RU2276461C2 - Method and device for classification of network connections - Google Patents

Abstract

FIELD: communications engineering, possible use for classification of connections.

SUBSTANCE: in method and device by means of computing block one or several distance coefficients are determined, while distance coefficients show efficient length of network connection depending on distance by air. On basis of known data about network connections, distribution coefficient of weak portions is determined, showing mutual relation to each other of weaker portions of network connection. Data transfer resource is determined to determine maximal for data transfer capacity for different types of modems. On basis of efficient length of network connection, weaker portions distribution coefficient and data transfer resources by means of computing block classification is performed (of subject network connection in accordance to its maximal data transfer capacity).

EFFECT: possible quick and flexible determining of service quality parameters.

3 cl, 9 dwg

Description

The present invention relates to a method and system for classifying network connections, wherein in the specified method and system, the geographical coordinates of the beginning and end of the classified network connection between the transmitter and receiver are known. In particular, the method relates to networks using cable connections with copper conductors.

Traditional telephony services, such as POTS (regular telephone network), traditionally connect residential premises and small businesses to the distribution station of the telephone network provider over copper wires, which are twisted together and are called a twisted pair. They were originally intended to provide transmission of analog signals, in particular, tonal and voice transmissions. However, these requirements later with the advent of Internet technology and the associated transfer of data streams have changed and are rapidly changing at present due to the need to provide real-time work and using multimedia applications both at home and at work.

Data networks, such as the intranet and the Internet, rely heavily on so-called shared media, for example, packet-oriented local area network (LAN) or wide area network (WAN) technologies, as for broadband network backbones between switches and gateways and local network connections with smaller bandwidths. It is common to use packet management systems such as bridges (LAN interface devices) or routers to provide a local area network connection to the Internet. An Internet router should be able to transmit packets based on a wide variety of protocols, such as IP (Internet Protocol), IPX (Internet Packet Exchange), DECNET (DEC Network Architecture Protocol), ApplTALK (Apple Network Protocol Stack), OSI ( open systems interconnection), SNA (IBM system architecture), etc. The integration of such networks in order to route packets globally is a problem for both service providers (providers) and manufacturers of the necessary hardware.

The most common local area network (LAN) systems work relatively well at data rates of the order of 100 Mbps. At transmission speeds above 100 Mbps, in most modern networks, the resources of network administrators, such as packet switches, are insufficient to control bandwidth allocation and user access. Of course, the usefulness of packet-oriented networks for the transmission of digital information, especially with short-term impulse transmissions, has long been recognized. Such networks usually have a point-to-point structure (point-to-point connection structure), with the packet going from one single sender to one single recipient, with each packet including at least the destination address (recipient address). A typical example of this is the well-known IP header of an IP data packet. The network responds to the data packet by sending the packet to the address of the corresponding header. Packet-oriented networks can also serve to transmit data types that require a continuous stream of data, such as high-quality tonal and audio transmissions or video transmissions. The commercial use of networks makes it particularly desirable that packet-oriented transmission is simultaneously possible to multiple destinations. An example of this is the so-called packet broadcast for transmitting video and audio data. In this way, pay-TV can be realized, that is, paid broadcasting of video data transmissions over the network.

However, in next-generation applications, such as real-time and multimedia applications, with their even higher demands on the bandwidth that must be guaranteed at any given time, packet-oriented networks run into limitations. So the next generation of networks should be able to dynamically change the configuration of networks in order to constantly guarantee the user a predetermined bandwidth for the required or agreed quality parameters (QoS - quality of service). Such quality parameters include, for example, access guarantee, access performance, error tolerance, data transfer reliability, etc. between all possible end systems. New technologies, such as asynchronous transfer mode (ATM), should at the same time help to create the necessary prerequisites for both the private intranet network and the public Internet access network in the long-term development of networks. These technologies promise a more economical and scalable solution for such high-quality QoS-guaranteed connections.

Changes in future systems will affect, in particular, the flow of data. The data stream is currently based on a server-client model, that is, data from multiple clients is transmitted to one or more or from one or more network servers. Clients usually do not create a direct data connection, but communicate with each other through network servers. This type of connection will be important in the future. Despite this, it should be expected that the amount of data transmitted between peers will increase significantly in the future. Since the ultimate goal of networks to satisfy the requirements will be a truly decentralized structure in which all systems will be able to act as a server and as a client, the data flow over connections between peer nodes will increase. Thus, the network will have to create more direct connections to various peer-to-peer nodes, and, for example, desktop computers will make direct connections via the Internet backbones.

Thus, it is clear that for promising applications, it is becoming increasingly important to ensure that the user can predefined QoS parameters and large bandwidths.

For data transmission to the end user, in particular, a traditional public switched telephone network (PTSN) and / or mobile landline public access network (PLMN) are used, which were originally designed to transmit only tonal signals, and not to transmit such volumes of digital data. Moreover, in determining the QoS parameters that the provider or telephone service provider can guarantee the user, the so-called “last mile” plays a decisive role. This term defines the distance between the last distribution station of the public switched telephone network and the end user. This last mile in some cases is formed by high-performance fiberglass cables, but most often it is based on ordinary cables with copper conductors, such as a cable with a core diameter of 0.4 or 0.6 mm. In addition, cables are not always laid underground in a protected conduit, and there are also sections of their laying above the terrain on telephone poles, etc. The consequence of this is additional interference.

An additional problem in determining the maximum QoS parameter is the so-called crosstalk issue. This problem occurs when the signal is modulated in the area, for example, from the end user to the distribution station of the telephone service provider and vice versa. For modulating digital signals in the prior art, for example, xDSL (digital subscriber line) technologies such as ADSL (asymmetric digital subscriber line), SDSL (symmetric digital subscriber line), HDSL (high speed digital subscriber line), or VDSL (ultra high speed digital subscriber line) are known line). Mentioned crosstalk is a physical phenomenon that occurs when modulating data transmitted over a copper cable. Nearby conductors inside one copper cable receive due to electromagnetic interaction the paired components of the signals that are generated by the modem. This leads to the fact that xDSL modems transmitting along adjacent cores of a cable create mutual interference. Distinguish between near-end crosstalk (NEXT), which refers to the unintentional input of transmitter signals at one end into signals at the receiver at the same end, and far-end crosstalk (FEXT), which refers to unintentional input of signals during transmission to the receiver at the other end moreover, the signals during transmission are introduced into the signals of neighboring copper pairs and appear as noise in the receiver.

Although there are currently many studies of xDSL crosstalk, such as Spectral management on metallic access networks; Part 1: Definitions and signal library, ETSI, TR 101 830, September 2000, due to the complexity of the crosstalk phenomenon and other noise parameters Currently, there are few practical and technically simple to use, as well as cost-effective tools for determining the QoS parameters for a specific end-user on the network. In the prior art, remote sensing systems are proposed by various companies, for example, Acterna (WG SLK-11/12/22, Eningen uA, Germany), Trend Communications (LT2000 Line Tester, www.trendcomms.com, Buckinghamshire, Great Britain), etc. .d. At the same time, the maximum data transfer rates at the last mile are determined by direct measurements using remote measurement systems: a digital signal processor is installed on each local distribution station of the telephone network provider (for example, several thousand in Switzerland). By means of a digital signal processor, the so-called "asymmetric measurement" is carried out, since the user does not need to install any devices on the other side of the last mile. But fundamentally, measurements based on the principle of "symmetric measurements" are also possible. But it is necessary to install measuring instruments at both ends of the line. The published international application WO 01/41324 A1 (Qwest Communications International Inc.) describes a method for classifying network connections. In this case, the geographical length of a separate line is determined, for example, by measuring a section of a route, and classification is carried out by comparison with known measurements of network connections.

