RU2543565C1 - Method of forming data transmission channel - Google Patents

Abstract

FIELD: radio engineering, communication.

SUBSTANCE: method of forming a data transmission channel includes searching for all active receiving-transmitting devices of the transmitting side, for each receiving-transmitting device, specifying a set of individual settings needed for optimum operation, including network connection parameters and allowable communication quality parameters, and creating a terminal device in which are recorded said individual settings and the assigned IP address of the corresponding receiving-transmitting device; the terminal devices are connected to each other by a terminal device switch; a data stream intended for transmission is converted through a converter to an internal format and divided into packets which are stored in an input buffer; data from the input buffer are allocated in a certain manner to the active receiving-transmitting devices of the transmitting side and sent to the receiving side, wherein the terminal device switch is used to reallocate data packets intended for transmission based on change in communication quality parameters of each communication channel.

EFFECT: efficient use of bandwidth of each channel when transmitting data through dynamic allocation of network level data between all transmitting channels according to current characteristics thereof.

12 cl, 8 dwg

Description

Technical field

The invention relates to the field of telecommunications, and more specifically to methods for transmitting digital information, and can be used, for example, in systems providing a broadband and stable data channel.

State of the art

The invention can be used in data transmission systems used in vehicles and in places where access to fixed lines is absent or difficult. Such systems are required, for example, to support the activities of field teams and specialists in operational services, the media, and many industries.

The well-known digital communication systems providing the possibility of two-way data transmission are characterized by the following advantages and disadvantages:

Wired communication systems (Ethernet and its analogues):

Advantages: wide bandwidth, high stability of the characteristics of communication channels, high reliability

Disadvantages: the inability to use in open areas and in motion

Short-range wireless communication systems (WiFi-connection and analogues):

Advantages: sufficient bandwidth for most applications, stability of the characteristics of communication channels, high reliability, the ability to use in open areas

Disadvantages: small radius of the radio coverage zone, inability to use in motion

Wireless cellular communication systems (modems for GSM / 3G / 4G networks):

Advantages: the ability to use in open areas and in motion

Disadvantages: insufficient bandwidth, which provides a single transceiver; instability of the characteristics of communication channels, insufficient radio coverage of each of the telecom operators for the stable operation of transceivers in remote regions

Long-range wireless communication systems (satellite communication systems):

Advantages: wide bandwidth, the ability to use in open areas

Disadvantages: high transmission delays, the dependence of the characteristics of communication channels on the climate, the inability to use in areas with dense buildings and in traffic.

Also known systems that simultaneously use several of the above communication systems for data transfer. This allows you to combine the advantages and eliminate some of the disadvantages of each communication system separately.

For example, from the description of patent EP 2466811 patent holder Alcatel Lucent, publ. 06/20/2012, there is a known method of forming a data transmission channel, which consists in the fact that at the first stage they search for all available communication channels; at the second stage, the data stream intended for transmission is divided into packets of equal magnitude; at the third stage, these packets are distributed over the communication channels, taking into account the congestion and bandwidth of the communication channels, and are queued; in a fourth step, receiving and assembling data packets are performed on the receiving side.

The disadvantage of this known technical solution is that it has low efficiency and transmission speed of large amounts of data. This is due to the fact that the bandwidth and quality indicators of the communication channels are monitored only before the start of packet transmission and do not take into account the change in the bandwidth of the communication channels during the transmission of packets.

In addition, the signal level of the radio network is used as an input characteristic of the channel bandwidth. The estimated bandwidth is calculated by the signal level, but this method is not applicable for such widespread communication channels as GSM, WCDMA and LTE, where there is no direct connection between the quality indicators of the communication channels (channel speed, channel delay and percentage loss) and signal level. In networks of this type, the characteristics depend solely on the state of the network, the current load on the network, etc. Thus, another significant drawback of this method is its versatility, since it is not applicable to all types of communication channels.

From the description of the patent for US invention No. 8259739 patent holder CISCO TECHNOLOGY, INC, publ. 09/04/2012, there is a known method of forming a data transmission channel in which several different communication channels are simultaneously used, between which data packets are distributed and sent to the receiving side. A feature of this known method is that the data packets are distributed between communication channels with different bandwidths depending on the sending queue in front of the receiving and transmitting devices (hereinafter PPU).

This method eliminates the downtime of data transmission channels and allows you to effectively use the entire bandwidth of each communication channel.

However, this method, as well as the analogue described above, is not universal, since it is far from effective for all types of communication channels. Significant disadvantages of this method will be manifested when using communication channels built on equipment that implements shaping and data buffering, for example, GSM, WCDMA and LTE networks. In such communication channels, all packets are instantly transferred to network equipment, without creating a queue on the transmitting interface, but as a result of schedule shaping, packets will be delayed or lost on the network. Thus, the method described in the patent cannot provide maximum efficiency, minimum delays in the transmission of data packets and versatility.

From U.S. Patent Application No. 2008,0219,281 to Mushroom Networks, publ. 09/11/2008, there is a known method of forming a data transmission channel, which consists in the separation of sessions between transceiver devices. Packages of one session can be distributed between one or more PUFs. The distribution algorithm is a round robin (uniform distribution of the choice of the communication channel) or distribution associated with the configuration of the channel capacity. In this known method, the schedule for the interaction of IP equipment is divided between several communication channels. The method is based on the principles of traffic separation for sessions at the 4th network level and the transmission of the session through one or more communication channels.

