RU2439705C1 - Method to assess information capacity and priority of computer network technical condition parameters - Google Patents

Abstract

FIELD: information technologies.

SUBSTANCE: in the method, using an analyser, computer network technical condition characteristics are measured, an average value of each parameter and a total average value of parameters by number of completed measurements are calculated in a data processing unit, a correlation matrix is built on their basis. Also probability is calculated for network failure to fulfil tasks to the end of time of testing, measurements in a unit of data processing after calculation of a correlation matrix of controlled parameters of computer network technical condition, probability is calculated for exceeded permissible values for every parameter of computer network technical condition, as well as probability for network failure to perform tasks to the end of testing time by number of faults in the network operation.

EFFECT: increased efficiency of computer network technical condition diagnostics and reduced time of diagnostics.

3 dwg

Description

The proposed technical solution relates to the field of computer technology, namely to the field of monitoring the technical condition (TS) of elements of computer networks, their monitoring and inspection and can be used to adjust and reduce the number of controlled characteristics of the TS of the computer network.

There is a method of determining the information content of the parameters of the studied system, described in (Iberla K. Factor analysis (Mathematical and statistical methods abroad.) / Translated from German by V.M. Ivanova. - M .: Statistics - 1980), including the identification of the most significant signs (factors) for a large number of experimental data. The number of factors identified is less than the set of initial values, the structure and relationships of the identified components of the process (system) are determined, the significance of each factor is calculated.

There is a method of determining the priority of the parameters of the studied system, described in (Petrichenko G.S., Dudnik L.N. Choice of parameters for monitoring the technical condition for digital blocks of the corporate network based on the use of factor analysis methods. // Automation and modern technologies. M.: Mechanical engineering. , 2010. - No. 2. P.16-21), including the identification of the most significant features (factors) and the transition from the description of the investigated system, given a large set of measured parameters, to a description of a smaller number of the most informative signs that reflect the most important properties of the system. The factor mapping matrix A defined in this method allows one to select for each factor a group of parameters that is most closely associated with it, calculate the priority coefficients of the parameters taking into account their interconnections and select the most significant parameter for monitoring the technical condition of digital network units.

However, these methods provide only an assessment of the priority of highlighting the most significant features of the system, but they do not allow determining the measure of information of the resulting set of features and estimating information loss when discarding the least informative parameters, which reduces the accuracy of diagnostics of the TS of a computer network (CS) based on the selected set of monitored parameters.

The closest in technical essence to the proposed technical solution is the method described in (Bulkin M.A., Dubitsky L.G. Information content and priority of parameters in assessing product quality. // Electronic Technology. Ser. 8, 1972. Issue 5) . The prototype method consists in determining the information content of the parameters using the Shannon function. It consists in determining the relationships between the considered parameters of the technical state of the system by sequentially calculating the average values of each parameter for all controlled systems, calculating the total average values of the parameters according to the number of measurements taken, calculating the covariance matrix and constructing a correlation matrix based on them, calculating an additional parameter - failure-free time the work or the probability of failure at the end of the test time, the average values of this parameter and the calculation of the coefficients correlating them, adding the last row and the last column of the obtained probability correlation matrix as the last row and column to the correlation matrix of the parameters of the technical state of the system, forming a transformation that describes the internal relationships of the considered parameters of the controlled object, the product of the columns of which gives the corresponding element of the resulting final correlation matrix, calculating the priority coefficients of each indicator based on the elements of the constructed transformations, evaluating the information content of the initial set of characteristics using the Shannon information measure function, discarding parameters with the lowest priority coefficients, determining the information content of the reduced parameter set, calculating the information loss of the system of the remaining parameters of the technical condition compared to the initial system of indicators, and finally selecting the set of the most important control parameters in compliance with the initially set maximum level of information loss.

The disadvantages of the method are the impossibility of applying it to intermittent, unstable processes, incorrect formulation of the calculation of the final correlation matrix after calculating a new parameter, the presence of errors in calculating the failure probabilities of the systems under study, as well as centered probabilities. These shortcomings lead to a decrease in the accuracy of calculations.

