In cases where a special connection is not established between the devices and there are common connections in communication, more than one device communicates at the same time over more than one communication, the same line, the same cable, or the same channel. Multiple Access Protocols is an OSI Layer 2 definition that determines which data can be transferred to which network environment in which time interval. The Data Link Layer is responsible for operations such as Link Controls and Multiple Access Protocols in the communication established between the sender and the receiver.
Multiple Access Protocols are used to reduce the collision that may occur in the channel during transmission, to enable more than one user to access the same channel at the same time, and to reduce the congestion that may occur in traffic. Multiple Access Protocols are examined under three different headings: Random Access, Channel Access, and Control Access. This article will talk about Random Access Protocols.
Random Access Protocols were discovered in 1971, it has preserved its core with the same working structure until today. Random Access Protocols deals with transmission medium in communications as part of multiple access protocols.
Generally, the number of routers or switches in networks is less than the number of users. Traffic density and congestion in packet transmission occur on routers that communicate from one point to many points. For this reason, when users want to access the target via routers, there may be conflicts between packets. Conflict can cause data loss in both packets. Packets involved in the collision are discarded or lost. The colliding packets need to be retransmission by the source devices so that the transmission can be completed. The period of confliction and unsuccessful transmission time is considered lost time in communication. Thus, time losses in transmission, extra energy consumption, traffic density, and low network performance occur. Due to the Random Access Techniques to be used at this point, the transmission medium can be used more efficiently and the loss of time between packets is minimized.
The working principle of Random Access Techniques varies depending on the occupation of the network and the traffic situation. Packet forwarding is random among users because there is no user-specific priority or scheduled time for access, transmission. The device that receives the communication waits for the moment when the medium is available (empty) to transmit the data. Depending on the traffic, data transfer is performed when the environment is suitable for data transmission. If different packets are sent to the same medium by two different devices at the same time conflict, there will be distortions and packet losses in the transmitted packets. To ensure communication, the packets are managed from a central point and sent again in a certain order. Thus, packet and time losses are minimized. Various methods have been developed to prevent packet corruption, packet loss or time to be experienced. At this point, there are some different working structures.
1- ALOHA
In communications with more than one source or multiple destinations sharing a common data transmission channel, packet losses occur when packet headers collide and existing information is damaged or lost. ALOHA is a simple transmission technology used to avoid collisions in the packet transmission of the receiving or transmitting devices in the network. Generally, ALOHA works very efficiently in wireless communications, antenna systems broadcasting wirelessly, or networks with half-duplex structures. But in its build, collision detection is not available. It has two different types, Pure Aloha and Slotted Aloha.
Pure ALOHA
Pure ALOHA is a transmission technique that does not initiate the sending of the next packet until the sending of the sent packet is completed, to ensure error-free data transmission.
Packets to be transmitted are first fragmented. Fragmentation is the process of dividing the data to be transmitted into sequential pieces, regardless of the size of the medium to be transmitted. Data is segmented and transmitted sequentially, depending on the bandwidth. The receiving device, on the other hand, combines the data in order after it receives it. After the packet becomes available for transmission, the sending device starts the transmission one after the other.
If the packet transmission is successful, a successful acknowledge is sent to the device. As a result of successful transmission, the next packet is queued for transmission. If the transmission message (acknowledge) that the transmission is successful is not transmitted to the sending device within a certain time, it is understood that the transmission has failed. If the transmission fails, the same packet is sent again after a random time. If the retransmission limit is exceeded, the packet is considered lost and the next packet is transmitted.
However, as the number of users sending and the number of packages to be sent increases, certain problems may occur. When the return message indicating that the transmission is successful, the packets that will start the transmission are queued or conflicts may occur at the first moment of transmission. Different packets that want to start transmitting at the same time are sent to the channel unaware of each other and overlap at the channel entrance. Also, the devices to initiate packet transmission may transmit data at random times, ie any number of senders may wish to transmit data at any given time. Users who are not aware of each other can also cause conflict. The Slotted ALOHA method has been developed to prevent losses.
Slotted ALOHA
Slotted ALOHA has the same working structure as Pure ALOHA. It has emerged as a solution to the problems that may occur because the transmission is not done in regular order at the beginning of the Pure ALOHA technique.
When the appropriate transmission environment is formed, the packets that are ready for transmission and sequenced are sent at periodic time intervals. Packets that cannot be transmitted in the time slot have to wait for the next transmission time.
When the transmission channel is suitable for packet transmission, the sender is notified with a “beacon” signal. The sender receiving the signal sends the packets to the transmission medium at certain time intervals. Thus, packets are not sent to the same environment at the same time and conflicts are avoided. Packets that do not return a transmission notification message indicating that the transmission was successful are re-queried by the sender and sent again. As a result, the number of collisions is halved and ALOHA efficiency is doubled.
