“Data communication knowledge is extremely important. Without theoretical knowledge of data communication, the field of modern data communication will not be able to progress. In the previous article, the editor made a partial introduction to the principle of data communication. In this article, the editor will explain the rest of the data communication principle.

“

Data communication knowledge is extremely important. Without theoretical knowledge of data communication, the field of modern data communication will not be able to progress. In the previous article, the editor made a partial introduction to the principle of data communication. In this article, the editor will explain the rest of the data communication principle.

1. Data Link Protocol

The data link layer is the layer just above the hardware (physical) layer in the OSI protocol stack. Protocols in this layer manage the stream of bits between connected systems. Data packets from upper layers are encapsulated into frames and sent over the data link. It also uses flow control and error correction techniques. The data link layer handles point-to-point or point-to-multipoint links. In the OSI protocol stack, the higher network layers are responsible for handling connections that connect data links through multiple routers.

2. Framing

Framing technology is a technology used to allocate or mark channels within a bit stream, providing telecommunications with the means to select basic time slot structures and management methods, error isolation and segmented transmission protocols.

Framing provides a method of controlling the data bits transmitted over the physical medium. It provides error control and can provide data retransmission services depending on the type of service. The bit block is framed with the frame header and a checksum is appended so that corrupted frames can be detected. If a frame is corrupted or lost, it is only necessary to resend the frame without resending the entire data group.

Frames have a specific structure that varies depending on the data link used. The frame structure of the popular data link protocol called HDLC (High Level Data Link Control) is shown in the figure below. Note that the “info” field is where the data is put, and it can be of variable length. The “info” field can fit into an entire packet of information. The “start marker” field represents the start of the frame, the “address” field contains the destination address, the “control” field describes whether the information field contains data, command, or response, and the FCS field contains error detection codes.

Graphical HDLC frame format

3. Error detection and control

There are basically two types of error control methods, one is called “feedback error correction” and the other is called “forward error correction”. On the basis of these two types, a type called “hybrid error correction” is derived. For different types of channels, different error control techniques should be used, otherwise it will be less effective. Feedback error correction can be used for two-way data communication, and forward error correction is used for unidirectional digital signal transmission, such as broadcast digital television systems, because such systems do not have a feedback channel.

The data link layer is also responsible for error detection and control. One method of error control is to detect errors and then request retransmission. Another method is for the receiver to detect an error and then reconstruct the frame. The latter method requires that enough additional information is sent with the frame so that the receiver can reconstruct the frame after an error has been detected. This method is used when retransmission is not possible, such as when transmitting information to a space probe.

Performing error recovery tasks in the data link layer is generally inefficient. Many network implementations thus rely on upper layer protocols to accomplish this task. In most cases, the data link layer is used to transfer data as quickly and efficiently as possible without performing extensive data recovery tasks. The upper layer protocol provides recovery services.

4. Flow Control

Flow control is to control data flow between computers and network nodes to achieve the purpose of data synchronization. Too much data arriving before the device can handle it can cause data to be discarded or retransmitted. For serial data transmission, the Xon/Xoff protocol is used for control. In the network, flow control is also involved in adding new devices. When the traffic is large, new devices cannot be added.

Data transfer can be thought of as water flowing through a pipe and filling a bucket at the receiving end. The receiver takes water from the bucket, but needs some way to reduce the flow so that the bucket doesn’t overflow. In this analogy, the bucket represents the data buffer used by the receiver, which holds the incoming data that must be processed. The buffers on some NICs (Network Interface Cards) are large enough to hold the entire incoming transfer. If the buffer overflows, the frame is usually dropped, so it would be useful for the receiver to have some way of telling the sender to slow down or stop sending frames.

Shared LAN network access and logical link control access methods are necessary for networks shared by multiple devices. Because only one device can transmit on the network at a time, a medium access control method is needed to provide arbitration.

In the local area network environment defined by the IEEE, the media access protocol is located at a sublayer of the data link layer called the MAC (Medium Access Control) sublayer. The MAC sublayer is located below the LLC sublayer, which provides the data link for any MAC drivers installed below it.The sub-partitions of this layer can be seen in the image below

Figure Data Link Layer consists of two sublayers: MAC (Media Access Control) and LLC (Logical Link Control)

The MAC sublayer supports a variety of different network types, each of which has a specific method of arbitrating network access. The three possible access methods are as follows:

Carrier Sense Method The carrier sense technology means that before sending a frame, the sending station first monitors whether there is a carrier signal sent by other stations on the channel. If there is no other carrier, the signal can be sent; otherwise, the frame is delayed. Using this technique, devices listen to network transmissions and wait until the line is free to transmit their own data. If two stations attempt to transmit at the same time, both stations exit and wait an indeterminate amount of time before retransmitting.

Token access tokens are special packets that control network access in Token Ring, Token Bus, and Fiber Distributed Data Interface (FDDI) networks. Token Ring constitutes a logical ring in which each transmission travels along the ring from one station to another. Only stations with special tokens can transmit.

Reservation method In this scheme, each transmission device has a specific time slot or frequency assigned to it. TDM (Time Division Multiplexing) is an example. Devices can selectively place data into time slots for transmission. If the device does not transmit any data, this technique can waste bandwidth.

