|
Arabic
Bengali
Chinese
English
French
German
Hebrew
Hindi
Italian
Japanese
Korean
Malay
Polish
Portuguese
Spanish
Turkish
Ukrainian
Vietnamese|
ENCYCLOPEDIA OF RADIO ELECTRONICS AND ELECTRICAL ENGINEERING What is Frame Relay? Encyclopedia of radio electronics and electrical engineering
Encyclopedia of radio electronics and electrical engineering / Computers In recent years, a data transmission method called frame relay has become widespread, and often in our literature you can also find its English name - Frame Relay. The main stimulus for the development of this method is the growing demand for high-speed communications for information and computing systems. The emergence of frame relay is due to the development of data transmission terminals (TDTD) with artificial intelligence, reliable digital transmission facilities and high-speed digital communication systems. In order to understand how and why this method appeared and to understand its features in more detail, it is more convenient to start with a brief history of the development of data transmission technology and even telegraphy that preceded it. First data transmission systems The development of data transmission systems is based on the use of more than a century of documentary communication experience accumulated in telegraphy. Telegraph transmission speeds cannot meet modern requirements, but many of the ideas underlying the technology of high-speed data transmission originated in the era of the telegraph. First of all, this refers to the methods of coding transmitted messages. In the course of the development of the technology for transmitting documentary information, the inconvenience of the five-element telegraph code No. 2, once recommended by the International Consultative Committee for Telephone and Telegraph Communications (CCITT), which is part of the International Telecommunication Union (ITU), became apparent. Code No. 2 allows the transmission of alphanumeric text, which is printed on tape and is sufficient for the transmission of simple messages, but it does not meet the modern requirements for the design of these messages in the form of printed text. Therefore, an important stage in the development of the telegraph was the creation of a teletype, i.e. a telegraph apparatus with a typewriter keyboard, for which the seven-element telegraph code No. 3 was established by CCITT Recommendation V.5. Among the 27 = 128 combinations of this code, not only uppercase and lowercase letters are provided alphabets, numbers and other typographic characters, but also code combinations for controlling devices and mechanisms in the transmission process (for example, a carriage return at the end of a line, moving to a new page, and much more). The same set of code combinations has been recommended by the International Organization for Standardization (ISO) as the standard international exchange code for information processing. It is also called an ASCII code (from the first letters of the English words meaning "American Standard Information Interchange Code"). Simultaneously with the problems of direct coding of the transmitted information, the problems of code protection against errors were also solved. There are two classes of error-correcting codes: error-correcting codes and error-detecting codes. The former are characterized by a large redundancy of transmitted messages. It allows, in the event of individual errors, to still correctly interpret the transmitted message. Such codes are used only in very sensitive channels, for example, in deep space communications, where the importance of correct reception justifies the reduction in the useful transmission rate. Another class is codes that detect errors. Such codes make it possible to detect only the fact of an error in a certain group of characters without a specific indication of an erroneous character. Therefore, upon such detection, usually the entire group of characters with a registered error is reset, and an automatic request for retransmission is sent to the transmitting side. It is this method that has found wide application in commercial data transmission systems, where it is important to maintain high channel performance.