The disadvantages of this prior art include, among other things, the high costs associated with the necessary installation of remote measurement systems in each local distribution station, inaccurately known uncertainty or unknown measurement error, since measurements are carried out only on one side (asymmetric measurements), and to determine errors would require two-way measurements. Bilateral measurement would be impractical due to labor, time and financial costs. In addition, in the prior art, algorithms with a hardware or software implementation for calculating or predicting the maximum possible bit rates of a network connection are not known. The installation of remote measurement systems on a smaller number of central distribution stations instead of local terminal distribution stations shows that the measurements are characterized by such a high degree of uncertainty that they are not suitable for determining the maximum possible data transfer rate for a particular line to the end user.

The objective of the invention is to provide a new method and device for classifying network connections that do not have the above-described disadvantages. In particular, the opportunity should be created to quickly and flexibly determine the QoS quality of service parameters and specifically the maximum guaranteed data transfer rates for a particular user, without requiring disproportionate technical, labor and financial costs. This should be ensured even in the case when the network includes only inaccurate complex structures of connections, such as, for example, the last mile.

In accordance with the invention, this problem is solved in particular due to the features of the independent claims. Preferred embodiments follow from the dependent claims and the description.

In particular, these results are achieved in the invention in that for the classification of network connections, the geographical coordinates of the transmitter and receiver of the classified network connection being known, based on the known network connection data, one or more distance coefficients are determined by the computing unit and the received data correlated with the determined probability transmitted to the storage medium of the computing unit, and the distance coefficients indicate the effective length of the networks connection depending on the distance through the air, and the determined probability of whether a certain length of a network connection is shorter or greater than its effective network length is determined by the reliability coefficient, based on one or more distance factors, reliability coefficient and geographical coordinates of the transmitter and the receiver of the classified network connection using the computing unit determines the effective length of the network connection and is transmitted to the storage medium subtraction the casting unit in accordance with the classified network connection, based on the known data of the network connection, at least one attenuation distribution coefficient is determined and transmitted to the storage medium of the computing unit, at least one attenuation distribution coefficient indicates a relation to each other attenuation of various sections of the network connection, the reserves of data transmission resources are determined to determine the maximum data transmission throughputs for various types in modems and stored on the data carrier of the computing unit in accordance with the physical length and thickness of the network connection cable, and using the power measuring device, the energy spectra for the modems are measured, the effective signal levels and the corresponding noise levels are determined using the computing unit based on the energy spectra and Gaussian transform module based on signal levels and noise level for various modulations of data transmission and / or modulating codes reserves of transmission resources are allocated for a predetermined bit rate, and based on the effective length of the network connection, the attenuation distribution coefficient and reserves of data transmission resources, the classified network connection is classified by the computing unit in accordance with its maximum data transmission capacity. The advantage of the invention, in particular, is that the method and system for the first time provide a simple and quick determination of the reserves of data transmission resources, without requiring large technical, labor and time costs. In particular, it is possible to correct the uncertainties due to the mentioned correction without requiring, as in the known remote measurement systems for measuring the reserves of data transmission resources and / or bit rates, the adjustment in each local distribution station of a different, precisely unknown uncertainty or unknown error, which Due to one-sided measurements, it is difficult to evaluate, since two-sided measurements would be necessary to determine the error.

In one embodiment, the slope coefficient (angular coefficient) and abscissa are determined as distance coefficients using a computing unit, and a linear relationship between the air distance and the effective length of the network connection is determined. This embodiment has, among other things, the advantage that it is sufficient for most dependencies of network structures and can produce results within the required accuracy. For a specialist, this is more than unexpected, since it was not expected that a linear function would be sufficient for such complex dependencies within the desired accuracy. In particular, linear relationships are easier and faster to determine and process than non-linear ones.

In another embodiment, the computing unit determines the distance coefficients as parameters of a polynomial of at least degree 2. This embodiment, among other things, has the advantage that it provides any accuracy depending on the order of the polynomial used and the required maximum deviation for the relationship between the distance through the air and the effective length of the network connection. What was unexpected in this case was that essentially no polynomials of a very high order are required to satisfy the requirements of this method.

In another embodiment, a probability ranging from 0.85 to 0.95 is selected by means of a reliability coefficient. The advantage of this option is that the error tolerance and maximum deviation are limited by the accuracy required for this method and device.

In one embodiment, the reliability coefficient has a value ranging from 700 to 800. The unit of measurement for this option is a meter. The advantages of this option are the same as for the previous embodiment.

In yet another embodiment, a linear dependence of the attenuation relative to each other is determined by the attenuation distribution coefficient. This option has the advantage that it is sufficient for most dependencies of network structures and can provide results within the required accuracy. For a specialist, this is more than unexpected, since it was not expected that a linear function would be sufficient for such complex dependencies within the desired accuracy. In particular, linear relationships are easier and faster to determine and process than non-linear ones. This embodiment is particularly suitable for networks with connections consisting of two different cables with different core thicknesses, for example, cables with core diameters of 0.4 mm and 0.6 mm.

In another embodiment, the computing unit determines the adjusted reserves of data transmission resources by means of at least one correction factor based on the stored reserves of data transmission resources and stores them in accordance with the physical lengths and thicknesses of the cores of the network connection cables on the data carrier of the computing unit, the correction factor includes the average deviation of the stored reserves of data transmission resources relative to the effective reserves of resources giving data. This embodiment has the advantage, inter alia, of taking into account factors that cause an additional deviation of the calculated reserves of data transfer resources relative to the effective reserves of data transfer resources. This includes, for example, deviations due to high-quality or poor-quality execution of the modem by the manufacturer or additional internal noise due to sampling noise or poor-quality mutual coordination of the correction unit.

In yet another embodiment, the noise levels are determined by the computing unit depending at least on the crosstalk parameters and the number of interference sources based on the energy spectra.

In another embodiment, at least one correction coefficient reproduces a non-linear dependence on the physical lengths and / or thicknesses of the cable cores, that is, the correction coefficient can be a non-linear function, for example, a polynomial function whose degree is greater than 1. An advantage of this embodiment is that in this way it is possible to take into account and correct more complex dependences than those described by linear correction factors.

In one embodiment, the energy spectrum is measured depending on the transmission frequency for the types of ADSL, and / or SDSL, and / or HDSL, and / or VDSL modems. Possible types of SDSL modems may include at least the type of G.991.2 modem, and / or types of ADSL modems may include at least the type of G.992.2 modem. Using the Gaussian transform module, the reserves of data transmission resources can be determined, at least for data transmission modulations of the form 2B1Q (2 binary, 1 quaternary), and / or CAP (amplitude / phase modulation without a carrier), and / or DMT (digital multitonal) , and / or PAM (pulse amplitude modulation). Also, through the Gaussian transform module, the reserves of data transmission resources can be determined, at least for coding using trellis code modulation. This embodiment has, among other things, the advantage that, for the types of xDSL modems, the above-mentioned modulations of data transmission and coding using trellis modulation use the usual standard technologies that are readily available on the market and which are widely used as in Europe and in the USA.

