Device for parametric estimation of message streams distribution law

FIELD: radio engineering, computer science.

SUBSTANCE: device for parametric estimation of message streams distribution law contains: input amplifier, parameters computation block, average arithmetic values computation block, block for determining distribution type, distribution computing device, control block, intensiveness analysis block, block for analyzing trustworthiness of messages. Thus, it is possible to estimate parameters of distribution of homogeneous and heterogeneous streams of multi-packet messages with high trustworthiness, specific signs of pauses for which may be identified both quantitatively and qualitatively - untrustworthily, due to comparative analysis in block of bit series received in binary code from position of their match to signs of pauses and mathematically correct transformation, on basis of neuron networks theory, of untrustworthily identified signs of pauses to form, useable for performing procedure of parametric estimation of message streams distribution law.

EFFECT: increased trustworthiness if estimating under conditions, appropriate for real functioning process of multi-channel radio-communication networks and local area networks under conditions of untrustworthy identification of signs of pauses between multi-packet messages, circulating in given networks.

4 cl, 8 dwg

The invention relates to radio engineering and computer science and is intended for parametric estimation of the law of distribution of a multipackage of messages in networks multi-radio (SMCRS) and local area networks (LAN).

A device for parametric estimation of the distribution law on author. St. USSR №1024935, G 06 F 15/36, G 06 G 7/52, 1983, bull. 23 that contains an input amplifier, a memory unit, the unit inspection agreement, the control unit and the synthesizer distribution function.

The disadvantage of this device is the relatively low rate of flow analysis in the evaluation of the distribution of a multipackage messages (IPU) in SMCRS and LAN, due to the need enumeration of the set M combinations of parameters for the distribution of message flow consisting of a set of packages.

A device for parametric estimation of the distribution of message flows that contain the input amplifier, the power calculation unit calculating average values, the block type definition of the distribution, the evaluator distribution and control unit (see RF patent №2094844, G 06 F 17/18, 1997, bull. 30).

However, this device has a narrow scope, because it is impossible to estimate the distribution of the heterogeneous (mixed) flow MEAs, have their place in SMCRS and LAN and due to the presence of sources of information, outstanding as expressed independent single informational messages to the user (COI), and packs of COI, the length of which is distributed according to a geometric law. In this case, the approximation of the real statistics of incoming flows COI impossible to reduce them to the three classic versions message flows.

The closest to the technical nature of the claimed device (prototype) is a device for parametric estimation of the distribution of message flows (see RF patent №2165100, 7 G 06 F 17/18, 2001, bull. 10)containing the input amplifier, the input of which is filed binary pulse sequence, the block parameter calculation unit calculating average values, the block type definition of the distribution, computer distribution, control unit, and the unit of analysis of the intensity, and the clock output control unit connected to the clock input of the transmitter distribution, Abdoulaye and control outputs of the control unit are connected respectively with boleushim and managing inputs of the computing unit parameters, an information input connected to the output of the input amplifier, the first and second information outputs of the computing unit parameters connected respectively to the first and second information input unit for computing the average value is, the first information output of which is connected to the first information input of the transmitter distribution and with the input of the block type definition of the distribution, the first, second, and third modification, the outputs of which are connected respectively to the first, second and third modification to the inputs of the transmitter distribution, the second information input of which is connected with the second information output unit for computing the average values, the output of the transmitter distribution is the output device, the input unit of analysis intensity is connected to the second information output unit for computing the parameters of the first and second parametric outputs of the unit of analysis intensity are connected respectively with the first and second parametric inputs evaluator distribution, control output unit rate analysis connected to the control input of the computing unit arithmetic mean values.

In the prototype implements the estimation of distribution parameters (intensity - λ_cand the ratio of the information length of the packets and messages ξ) homogeneous or heterogeneous (mixed) flow MEAs - streams, periodically changing its intensity during operation SMCRS and LAN, thus the extension of the application device is A.

However, the prototype has the disadvantage of relatively low reliability assessment in terms of unreliability (failure, incompleteness and inconsistency) characteristics (parameters) of the circulating flow of a multipackage messages, i.e. relatively low reliability evaluation thread state IPU SMCRS and LAN unreliable (insufficient, incomplete) identified by signs. This device allows high accuracy to estimate the parameters of the distribution of only those homogeneous and heterogeneous (mixed) message flows, source characteristics (parameters) of the pauses which quantitatively defined, consistent, and fully identified, while the signs of the beginning and end delays for a large number of high-speed flows of messages that are actually circulating in the computer and other telecommunication networks can be identified only qualitatively (at a qualitative level - of inaccurate, insufficient, incomplete), based on the opinions of a large number of experts.

Under "features" breaks the flow of the message refers to time intervals between messages - the time intervals between the last bit of the end flag information of the previous packet (the last one in the previous message) and the first bit of the flag information beginning of the next pack is a (first in the next message), and when the identification bit of the end flag information of the previous packet (the last one in the previous message) and the time of identification bits flag the start information of the next packet (the first in the next message).

The aim of the invention is a device for the parametric estimation of the distribution of message flows, providing increased reliability assessment in conditions inherent in the real process of functioning SMCRS and LAN, i.e. unreliability (failure, incompleteness and inconsistency) identification signs pauses circulating flows multipackage messages, devices with high reliability to estimate the parameters of the distribution of homogeneous and heterogeneous (mixed) flow map, identification signs pauses (signs of the beginning and end of the pause between messages) can be identified both quantitatively and qualitatively inaccurate, incomplete.