The disadvantage of the described method is its low efficiency, which is due to the inability to analyze the parameters of the communication quality of the transmitting channel and their changes in time, in conditions of high volatility of the characteristics of the communication channels. The method rigidly binds a session of the 4th network level to one or several communication channels, which can lead to a complete halt in the interaction with complete degradation of the selected channels, i.e. the appearance of large delays, on the one hand, and the inefficient use of unused channels, on the other.

From the description of the application for US invention No.20080267184, the applicant Mushroom Networks, publ. 10/30/2008, there is a known method of forming a data transmission channel, which extends the method described in the application US 20080219281 of the same applicant. This method consists in the fact that at the first stage (preparatory) they search for all active transmitting and receiving devices (PPU) of the transmitting side, determine the type of each PPU and its communication operator, each identified active PPU is assigned an IP address with the formation of a communication channel; at the second stage, the data stream intended for transmission by means of a converter is converted to the internal format and divided into packets, and, if necessary, into packet segments, the finished packets are placed in the input buffer; at the third stage, the data from the input buffer is distributed over the active control units of the transmitting side and sent to the receiving side; in the fourth step, data packets are received on the receiving side, and then data is converted from the internal to the original format using a converter.

This analogue is the closest to the claimed method of forming a data transmission channel.

It is worth noting that in the closest analogue, a buffer is used on the receiving side, in which the received data packets for each communication channel are stored. On the transmitting side with the receiving side for each received data packet, confirmation of its receipt is sent, the same communication channel is used to transmit the confirmation as for the delivery of data packets. If there is no acknowledgment on the transmitting side, retransmission of the unconfirmed data packet is performed.

Such an implementation of guaranteed delivery effectively solves the problem of eliminating packet loss during transmission, but has a number of disadvantages that will appear when the characteristics of the transmission channels in the forward and reverse directions change. So, with the degradation of the reverse direction of transmission, there may be a significant delay in the transmission of confirmations, which will lead to the retransmission of data packets for which no confirmation has been received, despite the fact that the data packets have been successfully delivered.

A significant drawback of the closest analogue is the effect of the degradation of one of the communication channels on the efficiency of using the others, since when changing the quality parameters of one or several communication channels, the method does not allow to quickly redistribute data packets between communication channels, as a result, the communication channel, the throughput of which has deteriorated, is is overloaded with data, and a communication channel, the throughput of which, on the contrary, has increased, transfers less data than it could transmit. Due to the reasons described above, the closest analogue (US 20080267184) does not allow to combine and use with maximum efficiency all available communication channels for data transmission.

Disclosure of invention

The objective of the present invention is to provide a method of forming a high-speed data channel, allowing with the highest possible efficiency to combine and use all available communication channels. Moreover, the present invention should ensure the universality of the use of virtually any known communication channel, minimizing delays, maximizing the use of bandwidth and eliminating the influence of degradation of one or more communication channels on the efficiency of use of the rest.

The task in the method of forming a data transmission channel, which consists in the fact that at the first stage they search for all active transmitting and receiving devices (PPU) of the transmitting side, determine the type of each PPU and its communication operator, each identified active PPU is assigned an IP address with the formation of the channel communication; at the second stage, the data stream intended for transmission by means of a converter is converted to the internal format and divided into packets, and, if necessary, into packet segments, the finished packets are placed in the input buffer; at the third stage, the data from the input buffer is distributed over the active PUs of the transmitting side and sent to the receiving side, at the fourth stage, receiving data packets is received at the receiving side, and then the data is converted from the internal to the original format using a converter; is achieved by the fact that at the first stage for each control panel, a set of individual settings is set up necessary for its optimal operation, including network connection parameters and acceptable parameters for communication quality, and a terminal device (OS) is created in which the mentioned individual settings and the assigned IP address are written corresponding PPU, terminal devices are interconnected by a switch of terminal devices; at the third stage, the data is transferred from the input buffer to the transmitting stream analyzer, which analyzes the transmitted data and logically separates the data into interaction flows, determined by the set of parameters of the sender, receiver, and type of interaction protocol, while the stream analyzer sets the processing mode for each of the transmitted packets and requirements for the quality of transmission, then separately for each of the interaction flows, by means of a numbering device, sequentially assign numbers to data packets, and so the same numbers, if any, to the packet segments, after which the data packets are sent to the DT switch, by means of which each of the packets is assigned the identifier of the forwarding terminal device (DT) and, thus, the packets are distributed according to the corresponding control points; in addition, by means of the generator of control packets of the transmitting side, a control packet is generated and transmitted to the Switch of terminal devices, with the help of which the control packet is sent via all available communication channels to the receiving side; in the fourth step, on the receiving side, by means of a decapsulator, the received stream is divided into control packets and data packets, the control packets being sent to the receiving packet control processor, and the data packets being sent to the receiving flow analyzer; by means of the control packet processor of the receiving side, for each communication channel, measurements and calculation of communication quality parameters are performed; by means of the control packet generator of the receiving side, regardless of the receipt of the control packet from the transmitting side, a counter control packet is formed in which information about the communication quality parameters of each communication channel is placed and the counter control packet is sent to the transmitting side; at the fifth stage, after receiving at least one oncoming control packet on the transmitting side by means of the transmitting control packet control processor, information about the communication quality parameters of each communication channel is extracted and the information is transmitted via the transmitting side control packet generator to the op-amp switch, taking into account the latter the received information updates the information on the parameters of the communication quality of each of the communication channels and writes these parameters to the corresponding OS; then, using the switch of the terminal devices, the data packets intended for sending are redistributed taking into account changes in the communication quality parameters of each communication channel; at the same time, communication channels that do not meet the conditions of the communication quality parameters specified in the first stage are excluded from the distribution; after a specified period of time after sending the previous control packet, the next control packet is formed on the control packet generator of the transmitting side; on the flow analyzer of the receiving side, the analysis of the received data is carried out and the data are logically divided into interaction flows, after which the interaction flows are sent to the sorter, which organizes the received packets and sends the ordered packets to the collector, where they are glued segmented packets, if any, after which the packets data is transmitted to a converter, by means of which the original data stream is restored; thus form a single high-speed data channel.