In particular, the above objectives are achieved in the present invention in that the known geographical coordinates of the transmitter and receiver of the classified network connection are used to classify network connections, based on the known data on network connections, one or more distance coefficients are determined by the computing unit, and the received data correlated with defined probability are transferred to the storage medium of the computing unit, and the distance coefficients indicate the effects the apparent length of the network connection depending on the distance through the air, and the determined probability of whether a certain length of the network connection is shorter or greater than its effective network length is determined by the reliability coefficient, based on the distance coefficients, the reliability coefficient and the geographical coordinates of the transmitter and receiver classifiable network connection using a computing unit determines the effective length of the network connection and is transferred to the storage medium will calculate In accordance with the classified network connection, at least one attenuation distribution coefficient is determined on the basis of the known network connection data and transmitted to the data carrier of the computing unit, at least one attenuation distribution coefficient indicates the relation to each other friend weakening of various sections of the network connection, bit rates are determined to determine the maximum data throughputs for various types of modems stored on the data carrier of the computing unit in accordance with the physical length and thickness of the network connection cable, and using the power measuring device, the energy spectra for the modems are measured, the effective signal levels and the corresponding noise levels are determined using the computing unit based on the energy spectra, and using the Gaussian module transformations based on signal levels and noise level for various modulations of data transmission and / or modulating codes are determined bit rates for a predetermined margin of transmission resources, and based on the effective length of the network connection, the attenuation distribution coefficient and reserves of data transmission resources, a classified network connection is classified by the computing unit in accordance with its maximum data transmission capacity. The advantage of this embodiment is, inter alia, that the method and system for the first time provide a simple and quick determination of the bit rate without requiring large technical, labor and time costs. In particular, it is possible to correct the uncertainties due to the mentioned correction, without requiring, as in the known systems, remote measurement for measuring the reserves of data transmission resources and / or bit rates, the adjustment in each local distribution station of a different, definitely unknown uncertainty or unknown error, which, due to one-sided measurements, is difficult to evaluate, since two-sided measurements would be necessary to determine the error.

In one embodiment, the slope coefficient (angular coefficient) and abscissa are determined as distance coefficients using a computing unit, and a linear relationship between the air distance and the effective length of the network connection is determined. This embodiment has, among other things, the advantage that it is sufficient for most dependencies of network structures and can produce results within the required accuracy. For a specialist, this is more than unexpected, since it was not expected that a linear function would be sufficient for such complex dependencies within the desired accuracy. In particular, linear relationships are easier and faster to determine and process than non-linear ones.

In another embodiment, the computing unit determines the distance coefficients as parameters of a polynomial of at least degree 2. This embodiment, among other things, has the advantage that it provides any accuracy depending on the order of the polynomial used and the required maximum deviation for the relationship between the distance through the air and the effective length of the network connection. What was unexpected in this case was that essentially no polynomials of a very high order are required to satisfy the requirements of this method.

In another embodiment, a probability ranging from 0.85 to 0.95 is selected by means of a reliability coefficient. The advantage of this option is that the error tolerance and maximum deviation are limited by the accuracy required for this method and device.

In one embodiment, the reliability coefficient has a value ranging from 700 to 800. The unit of measurement for this option is a meter. The advantages of this option are the same as for the previous embodiment.

In yet another embodiment, a linear dependence of the attenuation relative to each other is determined by the attenuation distribution coefficient. This option has the advantage that it is sufficient for most dependencies of network structures and can provide results within the required accuracy. For a specialist, this is more than unexpected, since it was not expected that a linear function would be sufficient for such complex dependencies within the desired accuracy. In particular, linear relationships are easier and faster to determine and process than non-linear ones. This embodiment is particularly suitable for networks with connections consisting of two different cables with different core thicknesses, for example, cables with core diameters of 0.4 mm and 0.6 mm.

In another embodiment, the computing unit determines the corrected bit rates using at least one correction coefficient based on the stored bit rates and stores them, correlated with the corresponding physical lengths and thicknesses of the cores of the network connection cables, on the data carrier of the computing unit, the coefficient includes the average deviation of the stored bit rates relative to the effective bit rates. This embodiment has the advantage, inter alia, of taking into account factors that cause an additional deviation of the calculated bit rates from the effective bit rates. This includes, for example, deviations due to high-quality or poor-quality modem performance by the manufacturer or additional internal noise due to sampling noise (analog-to-digital conversion) or poor-quality mutual coordination of the correction unit.

In one embodiment, the energy spectrum is measured depending on the transmission frequency for the types of ADSL, and / or SDSL, and / or HDSL, and / or VDSL modems. Possible types of SDSL modems may include at least the type of G.991.2 modem, and / or types of ADSL modems may include at least the type of G.992.2 modem. By means of a Gaussian transform module, data transfer resource reserves can be determined, at least for data transfer modulations of the form 2B1Q, and / or CAP, and / or DMT, and / or PAM. Also, through the Gaussian transform module, the reserves of data transmission resources can be determined, at least for coding using trellis code modulation. This embodiment has, among other things, the advantage that, for the types of xDSL modems, the above-mentioned modulations of data transmission and coding using trellis modulation use the usual standard technologies that are readily available on the market and which are widely used as in Europe and in the USA.

In another embodiment, at least one correction coefficient reproduces a non-linear dependence on the physical lengths and / or thicknesses of the cable cores, that is, the correction coefficient can be a non-linear function, for example, a polynomial function whose degree is greater than 1. An advantage of this embodiment is that in this way it is possible to take into account and correct more complex dependences than those described by linear correction factors.

In yet another embodiment, a bit rate for data transfer resource reserves between 3 and 9 dB is determined by a Gaussian transform module. This embodiment, inter alia, has the advantage that the range between 3 and 9 dB provides reception with a QoS parameter that satisfies most requirements. In particular, the specified range of data transfer resource reserves between 3 and 9 dB provides an optimization of the bit rate with respect to other QoS parameters.

In yet another embodiment, a bit rate for a data transfer resource margin of 6 dB is determined by a Gaussian transform module. This embodiment, among other things, has the same advantages as in the previous described embodiment. In particular, the specified margin of data transfer resources of 6 dB provides optimization of the bit rate in relation to other QoS parameters.

In addition, it should be noted that the claimed invention, along with the method corresponding to the invention, also relates to a device for implementing this method.

The following examples describe embodiments of the claimed invention. Examples of execution are illustrated by drawings, which show the following:

FIG. 1 is a block diagram showing the architecture of an embodiment of a system of the invention for determining data resource reserves or bit rates for network connection 12 with a specific physical length 13 between transmitter 10 and receiver 11.

FIG. 2 is a schematic representation of crosstalk with near-end crosstalk (NEXT) 51, which relates to the unintentional input of signals 50 of transmitter 10 at one end to signals 50 at receiver 11 at the same end, and far-end crosstalk (FEXT) 52, which refers to the unintentional input of signals 50 during transmission to a receiver 11 at the other end, wherein signals 50 during transmission are input to signals 50 of adjacent copper pairs and appear as noise in the receiver 11.

FIG. 3 is a schematic representation of a transmission section of a network connection depending on the transmission rate (bit rate) for ADSL modems, as it can be obtained using the system of the invention. Reference numerals 60 and 61 denote various noise conditions.

FIG. 4 is a schematic representation of the so-called “last mile” of a public switched telephone network (PSTN), which typically exists between an end user at home with a network that should be accessible through a public switched telephone network.

FIG. 5 is a diagram of an example of data sampling for an existing network, wherein the data sampling includes 200,000 measured network connections of the last mile of the telephone network.

FIG. 6 is a diagram of the average deviation of the effective length D _{e of the} network connection from the calculated length D _{a of the} network connection. The X axis indicates the average deviation ΔD in meters, and the Y axis indicates the value of the applied data sample, that is, the number N of known network connections.

FIG. 7 is a schematic representation of the relationship R _{1 of a} copper cable t ₁ with a core diameter of 0.4 mm to a copper cable t ₂ with a core diameter of 0.6 mm at the last mile in a public telephone network. The X axis indicates the effective length D _{e of the} network connection, that is, its physical length, and the Y axis indicates the percentage R _{t of the} corresponding type of cable.

FIG. 8 is a diagram of an example of determining 2011/2012 of one or more distance factors, as well as a reliability coefficient. By analogy with FIG. 5, the X axis indicates the effective length D _{e of the} network connection in meters, and the Y axis indicates the air distance of the network connections D _a , also in meters.

FIG. 9 is a flowchart of the claimed method. Four-digit reference numerals refer respectively to FIG. 9.