The simplest error detection methods began to be used back in the era of reperforating telegram reception, when transit telegrams were recorded on punched tape, this tape was torn off and transferred by the operator to the transmitter of the desired outgoing direction for further transmission. The punched tape was a paper tape, the width of which provided for eight positions in each row for punching holes that carry information about the binary digits of the code combinations. Seven of these positions were reserved for registering the bits of the seven-element code, and the eighth was for error detection by checking for parity. This meant that the value of the eighth binary digit was chosen in such a way that the sum of the elements is unlikely to be even. If the receiver found an odd sum in any row, this meant that an error had occurred. It is easy to see that this method of error control allows you to detect one error, but leaves unnoticed two errors in a row. Both in the case of the same sign of two errors, and with their different signs, the simultaneous occurrence of two errors cannot change the result of the parity check, and therefore such errors remain undetected. To further enhance the error detection capabilities, a longitudinal check can be additionally applied. If the parity check described above, which is called cross-check, is added to check the sum of identical bits in a fixed series of characters following one after another on the tape, the possibility of error detection will increase. For such a check, at the end of each series, additional bits of a longitudinal check have to be inserted, which look like another character, although they are not. The advent of electronic means of transmitting and switching messages made it possible to abandon punched tape and use more advanced codes for error detection. This made it possible not to use the eighth digit for parity checking and to include it in the code combination. As a result, the ASCII code turned out to be extended to 2*=256 code combinations. Of these, the first 128 characters (coded by numbers from 10 to 127) are common, and the second 128 characters (coded by numbers 128-255) are additional and are used, in particular, to encode the national alphabets of different countries. The use of the ASCII code allows you to work with texts containing both the Latin and any national alphabet, which creates great convenience for users. However, the circumstances with the coding of the letters of the Russian alphabet were not the most favorable. The root of the discrepancies goes to the unsuccessful design of the ST-35 telegraph apparatus, which in the first period of the development of computer technology in our country served as an input / output device of a computer. By definition, a teletype is a telegraph machine with a typewriter keyboard. The standard arrangement of letters on the keys of typewriters in different countries is determined by the statistics of the corresponding language. In other words, the more often a letter occurs, the closer its key is located to the middle of the keyboard, where the index fingers work. For example, the arrangement of letters in the first row of alphabetic keys of a Russian typewriter begins with the letters YTSUKEN, while on an English-language Latin typewriter this row begins with the letters QWERTY. On the CT-35 keyboard, the standard position of Latin letters is violated, they are located on the basis of phonetic proximity to the corresponding Russian letter (i.e., instead of QWERTY, the letters YCUKEN are in the first row). The assignment of code combinations to each character on a key (or, as they say, character encoding) cannot be arbitrary, since computer word processing requires that the binary numbers assigned to each letter increase in accordance with the alphabetical order of these letters. That's where the confusion came from. For the apparatus ST-35. working with a computer, the code KOI-8 was developed. Subsequently, when keyboards appeared with a standard arrangement of Latin letters, an alternative GOST code was adopted. Later, this code was modified and then adopted as the main one. Thus, in the USSR there were four standards for information processing codes. In the conditions of such a leapfrog, our country was not able to act in the international arena as a legislator for coding the letters of the Russian alphabet, as a result of which the Bulgarian MIC code, the "American" Russian code (RS-866), as well as American Cyrillic (RS-855). This means that there are at least seven different code combinations for Russian letters in the world, which creates great inconvenience for Russian-speaking users, making it difficult to exchange documents in Russian and hindering the introduction of materials in Russian on the Internet. Apparently, it's time to think about creating a program that automatically recognizes the encoding used for Russian letters and translates them into the code necessary for decryption. In the future, it is expected that the coding of typographic characters will move from a one-byte code to a two-byte code (Unicode), in which each letter of the alphabets of different languages, mathematical signs, decorative and other characters is assigned its own sixteen-bit combination. However, this will not solve the problem of encoding Russian letters, since translators between different one-byte and a single two-byte code will still be required. The described story with the coding of the letters of the Russian alphabet is not only of particular importance as an example of the detrimental consequences of a specific short-sighted decision. More important is the general methodological significance of this example, which shows the need for a deeper approach to the problems of standardization, taking into account the fact that the transmission of information is not limited to sending signals, but must be accompanied by the necessary processing and interpretation of the received information. Therefore, further we dwell on a brief description of approaches to standardization. ISO Open Systems Interworking Reference Model and X.25 Protocol The variety of functions that are performed by modern means of transmitting and processing information, the various possibilities for the technical implementation of such tools, as well as the trends in the continuous improvement of these functions and tools lead to the need to use the principle of multi-level (multi-layer) architectures in standardization. The essence of this principle is to separate the most important functions into independent levels (layers) of processing and describe the interactions between the levels, regardless of their implementation. With this approach, individual levels in a