An embodiment of the present invention is possible, according to which, in the fifth step, when ordering data packets, the sorter sends concurrently through all communication channels the receipt of data packets to the numbering of the transmitting side, from where the lost data packets are re-sent to the receiving side, in addition, in the case of receiving a duplicated packet on the receiving side, it is deleted by the sorter in the process of organizing the received data packets.

An embodiment of the present invention is possible, according to which, in the fifth step, acceleration is performed for each communication channel in order to increase its throughput by transmitting more data packets per unit time than can be transmitted at the calculated speed obtained from the control packet for this communication channel.

An embodiment of the present invention is possible, according to which, in the fifth step, bandwidth is backed up for each communication channel, maintaining the gained speed, by transmitting as many data packets per unit of time that can be transmitted at the calculated speed obtained from the control packet for this communication channel, and to ensure the availability of a sufficient number of data packets duplicate data packets intended for transmission.

The inventive method allows the use of the following wired and wireless control panels: LTE modem, GSM modem, CDMA modem, WIFI module, satellite modem, optical modem, ADSL modem, Ethernet modem. The method does not exclude the use of other types of foam.

The inventive method dynamically distributes the data of the network level of interaction between all communication channels in accordance with their current characteristics, which guarantees the efficient use of the bandwidth of each communication channel and allows to achieve minimal delays in data transmission. In addition, the method is universal and allows you to use almost all possible types of communication channels.

The list of explanatory figures

The technical essence of the proposed technical solution is illustrated by figures, where:

Figure 1. The process of processing data packets on the transmitting side before sending them.

Figure 2. The process of processing data packets after receiving them on the receiving side.

Figure 3. Diagram of the process of packet distribution.

Figure 4. A diagram of the process of assembling packages by the sorter into the initial interaction flow.

Figure 5. The application of the proposed method on the example of the agricultural industry Datacrosser.

6. The scheme for converting data streams in the process of their transmission through the Datacrosser APK

7. The process of processing data packets before sending them to the transmitting side with delivery confirmation.

Fig. 8. The process of processing data packets after receiving them at the receiving side with delivery confirmation.

The list of positions:

10 - input data stream;

12 - stream of prepared data packets;

14 - outgoing stream (that is sent from the transmitting side to the receiving side);

16 - control packets prepared for transmission from the transmitting side to the receiving side;

17 - control packets 27 received on the transmitting side;

18 - service acknowledgment packets received on the transmitting side correspond to service acknowledgment packets 28 sent from the receiving side;

19 - data on the quality parameters of the communication of each of the communication channels received from the processor of the control packets of the transmitting side 170 to the switch terminal devices 135;

110 - converter;

115 - input buffer;

120 — transmitting side flow analyzer;

125 - numbering;

130 - encapsulator;

135 - terminal switch;

140 _i - terminal device (where i corresponds to the control panel number from 1 to N, for example: 140 ₁ , 140 ₂ , ..., 140 _N );

150 _i - PPU driver (where i corresponds to the PPU number from 1 to N, for example: 150 ₁ , 150 ₂ , ..., 150 _N );

160 _i - transceiver (PPU) (where i corresponds to the number of the PPU from 1 to N, for example: 160 ₁ , 160 ₂ , ..., 160 _N );

170 - processor control packets of the transmitting side;

180 - a generator of control packets on the transmitting side;

20 is a received stream of encapsulated data and control packets;

23 - received stream of decapsulated data in the internal format;

23 _i is the received stream of decapsulated data in the internal format received from the corresponding transceiver (PPU), where i corresponds to the number of the PPU from 1 to N, for example: 23 ₁ , 23 ₂ , ..., 23 _N ;

24 - flow of interaction;

25 - restored data stream;

26 - control packets received on the receiving side;

27 - control packets sent from the receiving side to the transmitting side;

28 - service confirmation packets sent from the receiving side to the transmitting side;

210 - decapsulator;

220 - host flow analyzer;

230 - sorter;

240 - collector;

250 - converter;

270 — receiver packet control processor;

280 — host control packet generator;

520 - mobile gateway (client part of the agro-industrial complex);

530 - data transfer operator (cellular provider, satellite provider or Wi-Fi connection provider);

550 - server (server part of the agro-industrial complex);

61 - analog video signal;

62 - MPEG2TS digital video stream;

63 — video stream prepared for transmission encapsulated in UDP packets;

64 — received video stream encapsulated in UDP packets;

65 is a video stream of MPEG2TS;

66 - video stream provided by a video server via RTSP or RTMP;

67 - video stream prepared for transmission encapsulated in UDP packets transmitted in acknowledged mode;

68 — received video stream encapsulated in UDP packets transmitted in acknowledged mode;

69 - flows of interaction with mobile client devices via TCP / IP;

600 - an analog video camera;

605 — a mobile client device;

610 - video encoder;

620 - video server;

630 - DHCP server;

690 - local area network (IP equipment and services in the network of a mobile gateway);

70 - interaction with mobile client devices via DHCP;

71 - output streams of interaction.