In FIG. 1 shows an architecture that can be used to implement the invention. In this exemplary embodiment, for the method and device for classifying network connections, the geographical coordinates of the transmitter 10 and receiver 11 of the classified network connection 12 are known (block 1000 in FIG. 9). The coordinates can, for example, indicate degrees of longitude and latitude with sufficient accuracy, but other coordinates or location data can be used to indicate the relative geographical position of the transmitter 10 and receiver 11. For example, to be able to determine if a particular network connection is working for a connection For example, xDSL connection, you need to know the effective cable length up to a known deviation. In practice, with reasonable costs (financial, time, labor and material), only air distance can be determined. Using the coordinate data or location data for the relative geographical location of the transmitter 10 and the receiver 11, the distance between the transmitter 10 and the receiver 11 is determined by the computing unit 30, which, for example, can be stored on the storage medium of the computing unit 30. The computing unit 30 determines ( 3010) based on a sample of data (4010) from known data (5000) about network connections, one or more distance factors (2011). The process for carrying out the method of the invention is shown in FIG. 9, which refers to the four-digit reference position. The data 5000 may be, for example, experimentally determined data or otherwise known data related to network connections, which include air distance and the effective physical lengths of these network connections. The distance coefficients 2011 are thereby determined depending on the probability, the probability can be determined and describe the effective length D _{e of the} network connection depending on the air distance D _a . Then, the distance coefficients 2011, correlated with the determined probability, can be transferred to the data carrier of the computing unit 30. As the distance coefficients 2011, the slope coefficient (angular coefficient) and abscissa can be determined using the computing unit 30, and a linear relationship between the air distance is determined D _a and effective length D _{e of the} network connection. But it is also possible to determine the distance coefficients 2011 as parameters of a polynomial of the 2nd or higher degree using computing unit 30. A determined probability that can be established by the reliability coefficient 2012 indicates whether the calculated network connection length D _e is less than or greater than the effective network connection length D _e . Probability can be selected through a reliability coefficient ranging from 0.85 to 0.95. The reliability coefficient at the mentioned probability value in the case of the last mile (described below) has, for example, a value ranging from 700 to 800, and the unit of measurement is a meter.

In FIG. 5 shows an example of data sampling for an existing network, the data sampling includes 200,000 measured network connections of the last mile of the telephone network. In this network, there are connections mainly from traditional telephone connections with a copper cable with core diameters of 0.4 mm and 0.6 mm. The example clearly illustrates the correlation, although due to the complexity of some network structures, specialists would expect a more complex dependency. The X axis indicates the effective length D _{e of the} network connection in meters, and the Y axis indicates the air distance of the network connections D _a , also in meters.

In FIG. 8 is an example of determining one or more distance coefficients 2011, as well as a reliability coefficient 2012. By analogy with FIG. 5, the X axis indicates the effective length D _{e of the} network connection in meters, and the Y axis indicates the air distance of the network connections D _a , also in meters. Data points can be selected (4010) from a data sample with known 5000 network connection data. The determination of the distance coefficients 2011 as well as the reliability coefficient 2012 can occur, for example, by means of a fitting module. In this example, a linear relationship was determined between the air distance D _a and the effective length D _{e of the} network connection, and the angular coefficient a and abscissa b are determined as distance coefficients 2011 by the computing unit 30. Abscissa b is defined for different connection points (for example, city, suburb, rural area, highlands), as well as for different connection areas (for example, main distributor, junction box, transition point, etc.). The effective distance is then obtained as follows: D _e = y = aD _a + b. For y, approximately 50% of the calculated network connections are shorter than the effective network connections, i.e. with a probability of 0.5. Reliability coefficient 2012 S was also chosen linear, that is, as a constant. Thus, we obtain D _e = y _s = aD _a + b + S. Using S, it is possible to determine the probability of whether the network connection length is shorter or longer than its effective network length D _e . In the example shown, at y _s according to FIG. 8 probability through the coefficient of reliability 2012 S was set to 0.9. In this example, for the slope a = D _e / D _a for the last mile of the traditional telephone network, the following values were found: for urban conditions a _s = 1.27, for suburban conditions a _v = 1.28, for rural conditions a _l = 1.30 and for mountainous conditions a _g = 130. For a mixed data block (city, suburbs, rural areas, highlands), the value a _all = 1.30 is determined. Similarly, we obtain the values for b _s = 200, b _v = 355, b _l = 372, b _g = 391 and b _all = 328, and b is indicated in meters. The standard deviations σ for this embodiment are as follows: σ _s = 333, σ _v = 569, σ _l = 682, σ _g = 527, and σ _all = 598. The standard deviation σ determines the scattering of the differences between the effective network connection length and the calculated network deviation length. The average deviation in meters of the effective length D _{e of the} network connection from the calculated length D _{a of the} network deviation is approximately independent of the length of the network connection and is shown in FIG. 6 for this embodiment. The X axis indicates the average deviation ΔD in meters, and the Y axis indicates the value of the applied data sample, that is, the number N of known network connections. In order to obtain a probability of 0.9, this is provided in this embodiment for the reliability coefficient S, for example, S _s = 360, S _v = 640, S _l = 850, S _g = 670 and S _all = 730. In order to obtain a probability of 0.95, this is provided in this embodiment for the reliability coefficient S, respectively, S _s = 490, S _v = 1100, S _l = 1330, S _g = 930 and S _all = 1210.

Based on one or more distance factors 2011 and reliability factors 2012, using the geographical coordinates of the transmitter 10 and receiver 11 of the classified network connection 12, the computing unit 30 determines (1010) the effective length of the network connection, that is, its physical length, and transfers it to the computing data carrier unit 30 in accordance with the classified network connection 12. The physical length refers to the effective cable length, and not, for example, the length through the air, between the transmitter 10 and the receiver 11. The network connection 12 may consist of analogue transmission media, such as cables with copper conductors. In this embodiment, for example, copper cables with a core diameter of 0.4 or 0.6 mm were used, as they are typically used for the last mile of a public switched telephone network (PSTN). The last mile is shown schematically in FIG. 4. Reference numeral 70 denotes a router to a network that is connected, for example, via 10BT Ethernet 77 and a public switched telephone network (PSTN) 72 to a server 71 of the modem terminal. The server 71 of the modem terminal may be a DSL access multiplexer (DSLAM). As mentioned, reference numeral 72 denotes a public switched telephone network (PSTN) to which the modem terminal server 71 is connected, for example, with fiberglass cable 78. In addition, the public switched telephone network 79 and, accordingly, the modem terminal server 71 are typically connected via cable 79 with copper conductors through a telephone box 73 are connected to a modem 74 of a personal computer (PC) 75. Reference position 79 here refers to the so-called “last mile” from the distribution station of the telephone network provider End User. The end user 76 can thereby use his PC to directly access the router 70 through the described connection. The most common telephone wires can consist, for example, of a pair of 2-2400 copper wires. But other analog transmission media can be used, in particular copper cable, for example, with different core diameters. It must be pointed out that network connections 12 can not only have different diameters or thicknesses 114, 142, 143, 144, respectively, but also that a separate network connection can consist of a combination of cables with different diameters or thicknesses of wires, that is, that the network connection includes several separate cables with different core thicknesses.