This goal is achieved by the fact that in the known device for the parametric estimation of the distribution of message flows that contain the input amplifier, the input of which is the input of the block parameter calculation unit calculating average values, the block type definition of the distribution, the evaluator distribution of the population, the unit of analysis of the intensity and the control unit, clock output of which is connected to the clock input of the transmitter distribution, control output control unit connected with the control input of the computing unit parameters, the information output of which is connected to the first information input unit for computing the average value, the first information output of which is connected to the first information input of the transmitter distribution and with the input of the block type definition of the distribution, the first, second, and third modification, the outputs of which are connected respectively to the first, second and third modification to the inputs of the transmitter distribution, the second information input of which is connected with the second information output unit for computing the average values, the output of the transmitter distribution is the output of the device, the first and second parametric outputs of the unit of analysis intensity are connected respectively with the first and second parametric inputs evaluator distribution, control output unit of analysis intensity is connected to the control input of the unit for computing the average values have been added to the unit of analysis the reliability of the message, and the output of the input amplifier is connected with an information input unit for computing a pair of the EAN and the information input unit of analysis the reliability of the message, the information output of which is connected to the second information input unit for computing the average values and the input unit of analysis intensity, Abdoulaye output control unit connected to boleushim input unit of analysis the reliability of message verification the output of which is connected to a verification to the input of the computing unit parameters.

Unit of analysis the reliability of the message consists of a controller pauses, controller messages, and the message counter, the output of which is an information output unit, the input of the controller pauses is an information input unit, the output of the controller pauses connected to the input of controller messages, the output of which is the verification unit output and is connected to the direct input of the counter message, Abdoulaye entrance which is boleushim input unit of analysis the reliability of the message.

Block parameter calculation consists of a decoder packet headers, the counter information of the length of a multipackage messages and computer ratio information of the lengths of the packets and messages, the output of which is an information output unit, the input of the decoder packet headers is an information input unit and connected to the first input counter information length of a multipackage message, the output of the decoder headers of paketo is connected with the second input counter information length of a multipackage messages verification input of which is the verification of the input block, the output of the counter information of the length of a multipackage messages connected to the direct input of the transmitter correlation information of the lengths of the packets and messages, the control input of which is a control input of the computing unit parameters.

Thanks to the new essential features due to the introduction of unit of analysis the reliability of the message, providing a comparative analysis of the incoming binary bit sequence from the point of view of their conformity to the characteristics of pauses and conversion unreliable (incomplete) identified characteristics of pauses, in the inventive device allows a preliminary analysis and transformation characteristics of pauses identified unreliable (incomplete) to a form suitable for parametric estimation of the distribution of message flows, which causes an increase in the reliability assessment in conditions inherent in the real process of functioning SMCRS and LAN - in terms of unreliability (failure, incomplete, and often contradictory) signs pauses circulating flows multipackage messages.

The analysis of the level of technology has allowed to establish that the analogues, characterized by a set of attribute the Cove, identical to all features of the claimed technical solution is available, which indicates compliance of the device to the condition of patentability "novelty".

Search results known solutions in this and related areas of technology in order to identify characteristics that match the distinctive features of the prototype of the features of the declared object, showed that they do not follow explicitly from the prior art. The prior art also revealed no known effect provided the essential features of the claimed invention transformations on the achievement of the technical result. Therefore, the claimed invention meets the condition of patentability "inventive step".

The claimed device is illustrated by drawings, on which:

figure 1 - structural diagram of the device for the parametric estimation of the distribution of message flows;

figure 2 - block diagram of the unit of analysis, the validity of the message;

figure 3 - block diagram of the controller messages.

figure 4 is a structural diagram of the neural network evaluator;

figure 5 - block diagram of the computing unit parameters;

figure 6 is a structural block circuit diagram of calculation of arithmetic mean values;

figure 7 - block diagram of the control unit;

on Fig - structural is a block circuit diagram of the analysis of intensity.

Device for the parametric estimation of the distribution of the message flows shown in figure 1, consists of an input amplifier 1, the inlet of which is the input device, the computing unit 2 parameters, the unit for computing the average values of 3, block type definition of the distribution 4, evaluator distribution 5, a control unit 6, unit of analysis intensity 7 and unit of analysis the reliability of message 8. Clock output 61 of the control unit 6 is connected to the clock input 56 of the transmitter distribution 5, control output 63 of the control unit 6 is connected with the control input 23 of the computing unit 2 parameters, the information output 24 which is connected to the first information input block 31 calculation of arithmetic mean values of 3, the first information output 33 which is connected with the first information input 51 evaluator distribution 5 and to the input of the block type definition of the distribution 4, the first 41, second 42 and 43 third modification, the outputs of which are connected respectively to the first 53 and second 54 and 55 third modification to the inputs of the transmitter distribution 5, second information input 52 which is connected with the second information output block 34 calculation of arithmetic mean values of 3. The output 59 of the evaluator distribution 5 is an output device, the first 73 and second 74 couples the metric outputs block rate analysis 7 are connected respectively with the first 57 and second 58 parametric inputs evaluator distribution 5, control output 72 of the unit of analysis intensity 7 is connected to a control input 35 of the block of calculation of arithmetic mean values of 3. The output of the input amplifier 1 is connected with the information input 21 of the computing unit 2 parameters and information input 81 of unit of analysis the reliability of message 8, the information output 83 which is connected to the second information input 32 of the block of calculation of arithmetic mean values of 3 and the input 71 of the unit of analysis of intensity 7, Abdoulaye output 62 of the control unit 6 is connected to boleushim entrance 84 unit of analysis the reliability of message 8, the verification output 82 which is connected to a verification to the entrance 22 of the computing unit 2 parameters.

Unit of analysis the reliability of communication 8 (2) is intended to capture (registration) pauses, comparative analysis of the characteristics of pauses from the point of view of their full compliance to the feature space pauses, conversion unreliable (incomplete) identified characteristics of pauses and determine the current true values of the parameter characterizing the intensity of the flow of messages - λ.

Unit of analysis the reliability of message 8 consists of a controller pauses 8.01, controller messages 8.02 and the message counter 8.03. Output 833 counter messages 8.03 information is output 83 unit 8, entry 811 controller pauses 8.01 detail is rationem entrance 81 unit 8, the output 812 of the controller pauses 8.01 connected to the input of controller 821 messages 8.02, the output 822 of which is the verification of the output 82 of the block 8 and is connected with direct entry 831 counter messages 8.03, Abdoulaye input 832 of which is boleushim entrance 84 unit of analysis the reliability of message 8.