DETAILED DESCRIPTION OF THE INVENTION

The method of forming a data transmission channel, which consists in the fact that at the first stage they search for all active transceivers (PPU) of the transmitting side.

The search for available PUFs 160 _i (where i is from 1 to N, and N is the total number of PUFs) of the transmitting side is performed by monitoring the status of all internal and external interface connectors (ports). Found PPUs are checked by means of the operating system and drivers PPU 150 _i for readiness for work, thus identifying active PSUs.

The type of each PPU 160 _{i is} determined (where i is from 1 to N, and N is the total number of PPU) and its communication operator.

As the PUF 160 _i , both wired and wireless types of data transmission devices can be used, for example, LTE modem, GSM modem, CDMA modem, WIFI module, satellite modem, Ethernet modem, ADSL modem, optical modem. A telecom operator is understood as the corresponding data transmission service provider for each type of ISP.

On the transmitting side, several data transmission devices of the same type can be used, any combinations are possible, however it is preferable to have PPU 160 _{i of} various types to increase the universality of work in various conditions.

Each identified active PPU 160 _{i is} assigned an IP address with the formation of a communication channel.

Preparatory work with the PUF 160 _{i is} considered completed after receiving the PUF 160 _i IP address in the network that provides communication services for the corresponding PUF 160 _i . The networks that provide communication services for PPU are understood as: for WiFi and Ethernet - local computer networks available for connection, for 3G / 4G, CDMA and other cellular modems - networks of mobile operators, for satellite modems - satellite networks, etc.

For each PPU 160 _{i, a} set of individual settings is set that is necessary for its optimal operation, including network connection parameters and acceptable parameters for communication quality, where network connection parameters, for example for modems of cellular operators, are: network connection mode - one of: GPRS, EDGE, WCDMA, LTE, and the name of the network access point.

Communication quality is determined by four parameters:

- Bandwidth (Bandwidth), describes the bandwidth of the transmission medium, determines the width of the channel. Measured in bit / s (bps), kbit / s (Kbps), Mbit / s (Mbps), Gbit / s (Gbps).

- Delay in packet transmission (Delay), measured in milliseconds.

- Oscillations (jitter) of delay in transmission of packets - jitter.

- Packet loss. Determines the number of packets lost on the network during transmission.

Allowable parameters for each of the PPU 160 _i are set by the user of the system depending on the type of application being solved and consist in setting the following parameters on the transmitting and receiving sides:

- maximum packet waiting time on the receiving side;

- maximum waiting time for a delivery confirmation packet;

- the number of packet retransmissions;

- size of the input buffer;

- the ability to use packet duplication;

- permissible values of the minimum transmission rate;

- permissible values of the maximum delays and jigger during transmission for various types of used PPU and telecom operators.

Priority of time characteristics (low allowable waiting time, delivery shutdown with confirmation, tightening the requirements for delays and jitter) is recommended for solving problems for which the implementation of data packet transmission with minimal delays is most important, but taking into account the possibility of losing some of the packets during transmission. An example of using this mode is the transmission of streaming video and audio data.

The priority of quality characteristics (using delivery with acknowledgment, increasing the size of the buffer and the acceptable waiting time, mitigating the requirements for delays and jitter) is recommended for solving problems for which guaranteed transmission of all data packets is most important, taking into account a possible increase in transmission delays. An example of using this mode is data transfer via HTTP, HTTPS, FTP, etc.

The optimal mode of operation refers to the mode of operation of the PUF, in which the PUF provides the maximum transmission speed and minimum delay in transmission under current conditions.

For each PUF 160 _i , a terminal device (OS) 140 _{i is created} , into which the mentioned individual settings and the assigned IP address of the corresponding PUF 160 _{i are} written, the terminal devices are interconnected by a terminal device switch 135.

The terminal device 140 _i is a logical entity for organizing the transmission and reception of data packets through each PPU 160 _i and storing data on the PPU 160 _i (settings, operating mode, statistics on transmitted and received data, etc.)

At the second stage, the data stream 10 intended for transmission by the converter 110 is converted to the internal format and divided into packets, and, if necessary, into packet segments, the finished packets are placed in the input buffer 115.

By “data stream” 10 is meant a set of streams of IP packets intended for sending from various consumers and data recipients using data transfer protocols no lower than the network level (see OSI network model).

The data packet reduced to the internal format contains both the original IP packet and the logical structure for storing a set of service information, which is subsequently added by the enumerator 125 to organize the data transfer process: packet number, packet segment number, processing flags, and identifier of the sending PPU.

To wait for further processing, Converter 110 places the finished packets of the internal format in the input buffer 115.

