OSI seven-layer and TCP five-layer protocols, why TCP/IP protocol wins

  [[278277]]

1. OSI Reference Model

1. Origin of OSI

First of all, what is the OSI seven-layer model?

OSI is the abbreviation of Open System Interconnection in English. It is translated into Chinese as (Open System Interconnection Reference Model). It is a standard system developed by an international standardization organization for interconnection between computer or communication systems.

How to understand it? You can think of it this way. A long time ago, there were various manufacturers of computer hardware and network equipment in the world. Everyone did their own thing, which was very messy and incompatible with each other. At that time, there was a very powerful and recognized organization in the world. In order to unify the Internet, make it less complicated and easier to manage, it formulated a rule, and everyone must abide by this rule. This is the OSI seven-layer model.

2. What do the seven layers mean?

First of all, this seven-layer model is a virtual one, that is, a theoretical rule, not a tangible physical shape like a car model. So what does it look like? Look at the picture below.

[This is purely hand-typed, to preach for the teacher. I don't seek great achievements, but I hope to help everyone. In my spare time, I will share some work experience and various technical resource articles with everyone, hoping to help everyone make progress together. I also hope that my releases will be helpful to you. There are more similar articles in my releases that you can check out. If you like them, please follow me. Thank you. Your attention is the biggest motivation for me to keep releasing.]

3. What are these two seven-layer models?

First, you should think of the seven-layer model on both sides as two people writing letters. If you write to someone, does he have to write back to you? This is the arrow in the middle of the two models. It means that there is a back and forth. The letters between you two have to go through these seven steps before they can reach each other. This is why the seven-layer model is drawn on both sides.

4. What are the basic functions of the OSI seven-layer model?

The OSI model uses a seven-layered structure model to enable reliable communication between different systems and networks. Therefore, its most important function is to help realize data transmission between different types of hosts.

5. So what do these seven layers correspond to?

(1) Physical layer: It corresponds to network cards, network cables, optical fibers, etc.

(2) Data link layer: It corresponds to bridges, switches, etc.

(3) Network layer: routers, layer 3 switches

(4) Transport layer, session layer, presentation layer, application layer, these are at the software level and virtual level, so they do not correspond to related physical devices. These four levels correspond to related protocols.

For example: the transport layer has: TCP UDP protocol; the application layer has http ftp protocol, etc., that's what it means

However, the actual network management requires daily maintenance of only five layers: physical layer, data link layer, network layer, and application layer.

6. Definition of each layer’s functions

Here we only give a general explanation of the functions of each OSI layer, without going into detail, because each layer is actually a complex layer.

Let’s take a brief look at it here.

We start from the top layer - the application layer.

The whole process is explained using the example of a commercial quotation sent between Company A and Company B.

(1) Application layer

The layer closest to the user in the OSI reference model provides application interfaces for computer users and also directly provides users with various network services.

Our common application layer network service protocols are: HTTP, HTTPS, FTP, POP3, SMTP, etc.

In fact, the boss of Company A is the user we are talking about, and the commercial quotation he wants to send is a network service provided by the application layer. Of course, the boss can also choose other services, such as sending a commercial contract, sending an inquiry, and so on.

(2) Presentation Layer

The presentation layer provides various encoding and conversion functions for application layer data, ensuring that the data sent by the application layer of one system can be recognized by the application layer of another system.

If necessary, this layer provides a standard representation for converting the various data formats within the computer into a standard representation used in communications.

Data compression and encryption are also one of the transformation functions that the presentation layer can provide.

Since Company A and Company B are from different countries, they have agreed to use English as the language of communication. Therefore, the presentation layer (the company's secretary) is responsible for translating the information transmitted by the application layer into English.

At the same time, in order to prevent other companies from seeing it, people from Company A will also encrypt the quotation.

This is the role of representation, converting and translating data at the application layer.

(3) Session layer

The session layer is responsible for establishing, managing, and terminating communication sessions between presentation layer entities.