The controller pauses 8.01 unit of analysis the reliability of message 8 is designed to capture (registration) pauses, characterizing the end of the message. The controller pauses 8.01 may be technically implemented as a standard mass storage device on the basis of the shift register with serial input and output of information, described in [Bystrov Y.A., Welikson AM, Wagman E, and other electronics: Reference book / Ed. by Bystrov Y.A - SPb.: Energoatomizdat, 1996. S-292, RIS].

The message controller 8.02 unit of analysis the reliability of communication 8 (3) is designed for comparative analysis of the characteristics of pauses from the point of view of their full compliance to the feature space pause and replay unreliable (incomplete) identified characteristics pauses to a form suitable for accurate parametric estimation of the distribution of message flows. The message controller 8.02 consists of a shifting element 8.02-1 and neural network evaluator 8.02-2. Entrance 8021 shifting element 8.02-1 is the entrance 821 to the of ntroller messages 8.02, reliable output 8022 shifting element 8.02-1 is connected to prohibit entrance 8024 neural network evaluator 8.02-2, combined with the output 8027 neural network evaluator 8.02-2 and is output 822 message controller 8.02, inaccurate output 8023 shifting element 8.02-1 is connected to allows input 8025 and information input 8026 neural network evaluator 8.02-2.

Shift element 8.02-1 controller messages 8.02 designed to perform serial comparison (number of digits) indication of pauses and decision-making about their mathematical nature - signs pauses identified quantitatively (parametrically) or qualitatively (inaccurate, incomplete and in need of verification using analytically formulated subjective opinions (associations) of experts about specific values of time intervals (pauses) between messages. Shift element 8.02-1 can be technically implemented in the form of commercially available shift register to shift to the left as shown in the literature [Sidorov, A. M., Gusev V.V., Lebedev O. Fundamentals of pulse and digital techniques. - SPb.: SUVIUS, 1995. S-160, RIS(b)].

Neural network evaluator 8.02-2 controller messages 8.02 (figure 4) is intended for the procedure of conversion characteristics pauses identified unreliable (n is full), to a form suitable for parametric estimation. Neural network evaluator 8.02-2 is a programmable parallel four-digit arithmetical-logical element (ALE) with one permit and one prohibiting entrance, n informational inputs and one combined logical output, where n can take values from 2 to 23. Enable input M ALE is permissive input 8025 neural network evaluator 8.02-2, prohibiting entrance WITH ALE is prohibiting entrance 8024 neural network evaluator 8.02-2, n inputs ALE and United are information input 8026 neural network evaluator 8.02-2, combined with logical output F ALE is output 8027 neural network evaluator 8.02-2. The message controller 8.02 may be technically implemented as a parallel four-digit arithmetical-logical unit on the basis of commercially available programmable CMOS chip (a chip on the complementary field-effect transistors with the structure of a metal-oxide-semiconductor) series 564 (for example, CIS), as shown in [awl V.L. Popular digital circuits. The Handbook. - M.: Radio and communication, 1987. S-275, Fig. 2.70].

Message counter 8.03 unit of analysis the reliability of message 8 is designed to determine reliable values of intensity λ pulse output is of controller messages per time interval, determined by the control unit. Message counter 8.03 can be technically implemented on the basis of commercially available counter, as shown in [H. Sobotka, Old Ya Microprocessor systems. - M.: Energoizdat, 1981. Pp.96-100].

The computing unit 2 parameters, depicted in figure 5, is designed for analyzing packet flow definition information of the message length and calculate the current value of parameter ξcharacterizing the ratio of the information length of the packets and messages. The computing unit settings 2 consists of a decoder packet headers 2.01, counter information length of a multipackage messages 2.02 and computer ratio information of the lengths of the packets and messages 2.03. The output of the transmitter correlation information of the lengths of the packets and messages 2.03 is an information output 24 of block 2, entrance decoder packet headers 2.01 is an information input 21 of the block 2 and is connected to the first input 221 of the counter information of the length of a multipackage messages 2.02, the output of decoder packet headers 2.01 connected with the second input 222 of the counter information of the length of a multipackage messages 2.02, verification input 223 of which is the verification of the entrance 22 of the block 2, the output of the counter information of the length of a multipackage messages 2.02 connected to the direct input 231 of the transmitter correlation information on the in packets and messages 2.03, control input 232 which is a control input 23 of the computing unit 2 parameters.

The decoder packet headers 2.01 block parameter calculation 2 is intended for the analysis arriving at its input binary pulse sequence and detecting it identified signs of the beginning and end of the data portion of the packet. The decoder packet headers 2.01 can be technically implemented on the basis of commercially available decoder described in the book [Bogdanovich M.I., Grell I.N., Prokhorenko, VA and other Digital integrated circuits: a Handbook. - Minsk: Belarus, 1991. S-436, RIS].

The counter information of the length of a multipackage messages 2.02 block parameter calculation 2 is designed to determine the current values of the parameter m_withinformational message length. The counter information of the length of a multipackage messages 2.02 can be technically implemented as a binary sum counter with serial transfer on T-triggers, as described in the literature [Ugryumov, H.E. Design elements and components of the computer. - M.: Higher school, 1987. S.138-141, RES.].

The evaluator correlation information of the lengths of the packets and messages 2.03 block parameter calculation 2 is designed to determine the current value of parameter ξcharacterizing the ratio of the information length of the pack is impressive and messages. The evaluator correlation information of the lengths of the packets and messages 2.03 can be technically implemented on the basis of commercially available counter-divider, as shown in [Gusev V.V., Lebedev O., Sidorov A. M. fundamentals of pulse and digital techniques. - SPb.: SUVIUS, 1996. S-182, RIS(b)].

The input amplifier 1, which is the common structural scheme, designed to perform amplification, normalization amplitude and duration of the input stream - the binary pulse sequence. The input amplifier 1 can be technically implemented as a Schmitt trigger, as shown in [Gusev V.V., Lebedev O., Sidorov A. M. fundamentals of pulse and digital techniques. - SPb.: SUVIUS, 1996. S-119, ri(a)].