The input buffer 115 is used by the transmitting side to compensate for random delays in the transmission of data packets through the PPU 160 _i resulting from deterioration or loss of communication by individual PPU 160 _i . The type of buffer used is FIFO. In the input buffer 115, packets of the input stream 10 places the Converter 110, after they are converted into an internal format.

Stream analyzer 120 sequentially reads data packets from input buffer 115, after which it logically separates the stream of IP packets received for sending into interaction flows, which are determined by a unique set of parameters: a combination of a pair of addresses and ports of the sender and receiver, as well as the type of protocol used (hereinafter - “Flow of interaction”). For the selected interaction flows, the flow analyzer 120 analyzes the transmitted data (the used data type, protocol, port, etc.) and sets the processing flags: the flag of the need for duplication, sending confirmation after receiving the packet, etc.

Then, separately for each of the interaction flows, by means of the numbering device 125, numbers of data packets and numbers of segments of packets, if any, are sequentially assigned.

After that, the numbering device directs the data packets to the switch of the op-amp 135, by means of which each of the packets is assigned the identifier of the forwarding terminal device 140 _i (DT), and thus, the packets are distributed among the corresponding PSU 160 _i for sending to the receiving side.

To send prepared data packets 12 through the PU 160 _i , the terminal switch 135 selects one PU 160 _i from the set of active (currently operating) (see FIG. 3).

After the selection procedure is completed, the packet is transmitted to the Terminal device 140 _i (hereinafter - the DT), corresponding to the selected PPU 160 _i . The package from the OS 140 _i is transmitted to the driver PPU 150 _i and then to the PPU 160 _i , which transmits it to the receiving side. The total number of packets transmitted by the Terminal Switch 135 to all operable OS 140 _i per unit of time is calculated in the process of adapting the data rate in accordance with the actual total characteristics of the data transmission channels of all operable OS.

Encapsulator 130 packages (encapsulates) internal format packets into UDP protocol packets.

The quality control of the used communication channels is carried out by regularly generating, sending and measuring the transit time of control packets from the transmitting side to the receiving side and back for each OS 140 _i . The control packets 16 created (see FIG. 1) by the control packet generator 180 and sent via the OS 140 _i contain summarized information on all the OSs, for each OS this is the following information: the ID of the OS, the time of sending, and the amount of data received through the OS from the moment of sending previous controlling interest.

By means of the control packet generator 180 of the transmitting side, the control packet 16 is formed and transmitted to the terminal switch 135, by means of which the said control packet 16 is sent via all available communication channels to the receiving side.

At the fourth stage, the receiving side receives packets, processes them, and also analyzes the parameters of the data transmission channels.

The received stream of encapsulated data and control packets 20 processes the decapsulator 210 (see FIG. 2), extracting packets of the internal format from the received UDP packets and dividing them into control packets and data packets.

The decapsulator 210 sends control packets 26 to the control packet processor of the receiving side 270, and sends data packets to the flow analyzer of the receiving side 220.

By means of the control packet processor of the receiving side 270, measurements and calculations of communication quality parameters are performed for each communication channel, at least these parameters include data rates and transmission delays.

By means of the control packet generator of the receiving side 280, regardless of the receipt of the control packet 26, the counter control packet 27 is formed from the transmitting side, in which information about the communication quality parameters of each communication channel is placed and the counter control packet 27 is sent to the transmitting side.

Using the “send time” parameter, the Control packet processor 270 on the receiving side (see FIG. 2) calculates the delay in passing the control packet 16. The control packet generator on the receiving side 280 creates a similar control packet 27 that contains the values calculated by the Control packet processor of the receiving side 270 delays and information about the amount of data received from OS 140 _i from the moment of the previous measurement, after which the control packet 27 is sent back to the transmitting side, namely to the corresponding processing counter of control packets of the transmitting side 170.

In the fifth stage, after receiving at least one counter control packet, information (data) about the communication quality parameters 19 of each of the communication channels is extracted by the control packet processor 170 of the transmitting side and the information is transmitted via the control packet generator of the transmitting side 170 to the op-amp switch 135. At the latter, taking into account the information received, the information on the communication quality parameters of each communication channel is updated and these parameters are recorded in the corresponding OS 140 _i .

Then, using the switch of the terminal devices 135, the data packets intended for sending are redistributed taking into account the change in the communication quality parameters of each communication channel.

Moreover, communication channels that do not meet the conditions of the specified parameters of communication quality are excluded from the distribution. In this case, the communication channels that were previously excluded from the distribution and have restored their characteristics are included in the distribution.

Distribution of data packets on active channels is based on the results of quality control of communication channels. The main selection criterion for OS 140 _i is the maximum probability of passing the packet with minimal delays. To implement this choice, a list of active OS is created, from which OS 140 _i that do not meet the quality criteria are discarded. The communication quality parameters used for the selection of OA 140 _i are measured and calculated by processing the control packets 17 received from the receiving side. After selection, OA 140 _i remains in the list, providing data transmission with an acceptable delay and jitter. Based on a list of opamp 140 _i with suitable transmission characteristics. A switch of the terminal devices 135 probabilistically distributes the data packets over the data transmission channels, and the probability of transmitting a data packet through the op-amp is directly proportional to the quality of the communication channel generated by the op-amp.

