Reliability refers to the accuracy of received information, which is a symbol of the quality of information transmitted by the communication system. The indicator to measure the reliability of digital communication system can be expressed by the probability of signal error during transmission, that is, it is measured by the error rate. The larger the error rate, the worse the system reliability. Error rate is usually expressed in the following two ways. (1) Bit error rate P. The bit error rate refers to the proportion of the number of symbols that have errors during transmission to the total number of symbols transmitted. To be more precise, the bit error rate is the probability that symbols are transmitted incorrectly in the transmission system. It can be expressed as P by expression. =number of received error symbols/total number of transmitted symbols (2) Error rate P. The error rate refers to the proportion of the amount of information that has errors during the transmission process to the total amount of information transmission. In other words, it is the amount of information of the code element that is lost in the transmission system. Probability. The expression can be expressed as P.=Number of erroneous information/Total number of transmitted information. The effectiveness and reliability of the communication system are interrelated and mutually restricted. The reliability of the system can be improved by reducing the effectiveness. sex, or vice versa.

Simplex communication means that the information transmitted is always in one direction and is not transmitted in the opposite direction. As shown in Figure 2-10a, assuming A is the sending terminal and B is the receiving terminal, data can only be transmitted from A to B, but not from B to A. Simplex communication lines generally use a two-wire system, such as the transmission of radio broadcast and television signals.

Half-duplex communication means that information flow can be transmitted in both directions, but is limited to one direction at a time. As shown in Figure 2-10b, information can be transmitted from A to B or from B to A, so both communicating parties have transmitters and receivers. To achieve two-way communication, the channel direction must be changed. Half-duplex communication uses a two-wire line. When station A sends information to station B, station A connects the transmitter to the channel, and station B connects the receiver to the channel; and when station B sends information to station A, Station B will disconnect the receiver from the channel and connect the transmitter to the channel. Station A will also disconnect the transmitter from the channel and connect the receiver to the channel. This method of converting on one channel to achieve communication in two directions, A→B and B→A, is called half-duplex communication. Half-duplex communication is often used in industrial data communications. For example, walkie-talkies use this communication method to transmit information.

Full-duplex communication means that the communication system can perform two-way communication at the same time as shown in Figure 2-10c. It is equivalent to combining two simplex communication methods in opposite directions. This method is often used for communication between computers. For example, EIA-232 and EIA-422 use full-duplex communication.

①Redundancy A simple way to implement error checking is to send redundant data. ②Loopback Loopback is used in communication systems where operators manually enter data from the keyboard. ③Exact technical encoding When using precise technical encoding, the number of 1's in each character is the same. ④Parity check Parity check is often used for error checking in communications. ⑤ Find the checksum. This error checking method is to find an error checking byte in the communication data. ⑥ Cyclic redundancy check (CRC) Cyclic redundancy check performs a division operation on the transmission sequence, and appends the remainder of the division operation to the end of the transmission information.

The CRC method treats the data bit sequence to be sent as the coefficient of a polynomial f(x), and uses the generator polynomial G(x) pre-agreed by the sender and receiver to remove it at the sender to obtain a remainder polynomial. The remainder polynomial is added to the data polynomial before being sent to the receiving end. The receiving end uses the same generator polynomial G(x) to divide the received data polynomial f(x) to obtain the calculation remainder polynomial. If the calculated remainder polynomial is the same as the received remainder polynomial, it means there is no error in the transmission; if the calculated remainder polynomial is not equal to the received remainder polynomial, it means there is an error in the transmission, and the sender will resend the data until it is correct. The basic working principle of CRC is shown in Figure 2-11.

The physical layer is concerned with the raw bit stream transmitted over the communication. The design must ensure that when one party sends a binary "1", the other party receives a "1" instead of a "0".

The main task of the Data Link Layer is to enhance the function of the physical layer in transmitting original bits, which refers to providing an error-free line to the network layer.

The Network Layer controls the operation of the connection relationship between subnets. One of the key issues is determining how to route packets from source to destination. Routing can use a fixed static routing table in the network, or it can be determined at the beginning of each session. It can also determine routing for each packet with a high degree of flexibility based on the current network load condition. If too many packets appear in a subnet at the same time, they will block each other's paths and form a bottleneck. This type of congestion control is also within the scope of the network layer. When packets have to cross one network to reach their destination, new problems arise: the second network may be addressed in a completely different way than the first network; the second network may not be able to receive the packet because it is too long; c. The protocols used by each network may also be different. The network layer must address these issues so that heterogeneous networks can interconnect.

The basic function of the Transport Layer is to receive data from the session layer. When necessary, it is divided into smaller units and passed to the network layer, and ensuring that each piece of information reaches the other party is correct, and these tasks must be completed efficiently. In a sense, the transport layer makes the session layer immune to changes in hardware technology.

The Session Layer allows users on different machines to establish session relationships. The session layer allows the transmission of ordinary data similar to the transport layer, and provides enhanced service sessions useful for certain applications. It can also be used to remotely log in to a time-sharing system or transfer files between two machines.

The Presentation Layer performs certain functions. Since these functions are often requested, people want to find a universal solution rather than letting each user implement them. It is worth mentioning that the layers below the presentation layer are only concerned with reliable transmission of bit streams, while the presentation layer is concerned with the syntax and semantics of the transmitted information.