The power calculation of arithmetic mean values of 3, part of the General structural scheme, designed to obtain numerical values averaged over the observation interval of flow parameters (and). Its structural scheme is known as described, for example, in the prototype (see RF patent №2165100, 7 G 06 F 17/18, 2001, bull. 10) and is represented in figure 6 of this specification. The power calculation of arithmetic mean values of 3 consists of counter-divider values duration 3.01, counter-divider intensity values 3.02 and item BAN 3.03. The entrance of the counter-divider values dlitelnost is 3.01 is the first information input block 31 calculation of arithmetic mean values of 3, the output of the counter-divider values duration 3.01 is the first information output 33 of the block 3. The entrance of the counter-divider intensity values 3.02 is the second information input 32 of the block of calculation of arithmetic mean values of 3, the output of the counter-divider intensity values 3.02 connected to allows input element 331 BAN 3.03. Prohibiting entrance 332 item BAN 3.03 is a control input 35 of the block of calculation of arithmetic mean values of 3, the output element PROHIBITION 3.03 is the second information output 34 of the block 3.

The block type definition of the distribution 4, part of the General structural scheme is designed for the comparison of the parameter values of the ratio of the durationwith single threshold and deciding set of a homogeneous input stream to one of the three known variants of threads multipackage messages. The block circuit diagram for determining the type of distribution 4 and the principle of its action is known and described in the prototype (see RF patent №2165100, 7 G 06 F 17/18, 2001, bull. 10, 6).

The evaluator distribution 5, part of the General structural scheme, designed to produce estimates of the probability density of the flow. Its block diagram and principle of operation are known and described in the prototype (patent RF №2165100, 7 G 06 F 17/18, 2001, bull. 10, IG).

The control unit 6, which is a General block diagram is intended to adjust the monitoring interval (t_neb), packet length (m_p) and the number of packets (k). Block diagram of the control unit 6 is known, is described in detail in the prototype (patent RF №2165100, 7 G 06 F 17/18, 2001, bull. 10) and is presented on Fig.7. this description. The control unit 6 comprises a generator of clock pulses 6.01 and three divider with variable division factor (DPCD) 6.02-6.04. The generator output clock pulses 6.01 connected in parallel to the inputs of all three divider with a variable division ratio, the first of which (6.02) is clocked by the output 61 of the block 6 and is connected to the clock input 56 of the transmitter distribution 5. Output DPKG 6.04 and output DPKG 6.03 are respectively boleushim 62 and 63 managing the outputs of the block 6 and is connected to boleosoma 22 and the Manager 23 inputs of the computing unit 2 parameters.

The unit of analysis of the volume 7, part of the General structural scheme, intended for the current evaluation and comparison of previous (λ₁) and subsequent (λ₂) values of samples of the intensity parameter λ the input stream. Its structural scheme is known, described, for example, in the prototype (patent RF №2165100, 7 G 06 F 17/18, 2001, bul) and presented on Fig of this description. The unit of analysis intensity 7 is iselement comparison 7.01, memory element 7.02, the primary element And 7.03 and a secondary element And 7.04. Information input 701 element of comparison 7.01, the input memory element 7.02 and information input 706 of the secondary element And 7.04 connected in parallel with each other, and their input is 71 unit of analysis of intensity 7, auxiliary output 707 memory element 7.02 connected to the auxiliary input element 702 comparison 7.01, the output of which is control the output 72 of the unit of analysis intensity 7 and connected in parallel with signal inputs 703 and 705 primary 7.03 and secondary 7.04 elements And, respectively, the outputs of which are respectively the first 73 and second 74 parametric outputs of the unit of analysis of intensity 7. Information output 708 memory element 7.02 connected to the information input 704 of the primary element And 7.03. The element of comparison 7.01 is a digital site comparison and can be technically implemented in the form of node comparisons (digital comparator, as shown in [Gusev V.V., Lebedev O., Sidorov A. M. fundamentals of pulse and digital techniques. - SPb.: SUVIUS, 1996. S-152, RIS]. Random access memory element (OSE) 7.02 may be technically implemented on the basis of commercially available programmable memory device is istwa in accordance with the description, presented in [Gusev V.V., Lebedev O., Sidorov A. M. fundamentals of pulse and digital techniques. - SPb.: SUVIUS, 1996. S-199, figure 6.10].

Device for the parametric estimation of the distribution of the message flows is as follows. It is known that from the point of view of verification of the observed characteristics of pauses within the estimated flow of messages you can view the feature space pauses in the form of a set of (transposed vector) [1, 3, 5] type:

where- transposed vector characterizing (parametrically or incomplete) feature space breaks. The elements of this vector are: t_Bothe time of the identification bits of the end flag information of the previous packet (the last one in the previous message); t_bnthe time of identification bits flag the start information of the next packet (the first in the next message);- the time interval between messages (from t_Boto t_bn). There is a possibility of detection (identification, detection) signs pauses identified both quantitatively and qualitatively (false, incomplete, contradictory). This is accomplished by successive transformations using neural calculate what lnyh methods and algorithms, enabling the transition from false (inadequate, incomplete) the task of assessment to parametric.

In this case, the expression (1)characterizing false information (incomplete knowledge) about the facilities analyzed characteristic pauses to feature space pauses, is:

where- transposed vector space unreliable (incomplete) of the identifiable signs of pauses, the elements of which are obtained with the help of experts in the framework of the simplest hardware implemented neural network computational methods and algorithms. This interpretation of the model identification signs pauses allow you to enter the sequential algorithm inaccurate information (incomplete) of the identifiable signs pauses to mind, causing the possibility of parametric and unambiguous identification of these characteristics, to improve the reliability of the parametric estimation of the distribution of message flows.

In the framework of the traditional expert system for solving the problem of combining expert opinions, knowledge which is used in the form of pre-formed data about possible membership of a particular characteristic pause to the feature space, use one of the standard computational algorithms theory of neural networks - near the network extraprise computational algorithm, or the so-called extrapola neural network (Ann), which is a variety of well-known computational models of associative memory [3].