After a specified period of time after sending the previous control packet on the control packet generator of the transmitting side, the next control packet is formed.

To successfully restore the original data stream, the number of buffers used in ordering should be equal to the number of streams transferred.

On the flow analyzer 220 of the receiving side, the analysis of the received data is carried out and the data are logically divided into interaction flows 24, after which the interaction flows 24 are sent to the sorter 230. The flow analyzer of the receiving side 220 checks the incoming packets for belonging to the already processed interaction flows and if stream not found, initializes a new buffer in Sorter 230.

Sorter 230 arranges the received packets and sends the ordered packets to the collector 240, where they are glued segmented packets, if any, after which the data packets are transmitted to the converter 250, through which restore the original data stream; thus form a single high-speed data channel.

Sorter 230 arranges the packets from the received stream 23 according to the numbers set when sending to the interaction flows 24 (see Figure 2). Ordering is performed by selecting packets from the data streams 23 _i received from the PPU 160 _i and aligning the selected packets in the buffer in the initial sequence of the interaction stream 24 (see FIG. 4).

From the sorter 230, the interaction stream 24 is passed to the collector 240. The collector 240 glues the packets segmented before transmission and transfers them to the converter 250.

Using the converter 250, the ordered packets are converted from the internal to the original format and the restored data stream is transmitted to 25 recipients at the IP addresses and ports indicated in the packets.

An embodiment of the invention is possible where, in the fifth step, when ordering data packets, the sorter sends concurrently through all communication channels the acknowledgment of receipt of data packets to the numbering of the transmitting side, from where the lost data packets are re-sent to the receiving side, in addition, in the case of receiving a duplicated packet on the receiving side side, it is deleted by the sorter in the process of organizing the received data packets.

The re-sending of packets lost during forwarding is used for individual types of data being forwarded and is regulated by the flag of the need to send confirmation after receiving the packet, which is set by Flow Analyzer 120. When processing the packet being sent with this flag. Numbering Number 125 places a copy of the packet in its internal buffer, indicating the time the packet was sent. When receiving a packet with the specified flag, the receiving side sorter 230 (see Fig. 7, Fig. 8) generates and sends to the transmitting side a service acknowledgment packet 28 with identifiers of the interaction flow and the received packet. To increase the likelihood and minimize the delay in receipt, a confirmation packet 28 is sent in parallel through all channels generated by the op-amp to the transmitting side. On the transmitting side, the received confirmation packet 18 (corresponds to the sent confirmation packet 28) is sent to the numbering device 125. Upon receipt of the delivery confirmation, the Numbering device 125 considers the saved package to be successfully delivered and removes it from its buffer. Based on the stored packet sending time, Numbering device 125 calculates the waiting time for the confirmation packet and, if the waiting time for the confirmation packet exceeds the allowable waiting time defined in the flow parameter configuration, Numbering 125 re-sends the packet. The total number of retransmissions is also indicated in the parameters, since it depends on the permissible delays in data transfer.

It is possible that in the fifth stage, overclocking is performed for each communication channel in order to increase its throughput by transmitting more data packets per unit time than can be transmitted at the calculated speed obtained from the control packet for this communication channel. Overclocking is understood as a request for additional resources of the transmission medium in order to increase the throughput of the communication channel. The request for additional resources is implemented by overloading the communication channel, i.e. Attempts to transmit more data packets per unit of time than can be transmitted at design speed.

The channel is overclocked by constant attempts to increase the current channel capacity generated by the PND 160 _i (see FIG. 1) by issuing more data to the corresponding OS 140 _i than the PNU 160 _i is transmitting at the moment. If the channel-forming equipment has a reserve of bandwidth, it increases the channel width, compensating for the increased load on the channel. This channel overclocking method shows good results for all types of PPU 160 _i operating in a shared data transmission medium (Ethernet, WiFi, 3G / 4G).

To implement the described method, the Terminal Switch 135 receives the predicted total outgoing flow rate by multiplying the current total transmission rate of the used PPU 160 _i calculated on the basis of the analysis of the received packets from the receiving side by the value of the adjustable parameter “acceleration coefficient”.

In the process of probabilistic distribution of packets, the obtained predicted speed of the outgoing stream is used by the Terminal switch 135 as the upper limit of the total transmission rate of packets in the OS 140 _i .

When using the described channel acceleration technique, sooner or later a situation arises when the communication channel formed by the PPU 160 _i cannot cope with the increased load and the data transfer rate through this PPU 160 _i starts to decrease. In this case, there is an increase in the waiting time for sending data packets in the internal buffer of the PPU 160 _i and in the Input buffer 115, exceeding the maximum allowable delay in the transmission of data packets and, as a result, packet loss during transmission. For early tracking of such situations and minimizing their consequences, in OS 140 _i a history of changes in the values of the sending speed through the corresponding PPU 160 _i is stored. Based on this information, Terminal Switch 135 calculates the difference between the current and previous values of the channel speed, and if the difference value is below zero, it corrects the speed of the control panel 160 _i calculated based on the passage of the control packets by the difference value. In this case, the correction does not lower the adjusted speed value below the configuration parameter that determines the minimum speed for this type of PUF 160 _i .