Communication at this layer consists of service requests and responses between applications in different devices.

The colleagues at the session layer get the converted data from the colleagues at the presentation layer (the colleagues at the session layer are similar to the external relations department of the company). The colleagues at the session layer may have the contact information of the company and many other companies. Here, the company is the entity in the actual transmission process. They need to manage the contact sessions between the company and many external companies.

After receiving the data from the presentation layer, the session layer will establish and record this session. He will first find the address information of Company B, then put the entire document into an envelope and write the address and contact information. Prepare to mail the document.

Once it is confirmed that Company B has received the quotation, the conversation is over and the colleagues in the External Relations Department will terminate the conversation.

(4) Transport layer

The transport layer establishes an end-to-end link between hosts. The role of the transport layer is to provide end-to-end reliable and transparent data transmission services for upper-layer protocols, including handling issues such as error control and flow control.

This layer shields the details of lower-layer data communications from higher layers, so that high-level users only see a host-to-host, reliable data path between two transmission entities that can be controlled and set by the user.

We usually say that TCP UDP is at this layer. The port number is the "end" here.

The transport layer is equivalent to the person in the company who is responsible for sending and receiving express mail, the company's own deliveryman, who is responsible for delivering the information to be sent from the upper layer to the express company or post office.

(5) Network layer

This layer establishes a connection between two nodes through IP addressing, selects appropriate routing and switching nodes for packets sent by the transport layer at the source end, and transmits them to the transport layer at the destination end correctly according to the address.

It is commonly referred to as the IP layer.

This layer is what we often call the IP protocol layer.

The IP protocol is the foundation of the Internet.

The network layer is equivalent to the express delivery company's huge express delivery network, with different distribution centers across the country. For example, SF Express from Shenzhen to Beijing (for land transportation, it seems that air transportation can fly directly to Beijing) must first go to SF Express's Shenzhen distribution center, then be sent from the Shenzhen distribution center to the Wuhan distribution center, and then from the Wuhan distribution center to the Beijing Shunyi distribution center. Each distribution center is equivalent to an IP node in the network.

(6) Data Link Layer

It combines bits into bytes, and bytes into frames, uses link layer addresses (Ethernet uses MAC addresses) to access the medium, and performs error detection.

The data link layer is divided into two sublayers: logical link control sublayer (LLC) and media access control sublayer (MAC).

The MAC sublayer handles the CSMA/CD algorithm, data error checking, framing, etc.; the LLC sublayer defines some fields that enable the previous protocol to share the data link layer.

In actual use, the LLC sublayer is not necessary.

(7) Physical layer

The actual transmission of the final signal is achieved through the physical layer.

Transmits a bit stream over a physical medium. Specifies voltage levels, speeds, and cable pinouts.

Commonly used equipment includes (various physical devices) hubs, repeaters, modems, network cables, twisted pair cables, and coaxial cables.

These are the transmission media at the physical layer.

The means of transportation used in the express delivery process are equivalent to our physical layer, such as cars, trains, planes, and ships.

7. Communication characteristics: peer-to-peer communication

Peer-to-peer communication: In order for data packets to be transmitted from the source to the destination, each layer of the OSI model at the source end must communicate with the peer layer at the destination end. This type of communication is called peer-to-peer communication.

During the communication process at each layer, the layer's own protocol is used for communication.

2. TCP/IP five-layer model

The correspondence between the TCP/IP five-layer protocol and the OSI seven-layer protocol is as follows:

Different devices work at each layer. For example, the switches we commonly use work at the data link layer, and general routers work at the network layer.

The protocols implemented at each layer are also different, that is, the services at each layer are also different. The following figure lists the main protocols at each layer.

1. Application layer

The application layer provides services to applications. These services are grouped according to the characteristics they provide to applications and are called service elements.

Some are used by multiple applications, while others are used by a smaller class of applications.