Computational neural network algorithm (extrapola neural network) this class consists of two layers of solvers (neurons) - the input layer S_aand the output layer S_b. The input layer S_aconsists of n_Ineurons with a set of forward and backward linkages with n_othe neurons of the output layer S_band the number of input and output images n = (n=(n_I=n_o)depends on the number of experts and the corresponding number of input devices hardware implementation of neural network algorithm. In our case, n can take values from 2 (two) to 23 (twenty-three), the number of inputs chip CAP. In Enns uses the so-called cognitive map, is completely defined by matrix of relations of the form:

Cognitive map characterizes the causal relationship of the individual characteristics of pauses and is formulated by experts, the principle of formation of cognitive maps is described in detail in the works [2, 3].

In this case n branches of the algorithm responsible for converting unreliable (incomplete) expert opinion on a specific characteristic pause to the feature space. Each elem is NT w _ijmatrix (3) defines the relationship of the i-th element (individual characteristic) space unreliable identifiable signs of pauses to the j-th element, and positive relationships are coded 1, negative -1 and the absence of linkages between elements is coded 0.

At the entrance of computational neural network algorithm (EDF) receives an input imagecharacterizing the variety known as parametrically (reliably), and unreliable set of elements (the individual characteristics of pauses). Define the q elements (individual signs pauses)that comprise a subset of Ω_qfrom a variety of source identifiable signs pausesat this point in time, specified quantitatively (parametrically), and what elements (individual signs pauses) from many of the identifiable signs of breaks required (preferred) to perform the task, authentic assessments, identified indefinitely (false, inaccurate). In order to obtain reliable results of the estimation of flow parameters necessary to reconstruct unreliable (incomplete, inaccurate) characteristics specified pauses characterizing the intervals between messages.

The functioning of computational neural network algorithm is as follows. Activates the input of the second layer S _anetwork input image characterizing identifiable signs pauses. In other words, evaluators (neurons) in the input layer is given in the initial state

Is the initial processors (neurons) output layer in accordance with the expression:

Is the cast of neurons in the input layer to the state of the neurons of the output layer:

The computation of the new state evaluators (neurons) output layer for all i∈ Ω_qaccording to the formula:

Is the repetition of the calculations of expressions (6) and (7) up until Enns reach a stable state:

The output of computational neural network algorithm (EDF) have output vectorreceiving a number of state values, which are determined on the basis of expressions (7) and (accurately) describe integrated expert opinion about a specific characteristic of the pause between messages to the feature space.

Reviewed computational neural network algorithm allows to eliminate uncertainty (inaccuracy, incompleteness) in identifying submit the Cove pauses, thereby uniquely identifying the message flow MEAs circulating in functioning SMCRS and LAN. Clarification, reconstruction unreliable (incomplete) of the identifiable signs of pauses allow you to increase the objectivity of the estimated parameters message flows, and ultimately to increase the reliability of the assessment of their distribution in the presence of inaccurate, incomplete, inaccurate, and often conflicting) information about the state of the environment of radio wave propagation, interference conditions, schedule and network behavior network management system and the influence of other destabilizing factors.

With this in mind, is the evaluation of the distribution of the message flows in the claimed device. The analyzed input stream IPU users SMCRS and LAN in the form of a binary pulse sequence, enhanced and normalized by the amplitude and duration of the input amplifier 1, is fed to the input 21 of the computing unit 2 parameters, in which you will determine the current values of the parameter ξcharacterizing the ratio of the information length of the packets and messages, and information to the input 81 of the unit of analysis the reliability of message 8, which defines the current valid values for parameter λdescribing the flow of messages.

The block you is ilenia parameters 2 can be implemented in accordance with the scheme, the proposed figure 5. The calculation in the block 2 to the current value of parameter ξ as follows. The decoder packet headers 2.01 analyzes the input 21 of the computing unit 2 parameters of the pulse sequence and finds in it the sign of the beginning of the data portion of the packet through the second input 222 runs counter information length of a multipackage messages 2.02, moments of complete accounts are determined by the moments of detection decoder 2.01 completion information part of the package. Moments of receiving the values of the information length message m_cin the transmitter the ratio of the information length of the packets and messages 2.03 and the reset of the counter 2.02 determined by the moments of occurrence of pulses on verification input 223 of the counter information of the length of a multipackage messages 2.02 coming through the verification input 22 of the computing unit 2 parameters with a verification output 82 analysis of reliability messages 8 and the locking completion message. The evaluator correlation information of the lengths of the packets and messages 2.03 defined values ξ=m_p/m_cfor fixed length packets received in SMCRS and LAN and set by the control unit 6, the flow through the control input 23 of unit 2 to the control input 232 transmitter 2.03 pulses the number of which is relevant to the duty to regulate the value of m _c. The obtained current values of the parameter ξ through the information output 24 of the computing unit parameters are received in binary code to the first information input block 31 calculation of arithmetic mean values 3.