If the network consists of a combination of cables with different diameters or thicknesses of the conductors, then based on a 4020 data sample selected from the known network connection data 5000, at least one attenuation distribution coefficient 2020 will be determined (3020), which will be transferred to the computer storage medium block 30, wherein at least one attenuation distribution coefficient 2020 indicates a ratio of attenuations of different portions of a network connection. The attenuation distribution coefficient 2020 may be defined as a linear coefficient. At least one attenuation distribution coefficient 2020, if necessary, may include non-linear dependence. In this embodiment, network connections include copper cables with core diameters of 0.4 or 0.6 mm, as they are typically used for the last mile. Since only two types of cables are used, the determination of only one attenuation distribution coefficient 2020 is sufficient. Connection cables exhibit, in accordance with their various diameters, various electrical properties and various attenuation. Therefore, for this method it is important that at least the ratio of the shares of the copper cable with a core diameter of 0.4 mm and a copper cable with a core diameter of 0.6 mm for a particular network connection is known with the required accuracy. The public switched telephone network is usually designed in such a way that the full DC impedance (DC) lies within certain limits. This property is used to determine when the user picks up the phone to make a call. If the phone is used, that is, the user picked up the phone, then the phone changes its impedance, and this change is detected at the central station. Therefore, usually for longer lines, more cable with a core diameter of 0.6 mm is used (since the resistance Ω is less), and for short distances, more cable with a core diameter of 0.4 mm is used. Thus, the ratio of the thicknesses of the cable cores can be determined approximately in accordance with the logic of phenomena. In particular, the computing unit 30, using a fitting module based on known network connection data 5000, can determine (2020) a function of the attenuation distribution coefficient depending on the length of the connection. In this embodiment, a linear coefficient was used as the coefficient 2020 of the distribution of attenuation, with:

D _e ≤10:

Moreover, L _0.4 indicates the proportion of cable with a core diameter of 0.4 mm in km, and L _0.6 indicates the proportion of cable with core diameter of 0.6 mm also in km as a function of D _e (D _e is the effective length of the network connection) . FIG. 7 schematically represents the relationship R _t , where t ₁ is the proportion of copper cable with a core diameter of 0.4 mm, and t ₂ is a copper cable with core diameter of 0.6 mm. The X axis indicates the effective length D _{e of the} network connection, that is, its physical length, and the Y axis indicates the percentage R _{t of the} corresponding type of cable. You can see that the proportion of copper cable with a core diameter of 0.6 mm for distances over 10 km increases to 100%, that is, the network connection essentially exclusively consists of a copper cable with a core diameter of 0.6 mm. Based on the function for the attenuation distribution coefficient for the classified network connection, (1020) the attenuation distribution coefficient is determined depending on the length of the connection 2020 and the effective length of the network connection and transferred to the storage medium of the computing unit 30 in accordance with the classified network connection 12.

In the next step, the reserves of data transfer resources 2030 are determined (1030) for determining the maximum data transmission throughputs and are stored on the storage medium of the computing unit 30, respectively, to different types of modems and physical cable length 13 and cable thickness 141, 142, 143, 144 of network connection 12. For this, the energy spectrum PSD _Modem (f) is measured depending on the transmission frequency f for possible types of modems 101, 102, 103, 104 by means of a power measuring device 20 and transferred to the storage medium of the computing unit and 30. The energy spectrum is also denoted by the term “power spectral density” (PSD) and reproduces a continuous spectrum of frequencies for a certain bandwidth, that is, the total power of a certain bandwidth divided by a specific bandwidth. The division by the bandwidth corresponds to the standardization. Therefore, the power spectral density PSD is a function of frequency f and is usually given in watts per hertz. To measure power by means of the power measuring device 20, a simple analog-to-digital converter can be used in the receiver 11, the voltage being applied to the resistance. To modulate digital signals on line 12, for example, from the end user to the distribution station of the telephone network provider and vice versa, various types of modulation can be applied. In the prior art, for example, xDSL (digital subscriber line) technologies are known, the two main representatives of which are ADSL (asymmetric digital subscriber line) and SDSL (symmetric digital subscriber line). Other representatives of xDSL technology are HDSL (High Speed Digital Subscriber Line) or VDSL (Ultra High Speed Digital Subscriber Line). XDSL technologies are highly developed modulation schemes used to modulate data transmitted in copper wire transmission lines or other analog transmission media. XDSL technologies are sometimes referred to as “last mile technologies”, in particular because they usually serve to connect the last telephone distribution station to the end user in a bureau or at home, and are not used for connections between individual telephone distribution stations. XDSL technology is similar to ISDN (Integrated Services Digital Network) in that it can operate on existing copper wire transmission lines, and both require a relatively short distance to the next distribution station of the telephone network provider. XDSL technology provides, however, higher data rates than ISDN. xDSL achieves data rates of up to 32 Mbps for the downstream speed (bit rate when receiving data, i.e. when modulating) and speeds from 32 kbps to 6 Mbps for the upstream speed (bit rate, then there when demodulating), while ISDN supports data transfer rates per channel of about 64 kbit / s. ADSL has recently become a very popular technology for modulating data transmission lines on copper wires. ADSL supports data rates from 0 to 9 Mbit / s for speeds in the downstream direction and from 0 to 800 kbit / s for speeds in the upstream direction. ADSL is called asymmetric DSL, as it supports different bit rates in the upstream and downstream. SDSL, or symmetric DSL, is called, in contrast, symmetric because it supports the same speeds for upstream and downstream data streams. SDSL provides the ability to transfer data at speeds up to 2.3 Mbps. ADSL transmits digital pulses in the high frequency range of a copper cable. Since these high frequencies are not used for normal transmission of tones in the audible range (for example, voices), ADSL can simultaneously act for transferring telephone calls over the same copper cable. ADSL technology is most widely distributed in North America, while SDSL technology is primarily developed in Europe. ADSL and SDSL require specially equipped modems for this. HDSL is representative for Symmetric DSL (SDSL). The standard for symmetrical HDSL (SDSL) is currently G.SHDSL, known as G.991.2, developed as an international standard by the CCITT Committee of the International Telecommunication Union (ITU). G.991.2 supports the reception and transmission of a symmetrical data stream over a simple pair of copper wires with transmission rates from 192 kbit / s to 2.31 Mbit / s. The G.991.2 standard was designed in such a way that it includes the properties of ADSL and SDSL and supports standard protocols such as IP (Internet Protocol), in particular modern versions of IPv4 and IPv6 or IPng (Internet Engineering Task Force - IETF ), as well as TCP / IP (Transmission Control Protocol), ATM (Asynchronous Transfer Mode), T1, E1, and ISDN. As the latest xDSL technology, VDSL (Ultra High Speed Digital Subscriber Line) technology is mentioned here. VDSL transmits data in the range of 13-55 Mbit / s over short distances (usually in the range of 300-1500 m) over a twisted pair copper cable. For VDSL, the relation is true, that the shorter the distance, the higher the transmission speed. As the final section of the network, VDSL is used to connect a user's office or home with a neighboring optical network unit called an optical network unit (ONU), which is typically connected to a main fiberglass network (trunk), for example, of a company. VDSL provides the user access to the network with the maximum bandwidth over normal telephone wires. The VDSL standard is not yet fully defined. So, there are VDSL technologies that have a line coding scheme based on DMT (discrete multi-tone signal), and DMT is a multi-carrier system, which is very similar to ADSL technology. Other VDSL technologies have a quadrature amplitude modulation (QAM) line coding scheme, which, in contrast to DMT, is more economical and requires less energy. For this embodiment, modem types may include ADSL and / or SDSL and / or HDSL and / or VDSL modem types (101, 102, 103, 104). In particular, possible types of SDSL modems (101, 102, 103, 104) may include at least one type of G.991.2 modem, and / or types of ADSL modems (101, 102, 103, 104) may include at least one type of G.992.2 modem. However, it is understood that this list should in no way be considered limiting the scope of protection of the invention, and, conversely, other types of modems are possible.