The application layer is the highest layer and directly provides services for the application process.

Its function is to complete a series of services required for business processing while enabling multiple system application processes to communicate with each other.

Its service elements are divided into two categories: common application service elements CASE and specific application service elements SASE.

CASE provides the most basic services. It becomes the user of any user and any service element in the application layer, and mainly provides basic control mechanisms for application process communication and distributed system implementation.

Specific service SASE needs to meet some specific services, such as document transmission, access management, job transmission, banking transactions, order entry, etc.

These will involve virtual terminals, job transfer and operation, file transfer and access management, remote database access, graphics core systems, open systems interconnection management, etc.

The application layer standards include DP8649 "Common Application Service Elements", DP8650 "Common Application Service Elements Protocol", file transfer, access and management services and protocols.

2. Transport layer

The transport layer is the first end-to-end layer when two computers communicate data over a network and has a buffering effect.

When the network layer service quality cannot meet the requirements, it will improve the service to meet the requirements of the higher layer; when the network layer service quality is good, it only needs to do very little work.

The transport layer can also perform multiplexing, which is to create multiple logical connections on one network connection.

The transport layer, also known as the transport layer, is a very important layer.

Because it is the last layer that controls data transmission from source to destination from low to high.

It is an established fact that there are great differences in performance among various communication subnets in the world.

For example, telephone switching networks, packet switching networks, public data switching networks, local area networks and other communication subnets can be interconnected, but the throughput, transmission rate, data delay and communication costs they provide are different.

For the session layer, an interface with constant performance is required.

The transport layer takes on this function.

It uses splitting/merging, multiplexing/demultiplexing technology to adjust the differences of the above communication subnets so that the session layer cannot feel it.

In addition, the transport layer must also have functions such as error recovery and flow control to shield the session layer from the details and differences of the communication subnets in these aspects.

The data objects faced by the transport layer are no longer network addresses and host addresses, but interface ports with the session layer.

The ultimate goal of the above functions is to provide reliable and error-free data transmission for the session.

The service of the transport layer generally goes through three stages: transport connection establishment stage, data transmission stage, and transport connection release stage to complete a complete service process.

The data transmission stage is divided into general data transmission and accelerated data transmission.

3. Network layer

The emergence of the network layer is also the result of network development.

In the context of online systems and circuit switching, the functions of the network layer do not make much sense.

As the number of data terminals increases, they are connected by relay devices.

At this time, there will be a situation where a terminal requires to communicate not only with a single terminal but with multiple terminals. This is where the problem of linking the data of any two data terminal devices arises, which is routing or path finding.

In addition, after a physical channel is established and used by a pair of users, a lot of idle time is often wasted.

People naturally want to let multiple pairs of users share one link. To solve this problem, logical channel technology and virtual circuit technology emerged.

In order to establish network connections and provide services to upper layers, the network layer should have the following main functions:

· Routing and relaying.

Activate and terminate network connection.

Multiplex multiple network connections on one data link, often using time-division multiplexing technology.

· Error detection and recovery.

· Sorting, flow control.

Service selection.

Network management.

4. Data Link Layer

A data link can be roughly understood as a data channel.

The physical layer provides transmission media and connections for data communication between terminal devices.

Media is long-lived and connections have a lifespan.

During the lifetime of a connection, the sending and receiving ends can perform data communication one or more times.

Each communication must go through two processes: establishing communication contact and dismantling communication contact.

This established data sending and receiving relationship is called a data link.

However, data transmitted on physical media is inevitably affected by various unreliable factors and may cause errors. In order to make up for the deficiencies in the physical layer and provide error-free data transmission for the upper layer, it is necessary to be able to detect and correct errors in the data.

The establishment and dismantling of data links, as well as error detection and correction of data are the basic tasks of the data link layer.

The main function of the link layer: The link layer provides data transmission services for the network layer, and this service depends on the functions of this layer.