2– CSMA (Carrier Sense Multiple Access)
CSMA is a technology that detects the traffic of the medium to be transmitted before packet transmission. Thus, it is aimed to reduce the probability of collision by performing the transmission when the channel is empty.
In order to detect the busyness of the communication channel, a carrier signal is transmitted before the packet is sent to detect whether the channel is empty or full. If the channel on which the communication will be provided is empty, packet transmission can be performed. If the channel is busy, packet transmission is not performed until the traffic density decreases and the channel becomes available for transmission (until the channel is free).
However, there is still the possibility of collisions when using CSMA. Two different senders may have listened to the medium at the same time and detected the transmission channel as empty at the same time. In this case, both senders will transmit packets at the same time and the packets will overlap on the channel. With the advent of CSMA, two types of CSMA technologies have been developed. These; Collision Detection and Collision Avoid.
3- CSMA/CD (Carrier Sense Multiple Access with Collision Detection)
CSMA/CD has the same as CSMA in terms of its working structure, it also has an extra collision detection system.
A carrier signal is transmitted to examine the traffic density of the channel before transmitting. If the transmission channel is empty, packet sending is started. If the forwarding channel is full, packet sending is suspended until the channel is empty. A possible conflict after packet sending starts transmits a signal to the system called “Jam Signal”, indicating that there is a conflict in the channel. When the system receives the Jam Signal, it detects that the last transmitted packet has collided and is corrupted. Retransmission of the conflicting packet is achieved after waiting for an arbitrary amount of time so that the transmission can be completed. If the transmission is successful, the next packet is prepared for transmission. Also, when the transmission is completed without a collision, there is no acknowledgment message is sent to the system that the transmission was successful. Because if the system did not receive a Jam Signal, it means that the transmitted packets did not conflict and the transmission was successful.
If the packet is transmitted more than once and there is a conflict every time it is transmitted, the packet exceeds the transmission limit and is dropped. The reason for packet loss at this stage is to recover the system from a problematic packet.
4– CSMA/CA (Carrier Sense Multiple Access with Collision Avoid)
CSMA/CA has an improved structure for more secure communication. It has been developed to ensure the healthy transmission of the packet without conflict and the packet sending from the source to the destination. It is often used in networks with wired communication.
Before the packet transmission starts, the queued packets are sent to the channel to be transmitted in order. Packets sent to the channel wait for random times. The expired packet continues on its way to the transmission channel and starts listening for the busy status of the communication channel. The status of the channel has been listened to until the channel is empty. The packet, which is sure that the channel is empty, is chosen again at a random number, and counting is performed before continuing on its way. When the counting process is finished, the busyness of the channel has listened again and each time the channel is busy, the random number counted in the previous stage, that is, the waiting time, doubles and is counted again. The counted packet listens for the busy status of the channel again before continuing on its way for transmission. If the channel is full, it waits until the channel is empty. If the channel is empty, the packet continues on the transmission path and transmission takes place. After packet sending is complete, an acknowledgment is sent to the device that initiated the transmission, indicating that the transmission was successful. Thus, the transmission is considered successful. If the system does not receive a return message indicating that the transmission of the sent packet was successful within a certain period of time, the packet is considered to be conflicted or lost. The undelivered packet is sent again, taking all the transactions from the beginning. The sending of the failed packet continues until the packet sending limit is reached or the packet is successfully transmitted. The package that exceeds the packet sending limit is dropped and considered as packet loss, it will not be sent again.
If the random counts are not counted as much as the determined random number, that is, before the timer expires, the packet is considered as lost and the transmission is started again. Similarly, if a successful transmission message is delivered to the system after the timeout period, the packet is considered lost and retransmitted.
Transmission time is extended when the communication channel or transmission medium is busy. Collision is likely to be avoided, but packet transmission times will increase. The continuation of the communication path of the packets at certain intervals and at certain times reveals a structure similar to TDMA (Time Division Multiple Access).
Thanks to CSMA/CA, a possible collision with the waits created in packet transmission is minimized. In addition, thanks to the acknowledgment message is sent to the system indicating that the packet has been transmitted, it can be detected whether the packet is involved in any conflict process. The basis of being able to communicate without losing data on the transmission channel is provided.
The graph of the mentioned protocols in terms of transmission rate and experienced packet delay times is shown on the side. The point to be considered at this stage is that the protocols are not examined as good or bad. The area where the technologies will be used should be preferred depending on many factors such as the type of data to be transmitted, the condition of the transmission medium, the transmission time, and so on. According to the mentioned situations, the type of communication changes. The type of transmission should be preferred so that the communication is most efficient.