Using the computing unit 30, the attenuation H is determined for various physical lengths 13 and core thicknesses of cables 141, 142, 143, 144, such as, for example, 0.4 mm and 0.6 mm of network connection 12, and effective signal levels S (f) in the receiver 11, based on the attenuation of H (f) and the energy spectrum of PSD (f), correlated with the corresponding physical lengths L13 and core thicknesses of the cables D 141, 142, 143, 144, are stored in the first list on the storage medium of the computing unit 30. Attenuation H (f, L, D), like the effective signal level S (f), is a function depending on the frequency f. The signal sent from transmitter 10 corresponds to PSD _Modem (f), while the effective signal level S (f) = PSD (f) H ² (f, L, D) is received at the receiver. In the second list on the data carrier of the computing unit 30, the noise level N (f) is stored, correlated with the corresponding physical lengths 13 and the thickness of the cores of the cable 141, 142, 143, 144 of the network connection, and the noise level N (f) 40 is determined by the computing unit 30 depending on at least the Xtalktype crosstalk parameters and the number A of interference sources based on the PSD energy spectrum. I.e:

The sum is taken at index i for all interfering modulations (SModem), depending on their crosstalk parameter Xtalktype, which operate on parallel connections of a given network connection. PSD _{SModem (i)} is the energy spectrum of the i-th modem. Hxr represents attenuation due to crosstalk. As mentioned, the issue of crosstalk is related to the physical phenomenon that occurs when modulating data transmitted over a copper cable. The adjacent copper cable conductors inside the copper cable receive due to electromagnetic interaction the paired components of the signals that are produced by the modems. This leads to the fact that xDSL modems that transmit on adjacent wires create mutual interference. Crosstalk as a physical action is negligible for ISDN (frequency range up to 120 kHz), but is significant, for example, for ASDL (frequency range up to 1 MHz) and is a decisive factor for VDSL (frequency range up to 12 MHz). As described, the telephone lines used consist of copper cores from 2 to 2400. To, for example, be able to use four pairs, the data stream in the transmitter is divided into many parallel data streams, and it is restored again in the receiver, which increases the effective throughput by 4 times. This would allow the transfer of data at speeds up to 100 Mbit / s. Additionally, in the case of 4 pairs of copper wires, the same wires of four pairs are used in order to simultaneously transmit the same amounts of data in the opposite direction. Two-way data transmission on each copper wire of a pair doubles the information capacity that can be transmitted. In this case, the data transfer rate increases by eight times compared with conventional transmissions, in which two pairs are used respectively for only one direction. For data transmission, as described above, crosstalk noise is a very limiting factor. The types of crosstalk (Xtalktype) are near-end crosstalk (NEXT) 51, which refers to the unintentional input of the 50 signals of the transmitter 10 at one end to the 50 signals at the receiver 10 at the same end, and the far-end crosstalk (FEXT) 52 , which relates to the unintentional input of signals 50 during transmission to the receiver 11 at the other end, the signals 50 during transmission being input to the signals 50 of neighboring copper pairs and appear as noise in the receiver 11 (see Fig. 1). It is generally assumed that the NEXT 51 interference has only a near-end interference source. The Xtalktype parameter, thus, depends on the location and flow (up / down), that is, this dependence can be written as Xtalktype (flow, location). If there are more than two copper conductors, as is usually the case (in a typical case, there are from 2 to 2400 wires), then the above pairwise connection is no longer valid. For example, for the case when the conductors of four pairs are used simultaneously, then, therefore, there are three unintentional sources of interference that act on signal 50 with their energy. For A in this case, the relation A = 3 is valid. The same is true for crosstalk type FEXT 52.

Computing unit 30 determines the reserves of data transmission resources by means of a Gaussian transform module 31 based on the effective signal levels S (f) from the first list and the corresponding noise levels N (f) from the second list for various data modulations and / or modulating codes for a predetermined speed transmission of bits and saves the reserves of data transmission resources, correlated with the corresponding physical lengths 13 and the thickness of the cores of the cable 141, 142, 143, 144 of the network connection 12, on the data carrier in the computing unit oce 30. Based on the effective signal levels S (f) from the first list and the corresponding noise levels N (f) from the second list, it is possible using computing unit 30 to determine the ratio of signal S to noise N (SNR) in the following form:

This expression is valid only for modulations of the form CAP, 2B1Q and PAM, but is not applicable for modulation of the form DMT. Modulation of the DMT type is described in more detail below. T in this case denotes the symbol interval or half the inverse of the Nyquist frequency. The Nyquist frequency is the maximum frequency with which a sample can still be accurately taken. The Nyquist frequency is half the value of the sampling frequency, since unintended frequencies are generated when a signal is sampled whose frequency is higher than half the value of the sampling frequency. Index n is the summation index. In practice, it is usually sufficient when n takes values from -1 to +1. If this is not enough, then additional maximums of 0, ± 1 / T, ± 2 / T, etc. can be used. until the required accuracy is achieved. Resource data transfer resources depend on the modulation of the data transfer and / or modulating codes, as noted above. This embodiment shows the dependence for 2B1Q modulation used for HDSL modems and CAP modulation as an example for ADSL modulation of the DMT type, as well as for modulating codes using trellis coding signals. However, it is understood that the method and system of the invention can also be used for other modulations of data transmission and / or modulating codes, such as PAM (pulse amplitude modulation), etc. Both 2B1Q type modulation and CAP type modulation are used for HDSL modems and are characterized by a predetermined bit rate. Modulation of the form DMT is used for ADSL modems and, in contrast, has a variable bit rate. The types of modulation of ATS and DMT use the same underlying modulation technology: quadrature amplitude modulation (QAM), although this technology is used in different ways. QAM provides the possibility that two digital carrier signals have the same transmission bandwidth. In this case, two independent, so-called message signals are used to modulate two carrier signals that have the same carrier frequency, but differ in amplitude and phase. QAM signal receivers can distinguish whether a low or a high number of amplitude and phase states is required to overcome the effects of noise and interference, for example, in a single pair of copper conductors. 2B1Q modulation is also known as 4-level amplitude modulation (PAM). It uses two voltage levels for pulse signals, and not one level, as, for example, in the case of modulation type AMI (coding with alternating polarity of the elements). Since positive and negative differences in levels are applied for this, a 4-level signal is obtained. The bits are then evaluated in pairs, which pair corresponds to a certain voltage level (hence the name 2-bit modulation). Due to this, it is possible to halve the necessary transmission frequency for transmitting at the same bit rate as in the case of bipolar modulation of the AMI type. For an HDSL modem using modulation of the form 2B1Q or CAP, the following dependence of the reserve of data transfer resources on the SNR exists:

where the parameter ξ can be determined depending on the error rate (symbol error rate) ε _s . For local area networks of the Internet protocol, usually a sufficient error rate ε _s = 10 ^-7 , that is, every 10 ⁷ bits on average is transmitted with distortion. Firms typically require ε _s = 10 ^-12 for their branded networks. If the value of ε _s becomes of the order of magnitude of the transmitted data packet (for example, 10 ^-3 ), then this, on the contrary, would mean that each packet should be transmitted on average twice, until it is received correctly. For modulation of the form 2B1Q for the parameter ε _s , for example, the following is true:

для некодированных сигналов

for non-encoded signals

для сигналов решетчатого кодирования

for trellis coding signals

while for the modulation of the type of ATS the following relations are true:

для некодированных сигналов

for non-encoded signals

для сигналов решетчатого кодирования

for trellis coding signals

The parameter G _c for both types of coding is a complementary Gaussian function:

M denotes a moment number for modulation of the form 2B1Q, and M = 4 for 2B1Q, while for modulation of the CAP type, the grouping parameter is MxM. T denotes, as indicated above, the symbol spacing or half the inverse of the Nyquist frequency. For ADSL modems using modulation of the form DMT, the dependence is different. As noted above, ADSL has a variable bit rate. This is also manifested in the determination of the parameter M _s . In this case, the following relation holds:

ξ (f) denotes the signal-to-noise ratio S (f) / N (f). x _ref denotes the reference supply of resources, which in this embodiment was typically chosen to be 6 dB, that is, x _ref = 10 ^0.6 . However, for x _ref , other values can also be selected as reference resource reserves. Δf - the entire frequency band or the frequency band that is used for transmission. Integration is performed by frequency. D is the bit rate, for example, in bits per second (bps). G - correction factor. In this embodiment, G has a value, for example, of 9.55. The integration in this example is performed over the frequency f. Similarly, it can, however, be carried out in time or other physical quantity, moreover, the above expression should be accordingly consistent.