The objective of the proposed method is to assess the information content and priority of the monitored parameters of the CS CS, reduce the totality of the controlled characteristics by eliminating the least important of them, which will increase the efficiency of diagnostics of the CS CS, reduce the cost of acquiring diagnostic equipment, and carry out the diagnostic process, which will increase the availability and operability of the CS enterprises. As the parameters of the CS CS, the number of subscribers in the network, throughput, subscriber intensity, average customer service time, failure rate, network data loading, conflict loading, etc. can be considered.

The technical result of the invention is to increase the efficiency of diagnostics of the TS of a computer network by reasonably reducing the number of monitored parameters of the TS CS during its continuous operation and monitoring the state of the CS in generalized control points that provide the user with maximum information about the diagnosed object, which is achieved by calculating additional probabilities of the parameters of the CS CS for limits of acceptable values and probabilities of network failures. In addition, as a result of the transition to control of a smaller number of parameters of the CS CS, the time and financial costs of diagnosing and further monitoring the technical state of the CS are reduced.

The technical result is achieved by the fact that in the method of evaluating the information content and priority of the parameters of the TS CS, consisting of sending a signal by the analyzer measuring the characteristics of the technical state of the computer network, storing values in the data storage unit, recording the measured values in the data processing unit, in which average values are sequentially calculated of each parameter for all controlled systems (similar to the purpose and operating conditions of the sections of the computer network), calculate the total average value parameters according to the number of measurements taken, the elements of the covariance matrix are calculated and the correlation matrix is built on their basis, an additional parameter is calculated - the probability of the network not meeting the tasks by the end of the test (measurement) time, the average values of this parameter and the correlation coefficients are calculated by them, add the last row and the last a column of the obtained correlation matrix of probabilities as the last row and a column to the correlation matrix of parameters of the technical state of the computer network form a transformation that describes the internal connections of the considered parameters of the controlled network (or its section), the product of the columns of which gives the corresponding element of the resulting final correlation matrix, calculate the priority coefficients of each indicator by solving a system of equations obtained on the basis of the elements of the constructed transformation, evaluate the information content of the initial set of characteristics using the Shannon information measure function, discard the parameters with the smallest coefficient m priority determining information content reduced set of parameters is calculated system loss information remaining parameters technical state compared to the original system performance, the final selection is performed together the most important parameters of the technical state in accordance with the user-specified maximum level of information loss. The calculation of additionally estimated parameters is carried out according to the probability of going beyond the permissible values for each parameter and the probability of the network not meeting the tasks by the end of the test time. The calculation of the additional parameter is carried out in the data processing unit after calculating the correlation matrix of the controlled parameters of the technical state of the computer network, the probability of going beyond the acceptable values for each parameter of the technical state of the computer network and the probability of the network failing to complete tasks by the end of the test time by the number of network failures are calculated an additional parameter is carried out on the basis of average probabilities of network failures for all measurements for each characteristic the technical condition of the computer network and the average probability of network failures for all characteristics for each measurement.

An example implementation of the proposed method to an existing computer network is implemented as follows. Figure 1 shows a generalized diagram of the interaction between the CS and the analyzer, figure 2 shows the options for placing the analyzer on the network, figure 3 shows the implementation diagram of the proposed method. The figures are illustrated by the following notation:

1 - section of a computer network;

2 - analyzer;

3 - block storage values of the controlled parameters;

4 - data processing unit;

4.1 - correlation data processing;

4.2 - calculation of a new parameter, calculation of an additional correlation matrix;

4.3 - prioritization of parameters;

4.4 - regrouping of parameters;

4.5 - assessment of the information content of the original parameter system;

4.6 - exclusion of parameters, estimation of information loss;

5 - display of results on the screen;

5.1 - display of priority coefficients on the screen, output of the level of information loss;

5.2 - display on the screen of the final set of the most important parameters of the TS CS;

6 - workstations in the network;

7 - router;

8 - hub;

9 - server.

When building CS 1 on the basis of hubs 8, there is a common data transmission medium, all nodes of the network exchange packets competing for access to this medium, and a packet sent by one node of the network is transmitted to all ports of the hub and all other nodes of the network are listening to this packet, 7, 9, but only the node to which it is addressed receives it. A packet analyzer 2 is installed on one of the nodes of the concentrator 8, which will intercept all network packets related to this segment of the CS.