complex system can be replaced by new ones, if the accepted standard rules for their interaction with neighboring levels are not violated. A well-known example of such a layered architecture is the ISO Open Systems Interconnection (OSI) Reference Model shown in Figure 1. XNUMX. This shows the connection diagram of two end users A and B. which are included in the communication nodes that are end users for these users. The model contains seven levels, for which the following abbreviations are accepted: F - physical level, K - channel level. C - network level, T - information transportation level (or transport level), SU - session level, UE - presentation level, P - application level. Each of the listed levels of the transmitting side interacts only with the same level of the receiving side using procedures called communication protocols. However, communication between two peer layers does not occur directly, but only through the physical layer. To do this, each higher layer refers to its immediate lower layer as a service provider. For example, the topmost application layer II, which interacts with a real user, should, on the one hand, perceive the real world, and on the other hand, give this world the opportunity to access technical means of transmitting and processing information through the presentation layer. In other words, at the application level, the semantics (i.e., meaning, or sense) of the transmitted information is described. This information is provided with the necessary header and is transmitted in the form of an application layer block for further processing to the presentation layer of the UE. At this level, the "syntax" of the transmitted information is described and automatic negotiations are conducted with the interacting party about the rules for interpreting the data, taking into account, if necessary, the system of their compression or encryption. The data block of the presentation layer provided with a new header is transferred to the session layer of the CS. The latter serves to control the dialogue procedures, including the establishment of a connection, the mechanism for detecting and establishing the direction of transmission, tracking transmission control points in time. The session layer data block, provided with another header, is transmitted to the transport layer T1, which sets the network-independent standards for transmitting messages from user to user, including general requirements for error control, automatic recovery of communication interruptions, automatic control over the correctness of the sequence of received data, etc. information is reflected in the next header, and in this form the transport layer data block is sent for transmission to the network. The protocols of these four layers are called high-level protocols, and the functions they perform are related to the functions of the end user. They are usually performed by the host computer. The technical means of the communication network include three lower levels that provide network services. The data block of the transport layer arriving at the network level C is supplied with a new header, which contains information about the addresses of the sender and recipient, the serial numbering of the block, and some other service information. The network layer data block formed in this way is called a packet. In order to transmit a packet over the network, the network layer uses the services of the K link layer, which ensures that the packet is delivered only to the nearest node. To do this, the packet is supplied with another header - the channel-level header, which carries its own serial numbering of blocks transmitted over this section, the address of the destination node, and other service information. A block of data formed at the link level is called a frame. To transmit a frame to a neighboring node, the channel layer refers to the physical layer F service. This layer establishes standards for the mechanical connectors and electrical characteristics of the communication channel, as well as digital signals transmitted over it, including line seizure and release signals. To maintain the characteristics of the transmitted signals at the physical layer, regenerators can be installed. The frame received by the neighboring node is released from the link-layer header, i.e., it is turned into a packet. The received packet is transmitted to the network layer, where its header is analyzed and the direction of further transmission is determined. Further, a new frame is formed from this packet, which is transmitted over the next section. The described method of packet transmission is called the X.25 protocol. It is included in CCITT Recommendation X25. first approved in 1976 (revised versions published in 1980 and 1984). The X.25 Recommendations provide an interface specification covering the three lower layers of the considered OSI ISO Reference Model. From the above information, you can see that the idea of the X.25 protocol resembles the traditional reperforated transmission of telegrams. The difference is that it is not a sequence of parity-checked characters that is transmitted over the section, but a standard frame with better error control (this is discussed below). In the node, however, it is not an operator who transfers the paper tape to the device of the desired transmission direction, but an electronic switching device that records the packet, analyzes its header and then reads it for transmission in the required direction. However, this is where the similarities between the X.25 protocol and traditional telegraph technology end, and further consideration reveals fundamental differences. The main one is that a large number of simultaneously operating channels can be organized through the interface connecting the terminal data transmission device (TDTD) and the linear data transmission device (LUPD). All of these channels pass through the same PDSN output terminal and over the same wire line, but carry different messages that may be directed to different recipients (other PDSNs connected to the network through their LUPDs). Such channels are called logical or virtual. When organizing a multi-channel transmission system over a single line using frequency or time division equipment, each channel is loaded by its own transmission system or can be idle regardless of the load of other channels. The virtual channels, formed on the basis of statistical multiplexing, provide the possibility of more flexible use of the line bandwidth, maintaining transmission continuity in the presence of load. Channel layer technology development The process of transmitting frames over a duplex digital channel, provided for by Recommendations X.25, is called the balanced procedure