In the general case, the reserves of data transmission resources obtained above do not coincide with the experiment. Therefore, the computing unit 30 determines the effective reserves of data transmission resources by means of at least one correction coefficient based on the stored reserves of data transmission resources. A correction factor was chosen for this embodiment in such a way that sufficient consistency is achieved between the received data transfer resource reserves and the effective data transfer resource reserves. As a sufficient value in this case, it was accepted, for example, ± 3 dB, and other values can be used. To obtain this maximum deviation of ± 3 dB, two parameters are determined. M _imp takes into account the good or poor implementation of the modem by the manufacturer. The parameter M _imp was introduced on the basis of the fact that the same modems with comparable hardware and the same modulation of data transfer and / or modulating codes, however, issued by different manufacturers, when converting an analog signal to a digital signal and vice versa, gave different results, which influenced their maximum data transfer rate or their maximum range for a particular network connection. This should be adjusted in relation to the stock of data transfer resources. The parameter N _int was introduced as the second parameter. The parameter N _int takes into account the quantization noise in the modem (analog-to-digital conversion), as well as the possible poor adjustment of the correction block during transmission. If there is a transfer between the transmitter 10 and receiver 11, then the correction unit in the modem will coordinate the data transfer rate with the conditions of the network connection, for example, attenuation in the line, phase distortion, etc. through a test sequence that is sent between both leading modem information exchanges in both directions. Poor coordination due to the correction block leads to distortion of the results and should be adjusted. For a linear correction block, for example, the following expression can be used:

Where

The SNR _LinearEq denotes the signal-to-noise ratio, S _e is the signal received by the correction unit, N _e is the noise and f is the frequency. For a correction block with adaptive decision feedback (DFE), for example, the following expression can be used:

Where

In this case, again, SNR _DFE denotes the signal-to-noise ratio, S _e - as above, the signal received by the correction unit, N _e - noise and f - frequency. The computing unit 30 for determining the SNR _DFE may use, for example, the following approximation:

Thus, for effective reserves of data transfer resources, we obtain: S (f) = PSD _Modem (f) H ² (f, L, D), as before. Noises are adjusted as follows:

Correction can be implemented in the computing unit 30 by hardware or software in a single module. It should be noted that with the help of such a module, based on the correction N _{int, a} variable noise factor is introduced, which, for example, can take into account the setting of the correction block, etc. This solution is unknown from the prior art and belongs, among other things, to the essential advantages of the invention. The effective data transfer resource reserves M _eff are taken into account by the relation M _eff = M _c -M _imp , which is taken into account in addition to N _int , as mentioned above. Correct values for M _c and N _int can be obtained by computing unit 30 in comparison with experimental data. In a typical case, the computing unit 30 should then have access to the data of various experiments in order to be able to correctly determine the parameters within the desired deviations. By means of correction factors, which, therefore, include the average deviation of the stored reserves of data transmission resources with respect to the effective reserves of data transmission resources, the effective reserves of data transmission resources are determined, as well as in comparison with the corresponding physical lengths L 13 and thicknesses D cores of cables 141, 142, 143, 144 of the network connection are stored on the storage medium of the computing unit 30. It should be noted that the correction factors are not necessarily linear coefficients, that is, they must be constant, but with the same success, they can include correction functions with a nonlinear dependence. Thus, depending on the application, more complex deviations of experimental data can be taken into account. Using the stored data arrays with data transfer resource reserves, the computing unit 30 determines, based on the stored effective data transfer resource reserves, using the known physical length 13 of the network connection 12 between the transmitter 10 and the receiver 11, the data transfer resource reserve for a particular network connection 12. Data transfer resource reserves are indicated, as was repeatedly mentioned above, in decibels. For values> 0, the modem operates in a typical way, while for values <0 it does not work. To guarantee reliable quality operation, it may be appropriate to choose, for example, 6 dB as the lower limit. However, in general, other values of data transfer resource reserves for the lower boundary are suitable for use, for example, values ranging from 3 dB to 9 dB. Due to a similar configuration for ADSL modems, it is also possible, as follows from the above data, instead of data arrays with data transfer resource reserves, data arrays with bit rates for various network connections, for example, for data transfer resource reserves of 6 dB, can be respectively determined. Thus, to determine data arrays with bit rates of 6 dB = M _eff . For HDSL modems, this does not make sense in this regard, since in the case of HDSL encodings are used, for example, of the form 2B1Q or CAP with a constant data rate, in this case 2.048 Mbps. The reason for this difference with respect to ADSL modems lies in the fact that HDSL systems were designed only for connection with a higher bit rate, and only transmission reliability (SNR ratio) is of interest to them. In FIG. Figure 3 shows the transmission section of a network connection depending on the bit rate for ADSL modems. Reference numerals 60 and 61 denote various noise conditions. Bit rates, as described above, are presented based on stored data arrays or lists 2030.

Based on the stored data arrays or lists of 2030 data resource reserves / bit rates, (1030) the data resource reserves / bit rates for the classified network connection are determined and transferred to the storage medium of the computing unit 30 in accordance with the classified network connection 12.

Based on the effective network connection length, the attenuation distribution coefficient 2020, and the data transfer resource reserves 2030 using the computing unit 30, a classified connection (1040) can be classified according to its maximum data throughput. The classification may, in particular, include the highest possible data rate for the classified network connection. Classification results can be provided (1050) to the user through a screen, print module, or other output device. In particular, it is possible, for example, to connect to the Internet via a graphical interface via a device, and it is easy for any telephone subscriber of a telephone service provider to determine whether his connection (for example, in his living room) is suitable for a particular network connection.

Claims (28)