It is also possible that at the fifth stage, for each communication channel, bandwidth is backed up, maintaining the gained speed, by transmitting as many data packets per unit of time that can be transmitted at the calculated speed obtained from the control packet for this communication channel. To ensure the availability of a sufficient number of data packets, data packets intended for transmission are duplicated (partially or completely depending on the communication channel width and the number of data packets for transmission). At the same time, they further increase the reliability and speed of data transfer.

The decision on the possibility of duplication of the transmitted data packets for automatic switching on and off of the duplication process is performed by the Terminal device switch 135 by comparing the speeds of the input stream 10 and the predicted total speed of the outgoing stream 14. If the predicted total speed exceeds the incoming by at least 10% and the sending packet has the flag of the need for duplication; the terminal switch 135 makes a positive decision on the possibility of using remote control lation. In this case, the probability of duplication of a packet having the duplication flag set is proportional to the excess of the predicted total speed of the outgoing stream over the incoming speed.

The best embodiment of the invention

The described method of forming a fault-tolerant, broadband and highly reliable communication channel is sufficient to implement a system that provides the formation of a one-way communication channel that can be used to transmit data via UDP. An example of the use of such a channel is the transmission of a video signal from an analog surveillance camera located on a remote or moving object.

To form a communication channel with two-way data transmission as described above, the receiving and transmitting parts should be located symmetrically on both sides of the channel. Examples of the use of such a communication channel are ensuring the operation of a video conferencing system on a moving object, providing high-speed access to resources of a local (global) computer network from a mobile computer (laptop) located in a remote region.

When developing the Datacrosser hardware-software complex (hereinafter referred to as the Datacrosser AIC), in particular, the described method of forming forward and reverse communication channels was used. Here, the direct communication channel is understood as the data transmission channel from the mobile part of the complex to the stationary one, and the reverse - from the stationary part of the complex to the mobile one.

The general scheme of the Datacrosser APK is shown in FIG. 5. The minimum configuration of the Datacrosser AIC includes one server (Server part of the AIC) 550 and one mobile gateway (Client part of the AIC) 520 with six built-in 30 modems (PPUs) 160 _i connected to at least two mobile operators 530.

In the mode of transmitting video data from analog CCTV cameras, Datacrosser operates according to the following scheme (see Figure 6):

1. An analog signal 61 received from an analog video camera 600, Video encoder 610 converts MPEG2TS 62 into a video stream;

2. The received video stream is distributed by the client part of APK 520 between six 30 modems connected to six mobile gateways. For transmission, UDP 63 protocol is used, an unreliable delivery mode that allows to achieve minimal transmission delays, but does not guarantee the delivery of packets of transmitted data;

3. The server side of the APK 550, connected to the data transmission channels of cellular operators via the provider's transport network and / or the Internet, on its side receives UDP packets 64 distributed between the modems and restores the original MPEG2TS 65 video stream from them;

4. The video server 620 provides consumers on the local network 690 with access to the MPEG2TS 65 video stream via the RTSP and RTMP 66 protocols.

In this scheme, the Client part of the APK 520 implements a method of processing data packets before sending them through the communication channels generated by the PPU, The server part of the APK 550 implements a method of receiving and processing data packets after their transmission through the communication channels formed by the PPU. As the PPU, 3G modems are used that are built into the mobile gateway (the client part of the agro-industrial complex).

In the data transfer mode of the following types: video signal from digital (IP) cameras, graphics transmission of video conferencing devices, mobile and stationary workstations, servers and any other IP devices, the Datacrosser agro-industrial complex works according to the following scheme:

1. The connected mobile client device 605 via DHCP 70 receives an IP address from the DHCP server 630 built into the AIC;

2. The input data stream 69 received from mobile client devices 605 for devices or applications on the local (global) network The client part of the APK 520 distributes 3G modems connected to the mobile gateway. For transmission, a data stream prepared for transmission is used, encapsulated in UDP 67 packets, transmitted in a confirmation mode (reliable delivery mode), which guarantees packet delivery due to a slight increase in the average transmission delay;

3. The server side of the APK 550, connected to the data transmission channels of cellular operators through the provider's transport network or the Internet, on its side receives a received data stream encapsulated in UDP packets, transmitted in confirmation mode 68, distributed between modems, restores interaction flows from them and sends them to consumers on the local (global) network 690.

4. The output data packet 71 received from devices or applications 630 in the local (global) network for mobile client devices 605, the Server part of the APK 550 distributes between the channels created by 3G modems connected to the mobile gateway. The same protocol and mode are used to transmit the received data stream encapsulated in UDP packets transmitted in acknowledgment mode 68;

5. The client part of APK 520, on its side, receives a data stream prepared for transmission, encapsulated in UDP packets 67 distributed between 3G modems, restores 69 interaction streams from them, and sends them to consumers - mobile client devices 605.

In this scheme, both the Client part of the APK 520 and the Server part of the APK 550 implement methods for processing data packets before sending them through communication channels and a method for receiving and processing data packets after their transmission through communication channels formed by the PPU. 3G-modems built into the mobile gateway (Client part of the agro-industrial complex) 520 are also used as PPU.

The inventive method allows the use of the following possible wired and wireless control panels: LTE modem, GSM modem, CDMA modem, WIFI module, satellite modem, optical modem, ADSL modem, Ethernet modem. The method does not exclude the use of other possible types of PPU.