· Establishment, removal, and separation of link connections.

Frame delimitation and frame synchronization. The data transmission unit of the link layer is the frame. The length and interface of the frame vary depending on the protocol, but the frame must be delimited anyway.

Sequence control, which controls the order in which frames are sent and received.

· Error detection and recovery.

There are also link identification, flow control, etc.

Error detection often uses square code check and cyclic code check to detect bit errors in data on the channel, while frame loss and other conditions are detected using sequence number detection.

Recovery from various errors is often accomplished through feedback retransmission technology.

5. Physical layer

Although the physical layer is at the bottom layer, it is the foundation of the entire computer network.

The physical layer provides transmission media and interconnection equipment for data communication between devices, and provides a reliable environment for data transmission.

The media of the physical layer include overhead wires, balanced cables, optical fibers, wireless channels, etc. The interconnection equipment used for communication refers to the interconnection equipment between DTE and DCE.

DTE stands for data terminal equipment, also known as physical equipment, including computers, terminals, etc.

DCE is data communication equipment or circuit connection equipment, such as a modem.

Data transmission usually goes through the path of DTE-DCE and then DCE-DTE.

Interconnection equipment refers to the devices that connect DTE and DCE, such as various plugs and sockets.

Various thick and thin coaxial cables, T-connectors, plugs, receivers, transmitters, repeaters, etc. in the LAN are all media and connectors at the physical layer.

The main functions of the physical layer are:

(1) Provide a data transmission path for data terminal devices. The data path can be a single physical medium or a connection of multiple physical media.

A complete data transmission includes activating the physical connection, transmitting data, and terminating the physical connection.

Activation means that no matter how many physical media are involved, they must be connected between the two communicating data terminal devices to form a channel.

(2) To transmit data, the physical layer must form an entity suitable for data transmission needs and serve data transmission.

The first is to ensure that data can pass through it correctly, and the second is to provide sufficient bandwidth (bandwidth refers to the number of bits that can pass through per second) to reduce congestion on the channel.

The data transmission method can meet the needs of point-to-point, point-to-multipoint, serial or parallel, half-duplex or full-duplex, synchronous or asynchronous transmission.

3. TCP/IP Basics

1. The specific meaning of TCP/IP

In a literal sense, some people may think that TCP/IP refers to the two protocols TCP and IP. In real life, sometimes it does refer to these two protocols. However, in many cases, it is just a general term for the protocol group that must be used when using IP for communication. Specifically, IP or ICMP, TCP or UDP, TELNET or FTP, and HTTP are all TCP/IP protocols. They are closely related to TCP or IP and are an indispensable part of the Internet. The term TCP/IP refers to these protocols in general, so TCP/IP is sometimes called the Internet Protocol Group.

When the Internet communicates, a corresponding network protocol is required. TCP/IP is a protocol family originally developed for the use of the Internet. Therefore, the Internet protocol is TCP/IP, and TCP/IP is the Internet protocol.

Internet Protocol Suite

2. Data Packet

Packet, frame, data packet, segment, message

The above five terms are all used to express the units of data, which can be roughly divided into the following categories:

• Bag can be said to be a catch-all term;
• Frame is used to represent the unit of packet in the data link layer;
• Data packets are the unit of packets in layers above the network layer such as IP and UDP;
• Segments represent information in a TCP data stream;
• A message is a unit of data in an application protocol.

In each layer, a header is added to the data to be sent. This header contains the necessary information for that layer, such as the destination address and protocol-related information. Usually, the information provided to the protocol is the packet header, and the content to be sent is the data. From the perspective of the next layer, all packets received from the previous layer are considered to be data of this layer.

Packet header

The data packets transmitted in the network are composed of two parts: one is the header used by the protocol, and the other is the data transmitted from the upper layer. The structure of the header is defined in detail by the specific specifications of the protocol. In the header of the data packet, it is clearly indicated how the protocol should read the data. Conversely, by seeing the header, you can understand the necessary information of the protocol and the data to be processed. The packet header is like the face of the protocol.

