE802.11 standard is because;
"wireless stations are not capable of transmitting and receiving at the same time, so they are not capable of detecting a collision during their transmission. For this reason, 802.11 wireless networking uses CSMA/CA instead of CSMA/CD to try to avoid collisions"
Pg 252.. CWNA book
As quoted from the book, CSMA/CD cannot be used in the IEEE802.11 standard for wireless networks because since the wireless nodes cannot transmit and receive at the same time, it will be impossible for the nodes to detect collision during transmission between the nodes. For this reason, CSMA/CA is used that will try and avoid collisions rather than detect a collision that has already happened.
PPP is an acronym for Point-to-Point Protocol; this is a Data Link layer protocol that is used over asynchronous serial and synchronous serial media. It is used in order to allow communication between a set of nodes by using a serial interface. Its main job is to encapsulate the other layers data and be able to transmit this data on a serial interface, therefore it transports packets from the network layer over the data link layer's point-to-point link. This protocol uses the Link Control Protocol in order to form and sustain data link connections. It allows to use simultaneous network layer protocols.
Point-to-Point Protocol is made up of two major components, the Link control Protocol, which is used in order to create the point-to-point link, it also configures, tests and maintains the link connection between the two nodes, it also takes care of the termination of the link between the nodes. This link control protocol is a method that first, it establishes the configuration parameters before any datagram can be sent. Once the establishment is done, this protocol tests the link and checks whether the link has sufficient quality to bring up network layer protocols. After the quality test of the LCP protocol, the actual protocol configuration can begin, after the protocol configurations are set, the link can be terminated by a user or because of an event, like loss of carrier or expiration of the idle timer.
This protocol also offers different configuration options like authentication, compression, error detection, in order to monitor and avoid loops and multilink, which performs load balancing across links.
The other major component is the Network Control Protocol. This protocol can include several protocols and it is used for the establishment and the configuration of several different network layer protocols. This protocol is responsible for the negotiation of the network layer parameters. Once the link is established as discussed before, this protocol will determine what network layer protocols, like IP and IPX, are chosen and configured. After the protocols of the network layer are chosen and configured, datagrams of those protocols can be sent over the point-to-point communication link. (netcert.tripod.com, n.d.) (technet.microsoft.com, n.d.)
error control, addressing mechanism and flow control