for accessing the LAPB channel (in English, LAPB - Link Access Procedures, Balanced). The standard X.25 frame format for such a transmission is shown in Fig. 2, which shows that the "header" added to the packet contains 48 bits, which are actually placed both in the head and in the tail of the frame (24 bits each). In the head part there are, in particular, octets carrying the address, as well as control and management signals. Among the bits placed in the tail is a 16-bit frame check sequence (FRS), which makes it possible to detect even entire bursts of errors. Error detection is based on the theory of cyclic codes. It is reduced to algebraic transformations of the transmitted sequence using a specially selected generating polynomial of a certain form and comparison of the result of these transformations at the receiving end with the CPC obtained as a result of a similar transformation at the transmitting end. The SPDK procedure is an integral part of the high-level protocol used to control the channel (High-level channel control - VUK, or High level Data Link kontrol - HDLC). The latter provides for rather complex procedures for managing transmission over the channel, including establishing a connection, maintaining message transmission in both directions with control of frame sequence numbers and using the "window" mechanism (limiting the number of transmitted frames for which the receiving party has not yet received confirmation), rotation of the "window " as acknowledgments are received, error control and their correction by retransmissions, as well as termination of communication. This is a rather complicated protocol, the description of which takes up quite a lot of space. For example, the frame format shown in Fig. 2 may take the form of more than just an information frame carrying a packet. Along with this, the control and management octet code provides for the possibility of creating four different control frames that may not carry packets, or 32 unnumbered frames that do not carry packets, but only control processes such as connection or disconnection.
It should also be noted that a communication channel here means only a separate section between two network nodes (in English, link, i.e. literally "link"), and not the entire transmission path from the sender to the recipient (or , as they say, from end to end). In other words, the described procedure is repeated at each site, and the control over the transfer from end to end, as already mentioned above, is not a function of the channel, but a function of the network. An important task is the choice of frame length. As is clear from the above, it is determined by the length of the packet plus 48 bits. Thus, in fact, we are talking about choosing the packet length. With a small packet length, the overhead of 48 bits can be significant, which will negatively affect the performance of the channel. If the packet length is too long, the frame is more likely to be discarded due to error detection, and this will require retransmission, which also leads to a decrease in link performance. Thus, there is an optimal packet length, which depends on the probability of error in the channel. Taking into account the fact that different channels may meet, the standard does not determine the packet length, but leaves it at the discretion of the user. Since in this case the frame does not have a fixed length, it is necessary to designate its beginning and end with a special sequence like 01111110, called a flag (see Fig. 2). The introduction of flags imposes a serious limitation on the transparency of the channel. If there are six XNUMXs in a row in the transmitted message, they will be taken as a flag, and this will disrupt the entire transmission. To restore the transparency of the channel, at its transmitting end, after any five ones, except for the flag, a zero is inserted, while at the receiving end, the zero following after any five ones is always removed. This event allows you to restore the transparency of the transmission, and if seven units in a row are found in it, the corresponding frame will be dropped. Naturally, error checking in the frame is carried out on the sequence from the first bit of the address field to the last bit of the information field (packet) before introducing zeros into it after every five ones on transmission and after removing these zeros at reception. An important problem, often solved in the design of a communication system, is the problem of distribution of functions between the subscriber unit and the network. For example, when designing a telephone network, it is decided whether to provide the subscriber with the opportunity to install answering machines in his own telephone set or offer him the service of a centralized answering machine in a communication center (voice mail). Similar problems arise when organizing data transmission services, where the question of whether it is necessary to record packets at intermediate nodes becomes relevant. The solution of this issue depends on many factors that characterize the quality of the network and the level of development of the PDSN technology. If the network channels are not of very high quality, it is advisable to check for errors and correct them at each site, and then the recording of packets at the intermediate node is justified. At the same time, this may require a rather large volume of the recording device (memory) both for recording the packets themselves and for all the programs necessary to implement the protocols of the 2nd and 3rd layers (i.e., the channel layer and the network layer). With the growth of transmission speeds, the amount of such memory will grow. On the other hand, with increased reliability of network transmission and with more advanced PDSNs (eg personal computers), many functions of the network (ie intermediate nodes) can be transferred to the PDSN. Then, naturally, the idea of relaying frames in intermediate nodes without recording them arises. This idea is sometimes called fast packet switching, since packets are not separated from frames, and all procedures for processing them are concentrated at the link level. The proposal for frame relay as an alternative to the X.25 protocol was first proposed to the CCITT in 1984, but the development of standards and the development of equipment were not completed until 1990. An important limitation of the frame relay technique is that its application does not eliminate inherent to the X.25 protocol variable delays. Therefore, Frame Relay is not intended for telephony or video transmission, but it is ideally suited to the requirements of high-speed data transmission. The frame structure for relaying without accessing the network layer is shown in Fig. 3.