Current valid values for parameter λcharacterizing the intensity of the message flow are determined based on the detection and identification signs pauses between messages and are calculated in the unit of analysis the reliability of message 8, which may be implemented in accordance with the scheme proposed in figure 2. Thus removing uncertainty (inaccurate, incomplete) in identifying signs of pauses is implemented as follows. From the information input 81 of block 8 of the investigated input stream IPU users SMCRS and LAN is fed to the input 811 controller pauses 8.01, which detects (identifies, registers) pause characterizing the end of the message. Output 812 controller pauses 8.01 identified quantitatively or qualitatively (false, incomplete) signs of breaks in the form of a binary pulse sequence is fed to the input 821 controller messages 8.02. The message controller 8.02 unit of analysis the reliability of message 8 may be constructed in accordance with the scheme shown in figure 3. Controller messages 8.02 comparative analysis of the characteristics of the PA is C from the point of view of their full compliance to the feature space pauses and conversion unreliable (incomplete) identified characteristics pauses to mind suitable for reliable parameter estimation, as follows. Identified quantitatively or qualitatively (false, incomplete) signs of breaks in the form of a binary pulse sequence is fed to the input 8021 shifting element 8.02-1, which provides a comparative analysis of the characteristics of pauses from the point of view of their full compliance to the feature space pauses as follows. Shift element 8.02-1 is designed to hold five bits of the incoming data, if the number of digits exceeds this number, then, from the point of view of mathematics, this information contains redundancy characterizing the causal relationship signs pauses, and therefore, any signs of pausesor all of themidentified unreliable (incomplete). In this case, the shifting element 8.02-1 does not record this information, and performs the functions of a transit node, sending the information through its false output 8023 to allow 8025 and information 8026 inputs of neural network evaluator 8.02-2. If the input 8021 shifting element 8.02-1 comes in binary code information with signs of a pause in the number of five digits, then all the signs of breaks identified reliably (about naznaceno, full),, shifting element 8.02-1 records this information and with its reliable output 8022 sends it to deny entrance 8024 neural network evaluator 8.02-2, and through the output controller 822 message 8.02 for direct entry 831 counter messages 8.03 and after verification the output 82 of the unit of analysis the reliability of message 8 on verification entrance 22 of the computing unit 2 parameters.

Neural network evaluator 8.02-2 controller messages 8.02 technically implemented on a programmable (from the point of view of the matrix of weights (3) - causal cognitive relations formulated by experts) parallel four-digit ALE that implements computational neural network algorithm (4)-(8) in accordance with the scheme depicted in figure 4. The conversion characteristics of pauses identified unreliable (incomplete), to a form suitable for parametric assessment is carried out as follows. Reliable (full) information signs pauses, acting in binary code through prohibiting entrance 8024 neural network evaluator 8.02-2 on prohibiting entrance WITH ALE, blocks the chip. False (incomplete) information signs pauses comes in binary code via the enable input 8025 neural network evaluator 8.02-2 enable input M ALE, initiating not routeway computational process, and entered through the information input 8026 neural network evaluator 8.02-2 on n inputs programmable ALE.

Programmable ALE, based on the programmed values of the elements (w_ij) weights matrix W (3) is analytically described the causal cognitive relations formulated by experts, provides the procedure for calculating (extrapolation) in accordance with the computational neural network algorithm (4)-(8). In this case n inputs (1, ..., n) programmable ALE are equal n inputs (n_I) calculators (neurons) in the input layer S_aEnns, which in binary code are elements of the vectorhaving physical sense information from unreliable (incomplete) identified characteristics of pauses. The set of backward and forward linkages of n_Iwith n_oEnns, software-implemented within the programmable ALE, allows to take into account when computing the values of the elements (w_ij) weights matrix W (3) and obtain n outputs (n_oin the result of an arithmetic calculation in accordance with the computational neural network algorithm (4)-(8) extrapolated values of the vector elementscharacterizing preliminary decision about the degree of membership of a particular characteristic pause the feature space. The filing (and boy is deposited) in binary code on any i-th input ALE (in any i-th cell memory ALE) value false (incomplete) of the identifiable characteristic pause triggers the issuance of a binary code from the i-th output ALE (output of the i-th neuron of the output layer S_b) programmed in accordance with the computational neural network algorithm (4)-(8) values are mathematically transformed, relative to the identified characteristic pause. As a result, the output F of the logical element ALE engaged in the operation of the final verification (formation of a generalized opinion), and the output 8027 neural network evaluator 8.02-2 obtained in the binary code information reliably (full) identified characteristics of pauses produced quantitative characteristic values of breaks, fully (accurately) describing integrated expert opinion about a specific characteristic pause to the feature space.

Initially reliable information from credible exit 8022 shifting element 8.02-1 and verified the information obtained in the neural network calculator 8.02-2 quantitative (relatively accurate) values identified signs of pauses, in binary code is supplied via the output controller 822 message 8.02 for direct entry 831 counter messages 8.03 and after verification the output 82 of the unit of analysis the reliability of message 8 on verification entrance 22 of the computing unit 2 parameters. In the computing unit settings 2 this information is used as a command, the locking completion message, and is fordetermination moments of receiving the values of the information length message m _within the transmitter the ratio of the information length of the packets and messages 2.03 and the reset of the counter 2.02 block parameter calculation 2.

Current valid values for parameter λcharacterizing the intensity of the message flow is determined by the message counter 8.03 the pulses from the output controller 822 message 8.02 for the time interval defined by the control unit 6, by applying pulses (frequency f_nebwith Abdullayeva output 62 of the control unit 6 through Abdoulaye input 84 unit of analysis the reliability of message 8 on Abdoulaye entrance 832 counter 8.03.

Thus, based on the mathematical apparatus of neural networks, involving the identification criteria (causal cognitive relations), which are formulated by the experts, by mathematically correct calculation unit of analysis the reliability of message 8 managed to achieve that output 833 counter messages 8.03 via the information output 83 of the unit of analysis the reliability of message 8 to the input 71 of the unit of analysis intensity 7 and the second information input 32 of the block of calculation of arithmetic mean values of 3, despite often present in real communication channels (on the background noise and errors in the channels SMCRS and LAN) unreliability (incomplete) ID signs of breaks come in the form of verified binary the pulse sequence current valid values for parameter λ .