3. Data processing flow

The following figure takes user a sending an email to user b as an example:

Data processing flow

① Application Processing

• First, the application will perform encoding processing, which is equivalent to the presentation layer function of OSI;
• After the encoding conversion, the email may not be sent out immediately. This management function of when to establish a communication connection and when to send data is equivalent to the session layer function of OSI.

② TCP module processing

• TCP is responsible for establishing connections, sending data, and disconnecting connections according to the instructions of the application. TCP provides reliable transmission of data sent from the application layer to the other end. In order to achieve this function, a TCP header needs to be attached to the front end of the application layer data.

③ Processing of IP modules

• IP combines the TCP header and TCP data transmitted by TCP as its own data, and adds its own IP header to the front of the TCP header. After the IP packet is generated, the routing control table is referenced to determine the route or host that receives the IP packet.

④ Network interface (Ethernet driver) processing

• The IP packets transmitted from IP are data for Ethernet. The Ethernet header is attached to these data and sent, and the generated Ethernet data packets will be transmitted to the receiving end through the physical layer.

⑤ Network interface (Ethernet driver) processing

• After receiving an Ethernet packet, the host first finds the MAC address from the Ethernet packet header to determine whether it is a packet sent to itself. If not, the data is discarded.
• If it is a packet sent to itself, the data type is determined from the type in the Ethernet packet header and then passed to the corresponding module, such as IP, ARP, etc. The example here is IP.

⑥ Processing of IP modules

• The IP module also performs similar processing after receiving the data. It determines from the packet header whether the IP address matches its own IP address. If it matches, it sends the data to the corresponding module according to the protocol type in the header, such as TCP, UDP. The example here is TCP.
• In addition, in the case of a router, the receiving address is often not your own address. At this time, you need to use the routing control table to investigate the host or router to which the data should be sent before forwarding the data.

⑦ TCP module processing

• In the TCP module, the checksum is first calculated to determine whether the data is corrupted. Then it is checked whether the data is received according to the sequence number. Finally, the port number is checked to determine the specific application. After the data is completely received, it will be passed to the application identified by the port number.

⑧ Application Processing

The receiving application will directly receive the data sent by the sending end, parse the data, and display the corresponding content.

4. TCP and UDP in the transport layer

There are two representative transport layer protocols in TCP/IP, namely TCP and UDP.

• TCP is a connection-oriented, reliable streaming protocol. Stream refers to an uninterrupted data structure. When an application uses TCP to send messages, although the order of sending can be guaranteed, it is still like a data stream without any intervals sent to the receiving end. In order to provide reliable transmission, TCP implements "sequence control" or "retransmission control" mechanism. In addition, it also has many functions such as "flow control (traffic control)", "congestion control", and improving network utilization.
• UDP is an unreliable datagram protocol. It will be handed over to the upper-layer application to complete the subtle processing. In the case of UDP, although the size of the message can be guaranteed, it cannot guarantee that the message will arrive. Therefore, the application sometimes performs retransmission processing according to its own needs.
• The advantages and disadvantages of TCP and UDP cannot be simply and absolutely compared: TCP is used when reliable transmission is necessary at the transport layer; on the other hand, UDP is mainly used for communications or broadcast communications that have high requirements for high-speed transmission and real-time performance. TCP and UDP should be used as needed according to the purpose of the application.

1. Port number

The addresses in data links and IP refer to MAC addresses and IP addresses respectively. The former is used to identify different computers in the same link, and the latter is used to identify interconnected hosts and routers in a TCP/IP network. There is also a concept similar to addresses in the transport layer, which is the port number. The port number is used to identify different applications communicating in the same computer. Therefore, it is also called a program address.

1.1 Identify applications by port number

Multiple programs can run on a computer at the same time. The transport layer protocol uses these port numbers to identify the applications that are communicating on the local machine and accurately transmit data.