1. A method for classifying network connections, the geographical coordinates of the transmitter (10) and receiver (11) of the classified network connection (12) being known, characterized in that based on the known data (5000) of the network connection, one or more distance coefficients (2010) are determined by the computing unit (30) and the obtained data, correlated with the determined probability, is transferred to the data carrier of the computing unit (30), and the distance coefficients (2010) indicate the effective length of the network connection depending on the distance through the air, and the determined probability of whether a certain length of the network connection is shorter or longer than its effective network length, is determined t through the coefficient (2012) of reliability, based on one or more coefficients (2010) of distances, reliability coefficient (2012) and geographical coordinates of the transmitter (10) and receiver (11) of the classified network connection (12) using the computing unit (30) determine (1010) the effective length of the network connection and transferred to the storage medium of the computing unit (30) in accordance with the classified network connection (12), based on the known data (5000) about network connections, (3020) determine at least one attenuation distribution coefficient (2020) and transfer to the storage medium of the computing unit (30), at least one distribution coefficient (2020) attenuation indicates the relationship to each other of attenuation of various sections of the network connection, determine (3030) the reserve of resources (2030) of data transfer to determine the maximum data transfer throughputs for various types of modems and store on the data carrier of the computing unit (30) in accordance with the physical length (13) and cable thickness (141, 142, 143, 144) of the network connection (12), moreover, by means of a power measuring device (20), energy spectra for modem types are measured, and effective signal levels and corresponding levels of wavelengths are determined on the basis of energy spectra using a computing unit (30) Using the Gaussian transform module (31), on the basis of signal levels and noise levels for various modulations of data transmission and / or modulating codes, the reserve of transmission resources (2030) is determined for a predetermined bit rate and On the basis of the effective length of the network connection, the coefficient (2020) of the distribution of attenuation and reserves of resources (2030) for transmitting data, the computing unit (30) classifies (1040) the classified network connection in accordance with its maximum data transmission capacity. 2. The method according to claim 1, characterized in that as the coefficients (2011) of the distance using the computing unit (30) determine the angular coefficient and the abscissa, while determining a linear relationship between the distance through the air and the effective length of the network connection. 3. The method according to claim 1, characterized in that the distance coefficients (2011) are determined by means of a computing unit as parameters of a polynomial of at least degree 2. 4. The method according to claim 1, characterized in that by a coefficient (2012) of reliability, a probability is selected in the range from 0.85 to 0.95. 5. The method according to claim 1, characterized in that the reliability coefficient (2012) has a value in the range from 700 to 800. 6. The method according to claim 1, characterized in that by means of a coefficient (2020) of the distribution of attenuation, a linear relationship of the attenuation relative to each other is determined. 7. The method according to any one of claims 1 to 6, characterized in that the computing unit (30) determines the adjusted reserves of data transfer resources by means of at least one correction factor based on the stored reserves of data transfer resources (2030) and stores them correlated with the corresponding physical lengths (13) and thicknesses of the cable cores (141, 142, 153, 144) of the network connection (12) on the data carrier of the computing unit (30), and at least one correction factor includes the average deviation of the stored data apasses of data transmission resources relative to the effective reserves of data transmission resources and / or the coefficient of the correction block for setting the correction block. 8. The method according to claim 7, characterized in that the correction coefficient reproduces a non-linear dependence on the physical lengths (13) and / or thicknesses of the cable strands (141, 142, 143, 144). 9. The method according to claim 1, characterized in that the noise levels are determined by the computing unit (30) depending at least on the crosstalk parameters and the number of interference sources based on energy spectra. 10. The method according to claim 1 or 9, characterized in that the energy spectrum is measured depending on the transmission frequency for the types of modems (101, 102, 103, 104) ADSL, and / or SDSL, and / or HDSL, and / or VDSL . 11. The method according to claim 10, characterized in that the possible types of modems (101, 102, 103, 104) SDSL include at least the type of modem G.991.2 and / or types of modems (101, 102, 103, 104) ADSLs include at least the type of G.992.2 modem. 12. The method according to claim 1, characterized in that through the module (31) of the Gaussian transform determine the reserves of data transmission resources, at least for modulations of data transmission of the form 2B 1Q (2 binary, 1 quadruple), and / or CAP (amplitude / phase modulation without a carrier), and / or DMT (digital multi-tone), and / or PAM (pulse amplitude modulation). 13. The method according to any one of claims 1 to 6, 11 and 12, characterized in that through the module (31) of the Gaussian transform determine the reserves of data transmission resources, at least for coding using modulation trellis code. 14. A method for classifying network connections, wherein the geographical coordinates of the transmitter (10) and receiver (11) of the classified network connection (12) are known, characterized in that on the basis of known data (5000) on network connections, one or more distance coefficients (2011) are determined by the computing unit (30) and, in accordance with the determined probability, transferred to the data carrier of the computing unit (30), the distance coefficients (2011) indicating the effective length network connection depending on the distance through the air, and the determined probability of whether a certain network connection length is shorter or greater than its effective network length is determined by factor (2012) of reliability, based on the coefficients (2010) of distances, the coefficient (2012) of reliability and the geographical coordinates of the transmitter (10) and the receiver (11) of the classified network connection (12) using the computing unit (30) determine (1010) the effective length of the network connection and transfer it to the medium data of the computing unit in accordance with the classified network connection (12), based on the known data (5000) about network connections, (3020) determine at least one attenuation distribution coefficient (2020) and transfer to the storage medium of the computing unit (30), at least one distribution coefficient (2020) attenuation indicates the relationship to each other of attenuation of various sections of the network connection, determine the bit rate (2030) bits to determine the maximum data throughput for various types of modems and store on the data carrier of the computing unit (30) in accordance with the physical length (13) and cable thickness (141, 142, 143, 144) of the network connection (12), moreover, by means of a power measuring device (20), energy spectra are measured for types of modems, effective levels of signals and corresponding noise levels are determined on the basis of energy spectra using a computing unit (30) schyu module (31) based on the Gaussian transformation signal strength and noise level for different data transmission modulations and / or modulating codes determined transmission rate (2030) of bits for a predetermined reserve data transmission resources, and On the basis of the effective length of the network connection, the coefficient (2020) of the distribution of attenuation and reserves of resources (2030) for transmitting data, the computing unit (30) classifies (1040) the classified network connection in accordance with its maximum data transmission capacity. 15. The method according to 14, characterized in that as the coefficients (2011) of the distance using the computing unit (30) determine the angular coefficient and the abscissa, while determining a linear relationship between the distance through the air and the effective length of the network connection. 16. The method according to 14, characterized in that through the computing unit (30) determine the coefficients (2011) of the distance as the parameters of the polynomial of at least the 2nd degree. 17. The method according to 14, characterized in that through the coefficient (2012) of reliability choose the probability in the range from 0.85 to 0.95. 18. The method according to 14, characterized in that the coefficient (2012) of reliability has a value in the range from 700 to 800. 19. The method according to 14, characterized in that through the module (31) Gaussian transform determine the bit rate for the reserves of data transfer resources between 3 and 9 dB. 20. The method according to 14, characterized in that through the module (31) Gaussian transform determine the bit rate for the reserve of data transfer resources of 6 dB. 21. The method according to any one of paragraphs.14-20, characterized in that the computing unit (30) determines the corrected bit rates using at least one correction coefficient based on the stored bit rates (2030) and saves them in accordance with physical lengths (13) and thicknesses of cable cores (141, 142, 143, 144) of the network connection (12) on the data carrier of the computing unit (30), and the correction coefficient includes the average deviation of the stored bit rates relative to the effective rates bit transfer ratios and / or correction block coefficient for adjusting the correction block. 22. The method according to item 21, wherein the at least one correction factor reproduces a non-linear dependence on the physical lengths (13) and / or thicknesses of the cable cores (141, 142, 143, 144). 23. The method according to 14, characterized in that the noise levels are determined by the computing unit (30) depending, at least, on the crosstalk parameters and the number of interference sources based on energy spectra. 24. The method according to 14 or 23, characterized in that the energy spectrum is measured depending on the transmission frequency for the types of modems (101, 102, 103, 104) ADSL, and / or SDSL, and / or HDSL, and / or VDSL . 25. The method according to paragraph 24, wherein the possible types of modems (101, 102, 103, 104) SDSL include at least the type of modem G.991.2 and / or types of modems (101, 102, 103, 104) ADSLs include at least the type of G.992.2 modem. 26. The method according to 14, characterized in that through the module (31) Gaussian transform determine the bit rate, at least for modulations of data transmission type 2B 1Q, and / or CAP, and / or DMT, and / or RAM . 27. The method according to any one of claims 14 to 20, 25 and 26, characterized in that by means of the Gaussian transform module (31) bit rates are determined, at least for coding using trellis code modulation. 28. Device for classifying network connections, and the geographical coordinates of the transmitter (10) and receiver (11) of the classified network connection (12) are known, characterized in that the device comprises a computing unit (30) for determining and storing one or more distance coefficients (2011) in accordance with a determined probability based on known data (5000) on network connections, and distance coefficients (2011) indicating the effective length of the network connection depending on the distance by air, and the determined probability of whether a certain length of a network connection is shorter or longer than its effective network length is determined by the reliability coefficient (2012), computing unit (30) comprises means for determining and storing at least one attenuation distribution coefficient (2020) based on known data (5000), wherein at least one attenuation distribution coefficient (2020) indicates a relation to friend weakening of various sections of the network connection, the device also includes a power measurement device (20) for measuring energy spectra for various types of modems, a means (30) for determining effective signal levels and corresponding noise levels based on energy spectra, and a Gaussian transform module (31) for determining and maintaining resource reserves (2030) transmissions based on signal levels, noise levels for various modulations of data transmission and / or modulating codes and for a predetermined bit rate.

Images