The inventive method dynamically distributes the data of the network level of interaction between all communication channels in accordance with their current characteristics, which guarantees the efficient use of the bandwidth of each communication channel and allows to achieve minimal delays in data transmission. In addition, the method is universal and allows you to use almost all possible types of communication channels.

Claims (12)

1. The method of forming a data transmission channel, which consists in the fact that at the first stage they search for all active transmitting and receiving devices (PPU) of the transmitting side, determine the type of each PPU and its communication operator, assign an IP address to each identified active PPU with the formation of a communication channel ; at the second stage, the data stream intended for transmission by means of a converter is converted to the internal format and divided into packets, and, if necessary, into packet segments, the finished packets are placed in the input buffer; at the third stage, the data from the input buffer is distributed over the active PUs of the transmitting side and sent to the receiving side, at the fourth stage, receiving data packets is received at the receiving side, and then the data is converted from the internal to the original format using a converter; characterized in that at the first stage, for each control panel, a set of individual settings is set that is necessary for its optimal operation, including network connection parameters and acceptable communication quality parameters, and a terminal device (OS) is created in which these individual settings and the assigned IP address are recorded corresponding PPU, terminal devices are interconnected by a switch of terminal devices; at the third stage, the data is transferred from the input buffer to the transmitting stream analyzer, which analyzes the transmitted data and logically separates the data into interaction flows, determined by the set of parameters of the sender, receiver, and type of interaction protocol, while the stream analyzer sets the processing mode for each of the transmitted packets and requirements for the quality of transmission, then separately for each of the interaction flows, by means of a numbering device, sequentially assign numbers to data packets, and so the same numbers, if any, to the packet segments, after which the data packets are sent to the DT switch, by means of which each of the packets is assigned the identifier of the forwarding terminal device (DT) and, thus, the packets are distributed according to the corresponding control points; in addition, by means of a generator of control packets of the transmitting side, a control packet is formed, and it is transmitted to the switch of the terminal devices, by means of which the said control packet is sent via all available communication channels to the receiving side; in the fourth step, on the receiving side, by means of a decapsulator, the received stream is divided into control packets and data packets, the control packets being sent to the receiving packet control processor, and the data packets being sent to the receiving flow analyzer; by means of the control packet processor of the receiving side, for each communication channel, measurements and calculation of communication quality parameters are performed; by means of the control packet generator of the receiving side, regardless of the receipt of the control packet from the transmitting side, a counter control packet is formed in which information about the communication quality parameters of each communication channel is placed and the counter control packet is sent to the transmitting side; at the fifth stage, after receiving at least one counter control packet by the control packet control processor of the transmitting side, information about the communication quality parameters of each communication channel is extracted and the information is transmitted through the transmitting packet control packet generator to the op-amp switch; at the last, the information is updated taking into account the received information according to the parameters of the quality of communication of each of the communication channels and record these parameters in the corresponding OS; then, using the switch of the terminal devices, the data packets intended for sending are redistributed taking into account changes in the communication quality parameters of each communication channel; at the same time, communication channels that do not meet the conditions of the specified parameters of communication quality are excluded from the distribution; after a specified period of time after sending the previous control packet, the next control packet is formed on the control packet generator of the transmitting side; on the flow analyzer of the receiving side, the analysis of the received data is carried out and the data are logically divided into interaction flows, after which the interaction flows are sent to the sorter, which organizes the received packets and sends the ordered packets to the collector, where they are glued segmented packets, if any, after which the packets data is transmitted to a converter, by means of which the original data stream is restored; thus form a single high-speed data channel. 2. The method of forming a data transmission channel according to claim 1, characterized in that in the fifth step, when organizing data packets, the sorter sends concurrently through all communication channels the receipt of data packets to the numbering of the transmitting side, from where the lost data packets are re-sent to the receiving side, in addition, if a duplicate packet is received on the receiving side, it is deleted by the sorter in the process of organizing the received data packets. 3. The method of forming a data transmission channel according to claim 1, characterized in that in the fifth stage, for each communication channel, overclocking is performed in order to increase its throughput by transmitting more data packets per unit time than can be transmitted at the estimated speed obtained from control packet for this communication channel. 4. The method of forming a data transmission channel according to claims 1 to 3, characterized in that at the fifth stage, for each communication channel, bandwidth is backed up, maintaining the gained speed, by transmitting as many data packets per unit of time that can be transmitted at the estimated speed obtained from the control packet for this communication channel, and to ensure the availability of a sufficient number of data packets duplicate data packets intended for transmission. 5. The method of forming a data transmission channel according to claim 1, characterized in that an LTE modem is used as one of the control panels. 6. The method of forming a data transmission channel according to claim 1, characterized in that a GSM modem is used as one of the control panels. 7. The method of forming a data transmission channel according to claim 1, characterized in that a CDMA modem is used as one of the controllers. 8. The method of forming a data transmission channel according to claim 1, characterized in that a WIFI module is used as one of the control panels. 9. The method of forming a data transmission channel according to claim 1, characterized in that a satellite modem is used as one of the controllers. 10. The method of forming a data transmission channel according to claim 1, characterized in that an optical modem is used as one of the controllers. 11. The method of forming a data transmission channel according to claim 1, characterized in that an ADSL modem is used as one of the controllers. 12. The method of forming a data transmission channel according to claim 1, characterized in that an Ethernet modem is used as one of the controllers.

Images