Service Protocols
Service Protocols divide the data into packets when sending or receiving data from another computer in what is known as the packet switching technique. The beauty of this is that all of the installed protocols on the network will be able to connect to each other without any problem because most modern protocols are built to be background compatible with older protocols. What we are really interested in are IP addressing protocols and the TCP (Transmission Control Protocol) and UDP (User Datagram Protocol) transmission protocols.
TCP
Transmission Control Protocol (TCP) operating on the Transport Layer of the OSI Model is the most commonly used protocol in networks because it provides a guaranteed reliable connection between two machines. Most applications running on a network need to make sure that when they send a packet to the other end of the network, that packet will be received and then processed, so most applications request TCP. When you copy data over the network from Machine A to Machine B, TCP is used to ensure that the data has been copied fully without error. You also use TCP when you’re surfing the Web.
Some TCP applications retry requesting the data from the other machine for a certain time or certain number of attempts. If that fails, TCP informs you of the error. TCP needs to make sure that the packet being sent to the destination device will be received. Once the destination device receives the packet, it sends an acknowledgement that the packet was
received. After the ACK has been sent to Machine A, Machine B processes the request, and so on. For every packet that Machine A receives successfully, it informs Machine B that the packet was received successfully and it is ready for more packets. Usually a TCP confirmation is a blank confirmation without any error messages displayed on your screen. This means that if you use an application that uses TCP and you perform an operation that is completed without any error messages, that is your guarantee that the operation completed successfully.
Figure 1. Machine A tries to establish a TCP connection with Machine B
Here is the 3-step-handshake that occurs when the packet is carried over TCP from Machine A to Machine B:
- Machine A sends a TCP synchronized packet SYN). (First TCP Handshake)
- Machine B receives the SYN and sends an acknowledgement SYN-ACK to Machine
A. (Second TCP Handshake)
- Machine A receives the SYN-ACK from Machine B and sends back an ACK to Machine B. (Third TCP Handshake)
When Machine B receives the ACK from Machine A, the TCP connection is established between both machines. Both sides have to acknowledge the connection before it is established, which makes TCP a very reliable protocol. If Machine B can’t process all the packets sent from Machine A, B will ask A to slow down the process of sending data until it’s done processing. Once done, Machine B will send a signal back to A telling it to send the rest of the packets.
UDP
User Datagram Protocol (UDP) is the second most popular protocol after TCP. It’s used on applications that don’t need to guarantee delivery of the packet since it sends the packet and won’t acknowledge to the sender whether the packet has been received or not. When you watch the news online or listen to music on your favorite radio station online, these applications use UDP rather than TCP. When an application sends a UDP packet from Machine A to Machine B, it just sends the packet without an acknowledgement. This has one big disadvantage because there is no way of knowing whether Machine B received the packet, but transmission is faster because the transmissions overhead is eliminated, but UDP’s disadvantage is also its biggest advantage. Since UDP doesn’t use a 3-step- handshake, packets sent over UDP carry more data than TCP packets with less overhead in the transmission allowing the packet to get processed in less time. This is why UDP is used in real time applications when you want to compromise reliability for speed.
Figure 2. UDP connection
(Voice over Internet Protocol) VOIP is a real time application that uses UDP.
TCP Windowing
To solve the problem of TCP’s delay the concept TCP windowing was introduced. TCP windowing increases the amount of data that can be sent to the receiver before the receiver sends an acknowledgement that it received the data.
Table 1. TCP Communication without Windowing in Which Machine A Tries to Send 640 KB of Information to Machine B over TCP (3-Way-Handshake)
| Machine A | Machine B |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
For us to send 640 KB, we would need to send 10 SYNs and receive 10 ACKs. With TCP windowing, instead of sending 10 SYNs to Machine B, Machine A starts by sending the first 64 KB of information. When Machine B receives and acknowledges the initial packet, TCP windowing will increase the size of the SYNs, so instead of sending 64 KBs only, it will send 128 KB. When Machine B receives it and sends the acknowledgement back to Machine A, TCP windowing will increase the SYN size to 256 KB and so on until all the data has been received. This will save more time since it will transfer the data faster than a regular TCP 3-way-handshake. Compare Table 1 and Table 2 to see the difference.
Table 2. TCP Communication with Windowing
| Machine A | Machine B |
| Sending a SYN with 64 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 128 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 256 KB of Data | Acknowledged. Send the next one. |
| Sending a SYN with 192 KB of Data | Acknowledged. Send the next one. |
Using TCP windowing it took the data 4 SYNs instead of 10 for the transfer, which is why TCP is faster than we thought when we saw how the TCP 3-way-handshake works. It contradicts everything we knew about how fast we get things done over TCP. After explaining what TCP windowing does to the TCP connection, it makes more sense to think of our TCP connection as a reliable fast connection. TCP windowing’s primary job is congestion control where if a host sends data over the network with too much speed and the receiver host can’t acknowledge all of the data because of speed mismatch, TCP windowing will slow the sender’s speed down until the receiver can acknowledge all the data it received. Once the data has been received and acknowledged, TCP Windowing will speed up the SYNs once again.
IP Addressing
The IP address contains four fields separated by dots. Each field can hold any value from 0-255; an IP address could have any of the following values:
- 1.1.1.1
- 24.55.155.245
- 125.85.251.253
- 192.168.0.16
The general format of an IP address is: A.B.C.D. All devices on the network must have an IP address to communicate. IP addresses can be assigned by many methods, such as automatically using DHCP (server that leases IP addresses automatically), or manually by entering the IP address in the TCP/IP properties of your local area connection