Identify applications by port number

1.2 Communication identification by IP address, port number, and protocol number

• It is not enough to identify a communication by its destination port number alone.

Communication identification through port number, IP address, and protocol number

• The communication between ① and ② is carried out on two computers. Their destination port numbers are the same, both 80. Here they can be distinguished based on the source port number.
• The destination port number and source port number of ③ and ① are exactly the same, but their respective source IP addresses are different.
• In addition, when the IP addresses and port numbers are all the same, we can also distinguish them by the protocol number (TCP and UDP).

1.3 Determination of port number

• Standardized port number: This method is also called static method. It means that each application has its own designated port number. But it does not mean that any port number can be used at will. For example, the port numbers used in widely used application protocols such as HTTP, FTP, and TELNET are fixed. These port numbers are called well-known port numbers, which are distributed between 0 and 1023. In addition to the well-known port numbers, there are also some port numbers that are officially registered, which are distributed between 1024 and 49151, but these port numbers can be used for any communication purpose.
• Sequential allocation method: The server needs to determine the listening port number, but the client receiving the service does not need to determine the port number. In this method, the client application does not need to set the port number by itself, but leaves it to the operating system to allocate. The dynamically allocated port number ranges from 49152 to 65535.

1.4 Port Number and Protocol

The port number is determined by the transport layer protocol it uses. Therefore, different transport layer protocols can use the same port number.

In addition, those well-known port numbers have nothing to do with the transport layer protocol. As long as the port is the same, the same application will be assigned to handle it.

2. UDP

UDP does not provide complex control mechanisms and uses IP to provide connectionless communication services.

• And it is a mechanism that sends the data sent by the application to the network as it is at the moment of receipt. Even if there is network congestion, UDP cannot perform flow control and other behaviors to avoid network congestion.
• In addition, if packet loss occurs during transmission, UDP is not responsible for retransmission.
• There is no correction function even when packets arrive out of order.
• If the above detailed control is required, it has to be handled by the application using UDP.
• UDP is often used in the following aspects: 1. Communications with a small total packet volume (DNS, SNMP, etc.); 2. Multimedia communications such as video and audio (instant messaging); 3. Application communications limited to specific networks such as LAN; 4. Broadcast communications (broadcast, multicast).

3. TCP

• TCP is quite different from UDP. It fully implements various control functions during data transmission, can perform retransmission control when packets are lost, and can also perform sequence control on out-of-order packets. These are not available in UDP.
• In addition, as a connection-oriented protocol, TCP will only send data when it confirms the existence of the other party, thereby controlling the waste of communication traffic.
• Based on these mechanisms of TCP, high-reliability communication can be achieved on a connectionless network such as IP (mainly through mechanisms such as checksum, sequence number, confirmation response, retransmission control, connection management and window control).

3.1 Three-way handshake (key points)

• TCP provides connection-oriented communication transmission. Connection-oriented means that preparations between the two ends must be done before data communication begins.
• The so-called three-way handshake means that when establishing a TCP connection, the client and server need to send a total of three packets to confirm the establishment of the connection. In socket programming, this process is triggered by the client executing connect.

Let's take a look at the flowchart of the three-way handshake:

Three-way handshake

• First handshake: The client sets the SYN flag to 1, randomly generates a value seq=J, and sends the data packet to the server. The client enters the SYN_SENT state and waits for confirmation from the server.
• Second handshake: After receiving the data packet, the server knows from the flag bit SYN=1 that the client requests to establish a connection. The server sets the flag bits SYN and ACK to 1, ack=J+1, randomly generates a value seq=K, and sends the data packet to the client to confirm the connection request. The server enters the SYN_RCVD state.
• The third handshake: After the client receives the confirmation, it checks whether ack is J+1 and ACK is 1. If correct, it sets the flag ACK to 1, ack=K+1, and sends the data packet to the server. The server checks whether ack is K+1 and ACK is 1. If correct, the connection is successfully established, the client and server enter the ESTABLISHED state, and the three-way handshake is completed. Then the client and server can start transmitting data.