Compared to Fig. 2, here, instead of an eight-bit neighbor address, a ten-bit virtual channel pointer UVC (DLCI - Data Link Connection Identifier) is provided, on which frames are relayed to a specific destination. In the X.25 protocol, the virtual channel number is transmitted in the packet header (and contains 12 bits). Here it is moved to the frame header, since the network layer is completely dismantled during frame relay. The channel layer is also subjected to significant dismantling, with the exclusion of many functions, as a result of which the channel performance increases dramatically. The frame relay procedure at an intermediate node includes three steps: 1) checking the frame for errors using PPK and discarding the frame when an error is detected (but without a retransmission request!); 2) checking the ICC against the table and, if this pointer is not defined for the given channel, discarding the frame; 3) if the outcome of the first two operations is positive, the frame is relayed to the destination using the port or channel specified in the table. Frames can be dropped not only due to the detection of an error, but also when the channel is overloaded. However, this does not break the connection, as the missing frames will be detected by the upper layer protocol of the receiver (see above about the transport layer), which will send an appropriate request for the transmission of the missing frames. In addition to the UHC bits, octet number 1 contains the C/O (command/response) and PA (address extension) bits. The K / O category is provided for management purposes, but is not yet used. As for the RA bit, it is important, as it indicates an increase in the size of the frame header (in excess of 48 bits). A similar need exists in the X.25 protocol, since only three bits are allocated for frame numbering in the control and control octet of the frame header. Therefore, the "window" mechanism can allow no more than seven unacknowledged frames to be transmitted. However, when operating on a satellite link, there may be more than seven frames in transit, and therefore the "window" is expanded to 127. In this case, seven bits are needed for numbering, which requires expanding the frame header format. In the case of frame relay, a ten-bit virtual circuit number, sufficient for local communication, may not be sufficient for global communication, and this may require its expansion. In the second octet, three bits are used to control channel congestion. The Forward Explicit Congestion Notification (FECN) bit is set by the network to indicate that congestion is possible on the path from the sender to the recipient. The Backward Expkicit Congestion Notification (BECN) bit is set by the network in the reverse direction frames and notifies forward path congestion. The Discard Eligioility bit (DE) indicates a lower priority of the transmitted frame, which can be considered as a candidate for discarding during congestion. In X.25 transmission, the typical default packet size is typically 128 bytes, while in local area networks (LANs), transmitted packets can be 1500 bytes or more in length. Therefore, when communicating with a LAN through an X.25 network, the transport layer packets are split into smaller blocks of information, formed as X.25 packets, and their combination is carried out after transmission. This example clearly shows where and why the ideology of transition from the X.25 protocol to frame relay is being formed. Author: V. Neiman, Moscow
Artificial leather for touch emulation
15.04.2024 Petgugu Global cat litter
15.04.2024 The attractiveness of caring men
14.04.2024
▪ Ultra-fast Samsung PM1725 and PM1633 SSDs ▪ Low noise precision amplifier ▪ Nanovaccine will protect the brain from nicotine ▪ Additional features of the touch sensor B6TS
▪ section of the site Audio Art. Article selection ▪ article Gogol. Popular expression ▪ article Which plant holds the record for growth rate? Detailed answer ▪ article Movement in the jungle. Travel Tips ▪ article Lacquer made of hard rubber (ebonite). Simple recipes and tips ▪ article Combined HPF filter. Encyclopedia of radio electronics and electrical engineering
Home page | Library | Articles | Website map | Site Reviews
www.diagram.com.ua |
See other articles Section 
Latest news of science and technology, new electronics:
All languages of this page