Further stages parametric estimation of the distribution of message flows are implemented as follows. In the block of calculation of arithmetic mean values of 3, the first 31 and second 32 information the input of which is in parallel code act upon the current values of the parameter ξ accurate and current value of parameter λ, determined by the numerical values averaged over the observation interval of flow parameters (). The power calculation of arithmetic mean values of 3 can be implemented according to the scheme shown in Fig.6. In the case when there is a homogeneous (non-mixed) flow MEAs users SMCRS and LAN, when the source generates messages regularly, the calculation of the numerical values averaged over the observation interval of flow parameters (or) proceeds as follows. Parameter values ξ binary sequence received at the first input 31 of the block 3, and then to the input of counter-divider values duration 3.01 producing an arithmetic operation of summing successively incoming values ξ and dividing the sum by the number of samples. At the output of the counter-divider values duration 3.01 and on the first information output 33 Blo is and calculate the average values of 3 in parallel binary code get the value of the parameter averaged over the observation interval. Meaningful values for parameter λ binary sequence received at the second information input 32 of the block 3, and then to the input of counter-divider intensity values 3.02 producing an arithmetic operation of summing successively incoming values λ and dividing the sum by the number of samples. At the output of the counter-divider intensity values 3.02 and permissive input element PROHIBITION 3.03 in parallel binary code have a value foraveraged over the observation interval. The second information output 34 of the block 3 with the output element PROHIBITION 3.03adjudged only if at the moment on the control input 35 of the block 3 and for prohibiting the input element PROHIBITION 3.03 not submitted in parallel binary code signal from the control output 72 unit rate analysis 7. Otherwise, the second information output 34 of the block 3 with the output element PROHIBITION 3.03 valuenot served. The observation interval to calculate the parametersandis set by the control unit 6 depending on the required accuracy.

The control unit 6 may be implemented according to the scheme, the offer is authorized 7. Formation of control pulse sequences is carried out by setting the elements of the group divider with variable division factor (DPCD) the corresponding ratios of the pulse sequence received from the output of the generator 6.01. Installing coefficient element 6.03 is determined by the value of t_neb=1/f_nebby setting the ratio of the element 6.04 determined by length of service adopted for this Protocol is currently working SMCRS and LAN (m_p)sets the ratio of the element 6.02 determined necessary to calculate the number of packets (k).

When the first 33 and second 34 information output unit 3 are simultaneously the values of the parametersandaccordingly, the value ofwith the first data output 33 unit 3 is input to the block type definition of the distribution of 4, which is compared with a single threshold. Block diagram of the block type definition of the distribution of the 4 known and described in the prototype (see RF patent №2165100, 7 G 06 F 17/18, 2001, bull. 10, 6). Determine the type of flow distribution IPU as follows. coming from the output of block 3 in parallel binary code digital to analog preobrazovala the ü (DAC) 4.01 of the block type definition of the distribution 4, is converted into an analog voltage, which Comparators 4.02, 4.03 unit determining the type of distribution 4 is compared with the threshold voltage produced by the generator threshold voltage 4.04 unit determining the type of distribution 4. The value of the threshold voltage corresponds to a single. The output 42 of the comparator 4.02 a signal of logical units in excess of the value ofa single value and is supplied to the second modification of the input 54 of the transmitter distribution 5, in the opposite case, the emergence of logical units occurs at the output of comparator 41 4.03 and is supplied to the first modification of the input 53 of the transmitter distribution 5. When there are logical zeros on the outputs of the Comparators 4.02 and 4.03 decides that=1, this is evidenced by the presence of a logical unit at the output 43 of the element OR NOT 4.05 block type definition of the distribution 4 and the receipt of this logical unit to the third modification of the input 55 of the block 5. Thus, by comparing the results by filing a control pulse on one of the three modification inputs 53-55 evaluator distribution 5 the decision on the calculation of one of the three well-known and described in the prototype (see RF patent №2165100, 7 G 06 F 17/18, 2001, bul, expression (1-3)) analytical expressions for probability density for values of parametersandcoming from the first 33 and second 34 information output unit 3 respectively on the first 51 and second 52 informational inputs evaluator distribution 5.

Structural diagram of the transmitter of the distribution of the 5 known and described in the prototype (see RF patent №2165100, 7 G 06 F 17/18, 2001, bull. 10, figure 3). Computing in the computer distribution 5 can be carried out for different values of the number of packets (k), the choice of values is performed by the control unit 6 by submission on the clock input 56 of the block 5 and to the first input 510 item And 5.02 of pulses the number of which corresponds to the specified number of packets. Admission to the second input 511 item And 5.02 of pulses from the first information input 51 unit 5 provides the appearance of the output element And 5.02 evaluator distribution 5 of the control signal, the number of pulses which corresponds to the specified number of packets. This signal, when the clock input 506 arithmetical-logical element 5.01 evaluator distribution 5, initiates the beginning of the calculation procedure. The calculated values of the uniform distribution of flow MEAs output 59 evaluator distribution 5 served in binary is parallel to the recording device.

In the case when there is a heterogeneous (mixed) flow MEAs users SMCRS and LAN, when the sources of mixed type display messages irregularly - single ISP or packs COI with lengths distributed according to a geometric law, the calculation of the numerical values averaged over the observation interval parameter flow (occurs as described above, and the calculation of the numerical values of the parameter λintensity of flow, as follows. Accurate (verified) parameter value λ binary sequence received from the information output unit of analysis the reliability of message 8 to the second information input 32 unit 3 for calculating averaged over the observation interval parameter flow () and to the input 71 of the unit of analysis of intensity 7. The unit of analysis intensity 7 may be implemented in accordance with the scheme shown Fig. Parameter values λ received in parallel on the information input 701 element of comparison 7.01 unit 7, to the input of memory element 7.02 unit 7 and on the information input 706 of the secondary element And 7.04 unit 7. Online storage element 7.02 stores the value of parameter λ in the previous step value of λ₁. With the auxiliary in the course 707 memory element 7.02 value λ ₁arrives at the auxiliary input element 702 comparison 7.01, which compares the value of the previous reference parameter (λ₁with a value of the subsequent reference (λ₂) is received by the information input 701 element of comparison 7.01. The comparison is made according to two criteria: λ₁≠λ₂and λ₁=λ₂. If the values of the counts is equal to (λ₁=λ₂), this confirms that at the input device, we have a uniform (non-mixed) flow map, and therefore, the comparison element 7.01 does not generate a signal on the control output 72 unit 7 and the signal inputs of elements And 7.03 and 7.04. This means that the control input 35 of the block 3 and for prohibiting the input element PROHIBITION 3.03 unit 3 control output 72 unit 7, the signal is not received, therefore, the output element PROHIBITION 3.03 and the second information output 34 of block 3 in parallel binary code have a value forand further calculation of the parameters of the distribution of the message flows is as described above.