3.2 Four waves (key points)

• Four waves terminate the TCP connection, which means that when a TCP connection is disconnected, the client and the server need to send a total of 4 packets to confirm the disconnection. In socket programming, this process is triggered by either the client or the server executing close.
• Since TCP connection is full-duplex, each direction must be closed separately. The principle is that when one party completes the data sending task, it sends a FIN to terminate the connection in this direction. Receiving a FIN only means that there is no data flow in this direction, that is, no more data will be received, but data can still be sent on this TCP connection until FIN is sent in this direction. The party that closes first will perform an active close, while the other party will perform a passive close.

Let's take a look at the flowchart of four waves:

Four waves

• The disconnected end can be either the client or the server.
• First handshake: The client sends a FIN=M to close the data transmission from the client to the server, and the client enters the FIN_WAIT_1 state. This means "I have no data to send to you from the client", but if your server still has data to send, you don't have to rush to close the connection and can continue to send data.
• Second wave: After receiving FIN, the server first sends ack=M+1 to tell the client that I have received your request, but I am not ready yet, please continue to wait for my message. At this time, the client enters the FIN_WAIT_2 state and continues to wait for the FIN message from the server.
• The third wave: When the server determines that the data has been sent, it sends a FIN=N message to the client, telling the client that the data has been sent and the connection is ready to be closed. The server enters the LAST_ACK state.
• Fourth handshake: After the client receives the FIN=N message, it knows that it can close the connection, but it still does not trust the network and is afraid that the server does not know to close it, so it sends ack=N+1 and enters the TIME_WAIT state. If the server does not receive ACK, it can retransmit. After the server receives ACK, it knows that it can disconnect. After the client waits for 2MSL and still does not receive a reply, it proves that the server has closed normally. Well, my client can also close the connection. Finally, the four-way handshake is completed.

The above is a case where one party actively closes and the other party passively closes. In practice, there are also cases where both parties initiate active closures at the same time.

The specific process is as follows:

Wave at the same time

3.3 Improving reliability through sequence numbers and confirmation responses

• In TCP, when the data from the sender reaches the receiving host, the receiving host will return a notification that the message has been received. This message is called an acknowledgment (ACK). After the sender sends the data, it will wait for the acknowledgment from the other end. If there is an acknowledgment, it means that the data has successfully reached the other end. Otherwise, there is a high possibility of data loss.
• If the sender does not receive a confirmation response within a certain period of time, it can assume that the data has been lost and resend it. In this way, even if packet loss occurs, it can still ensure that the data can reach the other end and achieve reliable transmission.
• Failure to receive a confirmation response does not necessarily mean that the data has been lost. It is also possible that the data has been received by the other party, but the returned confirmation response was lost on the way. This situation may also cause the sender to mistakenly believe that the data has not reached the destination and resend the data.
• In addition, it is also possible that the confirmation response is delayed due to some other reasons, and it is not uncommon for it to arrive after the source host resends the data. In this case, the source host only needs to resend the data according to the mechanism.
• It is undesirable for the target host to receive the same data repeatedly. In order to provide reliable transmission for upper-layer applications, the target host must discard duplicate data packets. For this reason, we introduce sequence numbers.

The sequence number is a number that is assigned to each byte (8-bit byte) of the sent data in order. The receiving end queries the sequence number and the length of the data in the TCP header of the received data, and sends back the sequence number that it should receive next as a confirmation response. Through the sequence number and confirmation response number, TCP can identify whether the data has been received and whether it needs to be received, thereby achieving reliable transmission.

