Describe tcp and how it provides reliability of packet delivery.

Describe tcp and how it provides reliability of packet delivery.

How tcp works – sequence numbers

A reliable protocol in computer networking is a communication protocol that informs the sender whether or not data transmission to intended recipients was efficient. The ITU and the ATM Forum use the term reliability as a synonym for assurance.
Reliable protocols have more overhead than unreliable protocols, and as a result, they run slower and are less scalable. This is usually not a concern with unicast protocols, but it can be with stable multicast protocols.
The key protocol used on the Internet, Transmission Control Protocol (TCP), is a dependable unicast protocol. UDP is an insecure protocol that is commonly used in video games, streaming media, and other situations where speed is a concern and some data loss is acceptable due to the data’s transitory nature.
A connection-oriented unicast protocol is frequently used. TCP, for example, is connection-oriented, with source and destination IP addresses and port numbers forming the virtual-circuit ID. Some insecure protocols, such as Asynchronous Transfer Mode and Frame Relay, are connection-oriented. Furthermore, certain connectionless protocols, such as IEEE 802.11, have a high degree of reliability.

Go back n protocol | data communication | bhanu priya

The transport layer establishes logical communication between application layer processes running on separate hosts, giving the impression that they are linked directly. The transport layer is located between the Network and Session layers in the OSI model. TCP/IP, on the other hand, positions the transport layer between the application and network layers, as seen in 3.1. In this chapter, we’ll look at the TCP/IP protocol suite’s transport layer protocols.
We covered the TCP/IP transport layer protocols User Datagram Protocol (UDP) and Transmission Control Protocol (TCP) in the previous two chapters (TCP). This chapter builds on the previous chapter’s discussion of UDP and TCP.
UDP is a connectionless protocol in the sense that it does not involve the establishment of a connection between application processes before data can be exchanged. UDP does not promise that messages will be sent to the intended recipient. Messages can also arrive in a jumbled order. This feature allows UDP to provide fast message transmission between applications at the cost of reliability.

The internet: packets, routing & reliability

How do packets get from point A to point B, the next logical question? This section looks at two separate examples to address this question: Let’s pause for a moment before moving on to these two delivery examples.
This privacy notice summarizes our commitment to privacy and explains how we gather, secure, use, and distribute personal information collected on this website. Other Pearson blogs, online goods and services, and services have their own privacy policy.
We collect the inquiry or query, as well as the sender’s name, contact information (email address, phone number, and mailing address), and any other information voluntarily sent to us through a Contact Us form or email. This knowledge is used to refer to the request and answer the query.
We collect order details, name, organization name and address (if applicable), email address, phone number, shipping and billing addresses, credit/debit card information, shipping options, and any instructions for orders and transactions made via our online store on this site. This information is used to complete purchases, fulfill orders, interact with customers who place orders or visit the online store, and for other similar purposes.

Communication network services-synchronous-asynchronous

A typical shared memory design is similar to that of a computer, with each pair of input/output ports serving as an input/output unit. A device interrupt is caused by an incoming packet at a port. The packet is copied from the transceiver to the device’s memory by the operating system. The processor runs a network stack, with the network layer looking up the routing table to see where the packet should be forwarded. The packet then works its way back down the stack to the output port of choice. The design’s drawback is that only one memory operation can be done at a time. Furthermore, a single CPU and a single system bus may become bottlenecks.
This design integrates a CPU on each line card to relieve the CPU bottleneck. This enables each line card to process the three layers of the stack and decide the packet’s destination port without the need for a central CPU, shared bus, or main (shared) memory. The control plane is handled by the central CPU, which creates the forwarding table and provides the administrative interface. Each line card receives a copy of the forwarding table. To allow the processor on one line card to copy the packet to another line card, shared memory and shared bus are used. Despite the fact that the bus is shared and only one memory operation can be performed at a time in shared memory, transferring packets between line cards can still be a bottleneck.

About the author


View all posts