5. Bridging

“Bridging” refers to the process of forwarding network data packets according to the address of the link layer of the OSI network model. When a router is configured with bridging options, it processes all frames on all interfaces and investigates the location of each host in real time. If a frame is received on an interface, an entry will be built into a bridge, listing the host sending the data and the MAC address of the interface receiving the data frame, so that the routing table is continuously improved in the communication. Transparent bridging makes the router transparent to the host, and its role is equivalent to a LAN switch. If two hosts in the same LAN communicate, the data frame will not be sent to other interfaces, because in the bridge table, the data frames come from the same interface; if a frame is received, the MAC address is not in itself In the bridging table, the frame will be diffused to all interfaces, and the bridging will also diffuse all broadcast packets, occupying the effective bandwidth of the network and causing network congestion. Cisco IOS supports various types of bridging, such as: transparent bridging, encapsulating bridging, source-route bridging, source-route transparent bridging, and source-route translation bridging.

A bridge is a device that connects two or more network segments into a single LAN. All devices on the newly connected LAN can communicate with each other, but bridges provide filtering that prevents unwanted traffic from propagating from one network segment to another. Bridges are often used to separate a large LAN into two separate network segments. If the LAN is Ethernet, the bridge creates one broadcast domain and two collision domains. In Ethernet, it is better to have fewer computers in the collision domain, which facilitates dividing the network with bridges. Note that switches are basically multi-port bridges.

6. Swap

As mentioned above, bridges can be used to divide a LAN into two network segments, which in turn effectively create two smaller collision domains. A switch is an extension of the device based on this theory. Bridges usually have two ports connecting two LAN segments, while switches have a set of ports that can connect more segments. The following diagram illustrates how a switch provides bridging functionality for multiple hubs. Each hub has a collision domain, but the entire network shown in the diagram is a single broadcast domain. Each switch port is basically a separate LAN segment that can be “bridged” to any other port at any time through the internal circuitry in the switch.

Diagram showing a switched network

All the advantages of bridging are described in the previous section.

Most switching devices provide a way to configure VLANs (Virtual LANs). When building a network with switches, there is a tendency to build large flat networks rather than multiple distinct LANs (ie all nodes are part of the same broadcast network). VLAN technology can be used to create virtual LANs in a flat switching environment. For example, if the hub in the figure above is replaced with a VLAN-capable switch, workstations A and D can be configured into one VLAN; and workstations B, E, and H can be configured into another VLAN. The broadcast from A can be heard by D, and the broadcast from B can be heard by E and H. A router is then required to send data packets between VLANs.

7. Routing, network interconnection and network layer

While a bridge connects two separate LAN segments into a single broadcast domain (or splits a large LAN into two or more distinct collision domains), routers provide network interconnection. At the bridge level, information is sent to other systems in the form of frames (frames are defined in the data link layer). At the router level, the information must be encapsulated within data packets containing the destination network address and then forwarded across router boundaries. The router connects the network to the Internet.

Sometimes a letter can be mailed to someone in the same city by writing the specific address and the word “city” below the address. But if the letter has an “intercity” address, you will need to write the city name and ZLP code (postal code) on the envelope. Likewise, the Internet is made up of many interconnected networks. The Internet is the largest Internet network. To send data packets between different networks requires a hierarchical naming scheme, where each network is identified by a name or number for routing purposes. This is what the ZLP coding scheme does in the postal system. IP (Internet Protocol) is the addressing and routing protocol of the Internet.

In the diagram below, the LAN is connected to routers and the routers form an interconnected network of paths through which data packets can travel to their destinations. Note that any router and connected LAN in the network can be reached from any other point.

Diagram showing the network that routers use to create multiple connection points and multiple paths.

8. Transport layer

The service transport layer provides connection-oriented services. This means that two systems can establish a session through which they “talk” about the state of the data exchange. Although establishing a connection takes some time and adds some overhead to the data transfer, it provides the sender with the service of guaranteeing that the receiver receives all the data sent. The sender sends a set of data packets, and the receiver acknowledges that it has received the data packets. If the receiver does not acknowledge receipt, a retransmission data packet is sent. Session control also provides flow control to prevent receiver overflow or in some cases network overflow.

The following diagram illustrates how a transport layer session becomes a logical end-to-end connection across intermediary devices such as routers. Two peer transport layers talk over a connection-oriented virtual wire.

The transport layer is able to perform end-to-end translation between networks

The functions of the transport layer include whether to choose an error recovery protocol or a non-error recovery protocol, and to multiplex the input of data streams of different applications on the same host, and also to reorder the received packets out of sequence. For example: TCP, UDP, SPX, etc.

The transport layer provides reliable connection-oriented services. For example, if a network link fails temporarily, a connection-oriented session does not immediately abort the connection, but instead attempts to keep the connection alive until the underlying link is re-established. After the session is re-established, data continues from where it was interrupted.

9. Application layer

Applications running at the highest layers in the protocol stack are not actually involved in communication, but they do use communication devices and implement functionality in their user interface (using the underlying network). The role of the application layer is to complete a series of services required for business processing while implementing multiple system processes to communicate with each other. It not only provides the information exchange and remote operation required by the application process, but also acts as the user agent of the interacting application process. Network file sharing services such as NCP (NetWare Core Protocol), NFS (Network File System) in the UNIX environment, or SMB (Server Information Block) in the Windows environment were developed specifically to use network services so that users can shared documents.

The application layer is the only window through which the network can provide application services to end users, and its purpose is to support the requirements of user networking applications. Due to the different requirements of users, the application layer contains a variety of application entities that support different applications and provide a variety of application services, such as e-mail, file transfer, virtual terminal, Electronic data interchange and so on.

In a TCP/IP environment, the socket API provides a programming interface between applications and underlying network services.