Sequence Number and Confirmation Response

3.4 Determination of retransmission timeout

• The retransmission timeout is the specific time interval for waiting for the confirmation response before retransmitting the data. If the confirmation response is not received after this time, the sender will retransmit the data. Ideally, a minimum time is found that can guarantee that the confirmation response will be returned within this time.
• TCP requires high-performance communication regardless of the network environment, and this feature must be maintained regardless of changes in network congestion. To this end, it calculates the round-trip time and its deviation every time a packet is sent. Adding the round-trip time and the deviation time, the retransmission timeout is a value slightly larger than this sum.
• In BSD Unix and Windows systems, the timeout is controlled in units of 0.5 seconds, so the retransmission timeout is an integer multiple of 0.5 seconds. However, the default value of the retransmission timeout is generally set to about 6 seconds.
• If the data is resent and still no confirmation response is received, it will be sent again. At this time, the time to wait for the confirmation response will be extended by 2 times and 4 times the exponential function.
• In addition, data will not be resent indefinitely and repeatedly. If there is still no confirmation response after a certain number of resends, it will be judged that an abnormality has occurred in the network or the other host, and the connection will be forcibly closed. The application will also be notified that the communication abnormality has been forcibly terminated.

3.5 Sending Data in Segments

• When establishing a TCP connection, you can also determine the unit of sending data packets, which we can also call the "maximum message size" (MSS). Ideally, the maximum message length is exactly the maximum data length in IP that will not be fragmented.
• When TCP transmits a large amount of data, it divides the data into segments based on the size of the MSS and retransmits the data in units of the MSS.
• The MSS is calculated between the two hosts during the three-way handshake. When the hosts at both ends send a request to establish a connection, they will write the MSS option in the TCP header to tell the other party the size of the MSS that their interface can adapt to. Then they will choose a smaller value between the two and put it into use.

3.6 Using Window Control to Improve Speed

• TCP uses one segment as a unit and performs a confirmation response process every time a segment is sent. This transmission method has a disadvantage that the longer the round-trip time of the packet, the lower the communication performance.
• To solve this problem, TCP introduced the concept of window. Acknowledgement is no longer confirmed in each segment, but in larger units, and the forwarding time will be greatly shortened. That is to say, the sending host does not need to wait for acknowledgement after sending a segment, but continues to send. As shown in the figure below:

·

• Window Control
• Window size refers to the maximum value that can continue to send data without waiting for acknowledgement. The window size in the figure above is 4 segments. This mechanism implements the function of using a large number of buffers and confirming responses to multiple segments at the same time.

3.7 Sliding window control

Sliding Window

• The data in the window in the figure above can be sent out even if there is no acknowledgment. However, before the acknowledgment of the entire window arrives, if some of the data are lost, the sender is still responsible for retransmission. To this end, the sender host needs to set a cache to retain the data to be retransmitted until they receive their acknowledgment.
• The part outside the sliding window includes unsent data and data that has been confirmed to have been received by the peer. When the data is sent out, if the acknowledgement is received as scheduled, you can no longer resend it, and the data can be cleared from the cache area.
• When acknowledgement is received, slide the window to the position of the serial number in the confirmation response. This allows multiple segments to be sent simultaneously in sequence to improve communication performance. This mechanism is also also known as sliding window control.

3.8 Retransmission control in window control

In the use window control, packet loss is generally divided into two situations:

• ① Confirm that the reply cannot be returned. In this case, the data has reached the peer and does not need to be resent, as shown in the figure below:

Partial confirmation response is lost

• ② The situation of a certain segment being lost. If the receiving host receives data other than the serial number it should receive, it will return a confirmation response for the data it has received so far. As shown in the figure below, when a certain segment is lost, the sending end will receive a confirmation response with the serial number of 1001. Therefore, when the window is larger and the message segment is lost, the confirmation response of the same serial number will be repeatedly returned. If the sending host receives the same acknowledge response three times in a row, it will resend its corresponding data. This mechanism is more efficient than the timeout management mentioned earlier, so it is also called high-speed retransmission control.