If the counts are not equal (λ₁≠λ₂), this confirms that at the input device, we have a heterogeneous (mixed) flow map, and therefore, the comparison element 7.01 outputs a signal on the control output 72 unit 7 and the signal inputs of the element is in And 7.03 and 7.04. The signal from the control output 72 unit 7 is supplied to the control input 35 of the block 3 and for prohibiting the input element PROHIBITION 3.03, locking him out and not allowing you to giveon the second information input 52 of the block 5. From the information output 708 memory element 7.02 unit 7, the value of parameter λ₁is supplied to the information input 704 of the primary element And 7.03, and online storage element 7.02 records the value of the new parameter. On the information input 706 of the secondary element And 7.04 unit 7 have the value λ₂. The output of the comparison element 7.01, acting on the signal inputs 703 and 705 of the primary and secondary elements And (7.03 and 7.04, respectively, opens the device, allowing you to get on their outputs, and hence the first 73 and second 74 parametric output unit 7 parameter value λ₁and λ₂respectively. Parameter values λ₁and λ₂come to the first 57 and second 58 parametric inputs evaluator distribution 5, and then on the first 507 508 and the second parametric inputs arithmetical-logical item 5.01, which calculates the values of the distribution in accordance with the analytical expression to determine the convolution of the corresponding probabilities (see RF patent №2165100, 7 G 06 F 17/18, 2001, bull. 10, the expression is s (4)), moreover, the presence of a control pulse on any of the three modification inputs 53-55 evaluator distribution 5 in this case does not play a role. The calculated values of the distribution is heterogeneous (mixed) flow MEAs output 59 calculator distributions 5 are served in parallel binary code on the recording device.

In the end, the output of block 5 are written in binary code reliable, verified using the mathematical apparatus of theory of neural networks, the values of the distribution of either homogeneous or heterogeneous (or mixed, depending on the characteristics of the graph SMCRS and LAN) flow is uniquely identified multipackage messages.

Thus, the analysis of the principle of operation of the inventive device for the parametric estimation of the distribution of message flows shows the obvious fact that along with saved possibilities of estimating the distribution parameters of a heterogeneous (mixed) flow MEAs device capable of improving the accuracy of estimation of distribution parameters message flows, identification signs pauses (signs of the beginning and end of the pause between messages) can be identified both quantitatively and qualitatively inaccurate, incomplete.

This device provides increased reliability OC is nivenia in terms inherent in the real process of functioning SMCRS and LAN - in terms of unreliability (failure, incompleteness and inconsistency) identification signs pauses circulating flows multipackage messages, due to the volatility (variability, stochasticity) conditions of functioning networks of this class, which significantly extends the application of the device, extends the functionality of the equipment, where the claimed device for the parametric estimation of the distribution of the message flows will be used.

SOURCES of INFORMATION

1. Wasserman F. Neurocomputer technique: Theory and practice. - M.: Mir, 1992. - 240 S.

2. Kosko Century Fuzzy cognitive maps // International Journal of Man-Machine Studies. V.24. N.Y., 1986. P.16-22.

3. Shcherbakov M.A. Artificial neural network. - Penza: PSTU, 1996. - 44 S.

4. Trahtengerts E.A. Computerized decision support. - M.: SINTEG, 1998. - 342 C.

5. Gorban A.N., Rossii D.A. Neural network on a personal computer. - Novosibirsk: Nauka. Siberian publishing firm Russian Academy of Sciences, 1996. - 146 C.

1. Device for parametric estimation of the distribution of message flows that contain the input amplifier, the input of which is the input of the block parameter calculation unit calculating average values, the block type definition of the distribution, the evaluator RA the distribution the unit of analysis of the intensity and the control unit, clock output of which is connected to the clock input of the transmitter distribution, control output control unit connected with the control input of the computing unit parameters, the information output of which is connected to the first information input unit for computing the average value, the first information output of which is connected to the first information input of the transmitter distribution and with the input of the block type definition of the distribution, the first, second, and third modification, the outputs of which are connected respectively to the first, second and third modification to the inputs of the transmitter distribution, the second information input of which is connected with the second information output unit for computing the average values, the output of the transmitter distribution is the output of the device, the first and second parametric outputs of the unit of analysis intensity are connected respectively with the first and second parametric inputs evaluator distribution, control output unit of analysis intensity is connected to the control input of the computing unit arithmetic mean values, characterized in that it further introduced the unit of analysis the reliability of the message, and the output of the input amplifier is connected to the information input of the Loka calculation parameters and information input unit of analysis the reliability of the message, the information output of which is connected to the second information input unit for computing the average values and the input unit of analysis intensity, Abdoulaye output control unit connected to boleushim input unit of analysis the reliability of message verification the output of which is connected to a verification to the input of the computing unit parameters.

2. Device for parametric estimation of the distribution of message flows according to claim 1, characterized in that the unit of analysis the reliability of the message consists of a controller pauses, controller messages, and the message counter, the output of which is an information output unit, the input of the controller pauses is an information input unit, the output of the controller pauses connected to the input of controller messages, the output of which is the verification unit output and is connected to the direct input of the counter message, Abdoulaye entrance which is boleushim input unit of analysis the reliability of the message.

3. Device for parametric estimation of the distribution of message flows according to claim 1, characterized in that the computing unit parameters consists of a decoder packet headers, the counter information of the length of a multipackage messages and computer ratio information of the lengths of the packets and messages, the output of which is information which was output unit, the input of the decoder packet headers is an information input unit and connected to the first input counter information length of a multipackage message, the output of the decoder, the packet header is connected with the second input counter information length of a multipackage messages, verification input of which is the verification of the input block, the output of the counter information of the length of a multipackage messages connected to the direct input of the transmitter correlation information of the lengths of the packets and messages, the control input of which is a control input of the computing unit parameters.