The Application Layer contains a large number of protocols required by users. For example, there are hundreds of incompatible terminal signals in the world, and you want a full-screen editing program to work on many different types of terminals on the network, each with different screen formats, escape codes for inserting and deleting text Sequence, cursor movement, etc. The difficulty can be imagined. One way to solve this problem is to define an abstract network virtual terminal (NetworkVirtualTerminal), to which the editing program and all other programs are oriented. For each terminal type, a piece of software is written to map the network virtual terminal to the actual terminal.

All these requirements to achieve seamless connections between multiple networks have led to the emergence of packet-switched networks based on connectionless interconnections. This layer is called the Internet Layer, and it is a key part of the entire architecture. Its function is to enable the host to send packets to any network and to independently transmit the packets to the target. The order in which these packets arrive may be different from the data sent, so the higher layers must sort the packets if they need to be sent and received in order.

In the TCP/I reference model, the layer above the Internet layer is usually called the transport layer. Its function is to enable peer entities on the source and target hosts to conduct conversations, which is the same as the transport layer of the OSI reference model. Two end-to-end protocols are defined here. (1) The first is the Transmission Control Protocol (TCP), which is a connection-oriented protocol that allows a byte stream sent from a machine to be sent to the Internet without errors. other machines. It divides the input byte stream into message segments and passes them to the Internet layer. At the receiving end, the TCP receiving process reassembles the received messages into an output stream. TCP also handles flow control to prevent a fast sender from sending too many packets to a slow receiver for the receiver to handle. (2) The second protocol is the User Datagram Protocol (UDP), which is an unreliable connectionless protocol and is used for applications that do not require the sorting and flow control capabilities of TCP to complete these functions by themselves. .

The TCP/IP reference model does not have a session layer or a presentation layer. Above the transport layer is the application layer, which contains all high-level protocols. The first ones introduced were Virtual Terminal Protocol (Telnet), File Transfer Protocol (FTP) and Simple Email Protocol (SMTP). The Virtual Terminal Protocol allows users on one machine to log in and work on a remote machine. File transfer protocols provide an efficient way to move data from one machine to another. The email protocol was originally just a file transfer, but later specialized protocols were proposed for it. Many protocols have been added over the years, such as Domain Name Service (DNS) for mapping host names to network addresses; Network News Transfer Protocol (NNTP) for delivering news articles; and Hypertext Transfer Protocol (HTTP) ), used to get home pages on the World Wide Web, etc.

There is nothing below the Internet layer. The TCP/IP reference model does not really describe this part, except that the host must use some kind of protocol to connect to the network in order to pass I packets over it. This protocol is undefined and varies from host to host and network to network.

There are three CSMA insistence backoff algorithms The first type: not adhering to CSMA. If the medium is free, send; if the medium is busy, wait for a while. machine time, repeat the first step. Second type: 1-Adhere to CSMA. If the medium is idle, send; if the medium is busy, continue to listen until the medium is idle and send immediately; if a conflict occurs, wait for a random period of time and repeat the first step. The third type: P-Adhere to CSMA. If the medium is idle, it will send with a probability of P, or it will listen after a delay of a time unit with a probability of 1-P. This time unit is equal to the maximum propagation delay; if the medium is busy, it will continue to listen until the medium is idle, and repeat the first step. step.

Token Ring is an access control method used in ring LANs. Tokens are continuously transmitted on the network ring. Only the site that owns this token has the right to send messages to the ring, while other sites are only allowed to receive messages. . After a node completes sending, it will hand the token to the next site on the network. If the next site does not send a message, it will immediately pass the token to its next site in order, so the token representing the right to send Loop continuously on the ring channel. Each node on the ring can get the opportunity to send messages, and only one node can use the ring to transmit messages at any time, so it is guaranteed that no access conflicts will occur on the ring.

Time Division Malfiplexing (TDM) pre-allocates a specific period of time for each node so that each node can occupy the bus during this period. The way in which multiple nodes occupy the bus in divided time order is called multiplexing. For example, let nodes A, B, C, and D occupy the bus in the order of 1, 2, 3, and 4 respectively. If the time each node will occupy the bus, the required communication time, or the number of bytes of messages to be transmitted can be estimated in advance, the cycle period between each node occupying the bus twice can be accurately calculated. This is useful in controlling network applications to meet certain inter-surface requirements. Time division multiplexing is divided into two types: synchronous time division multiplexing and asynchronous time division multiplexing. The meanings of "synchronization" and "asynchronous" here are different from the previous concepts of synchronization in bit synchronization and frame synchronization. Synchronous time division multiplexing refers to allocating equal time to each node regardless of the amount of communication each device needs to communicate. Whenever the time slice assigned to a node arrives, the node can send data. If the node has no data to send at this time, the transmission medium is empty during the time slice. This means that the average allocation strategy of synchronous time division multiplexing may cause a waste of communication resources and cannot effectively utilize the full capacity of the link.

Concurrent Time Domain Multiple Access (CTDMA) is one of the characteristic technologies used in ControlNe network system communications. Parallel time domain multiple access is a function completed by the physical layer and data link layer in the communication system. Parallel time-domain multiple access relies on the producer/consumer communication model. The generator of message data (data source) acts as the producer in this communication model, and the node that takes this data from the network is called a consumer. Messages sent by parallel time domain multiple access are identified by content. When a node receives data, it only needs to identify the specific identifier (Identifier) ​​associated with this message. The data packet no longer requires a destination address, and the data source only needs The data needs to be sent once. Multiple nodes that need the data simultaneously obtain message data from the same producer from the network by identifying this identifier on the network, so it is called parallel time domain multiple access.