Using analyzer 2 (software or hardware), the CS is loaded, its individual section or separate workstation 6 with data, after which a feedback signal is sent to the analyzer, which determines the values of the CS parameters it controls in this section, the data is stored in a separate file in memory the processor device 3, then entered into the processing unit of the accumulated values of the parameters 4, in which the calculation according to the proposed method and analysis are performed in order to identify the most and least important characteristics stick TS KS.

Data is sent on any segment of the computer network, to each computer (workstation - PC).

The implementation of the claimed method consists in the fact that the return signal received by the analyzer 2 is processed, the technical characteristics are evaluated, the obtained values are entered into the data storage unit 3, which contains statistical information about the change in the values of various characteristics of the CS. The accumulated information is entered by the network operator or administrator in a specially prepared block 4, which performs the calculation in the following sequence: in the data correlation processing block 4.1, the additional parameter calculation block 4.2 — the probabilities of the network not fulfilling tasks by the end of the measurement time, in which the average values of the new parameter are calculated , calculating the correlation coefficients, adding the last row and the last column of the obtained correlation matrix of probabilities as after them to the row and column of the correlation matrix TC KS parameters calculated in the previous block. Based on the obtained matrix, in the next block 4.3, a transformation is formed that describes the internal relationships of the considered parameters of the controlled network (or its section), the product of the columns of which gives the corresponding element of the resulting final correlation matrix, the priority coefficients of each indicator are calculated by solving the system of equations obtained on the basis of the elements constructed transformations, in block 4.4, the rows and columns of the correlation matrix are rearranged in increasing order The obtained priorities are controlled by the analyzer characteristics of the CS. In the next block 4.5, the information content of the considered system of parameters is estimated using the Shannon information measure function, in block 5.1, the obtained values of the information function and characteristics priority coefficients are displayed on the screen, in block 4.6, the parameters with the least significance (priority) are excluded, the information content of the reduced set of parameters is determined , the value of the loss of information of the system of the remaining parameters of the technical condition is calculated in comparison with the original system indicators, the result is displayed on screen 5.1. The final set of vehicle parameters is selected until the level of information loss set by the network administrator, which was initially set by the network administrator, is reached. In block 5.2, the final set of the most important parameters of the CS CS is displayed on the screen.

The technical result is achieved by sequentially calculating the priority coefficients of the characteristics of the CS CS, step-by-step control of information loss while excluding the least significant parameters and preventing the user from exceeding the interval of the acceptable level of information loss in order to maintain the computer network in the required standby mode.

The application of the proposed method entails an increase in the operability of the investigated object, improvement of its technical characteristics, prevention of sudden failures, and the possibility of choosing a small set of the most important parameters of the CS CS, which reduces the cost of the diagnostic process and further forecasting, as it significantly reduces the cost of acquiring diagnostic equipment or software providing.

The implementation of the method for evaluating the information content and priority of the parameters of the TS CS is as follows. The information approach used in the data processing unit 4 includes the process of compiling a correlation matrix for the values of the initial parameters of the CS CS, taking into account the probabilities of matching the values of the considered characteristics to the required level, constructing a transformation that allows determining the priority coefficients of the parameters, using the Shannon information function to determine the information content of various sets of parameters, and also information loss with the exclusion of unimportant characteristics. The following problem is posed.

. Необходимо найти такую функцию f(x), которая бы, описывая свойства системы, отвечала следующим требованиям: вектор

содержал бы ту же информацию о свойствах системы, что и вектор

; число компонент M вектора

было бы меньше числа компонент N вектора

.Let some system be characterized by a set of N parameters x ₁ , ... x _N , which we denote by the vector

. It is necessary to find a function f (x) which, describing the properties of the system, meets the following requirements: vector

would contain the same information about the properties of the system as the vector

; the number of components M of the vector

there would be less than the number of components N of the vector

.

The observational data obtained by monitoring P of similar systems at various time points are used. In this case, there is a set of N parameters for each system {x _ij (k)}, where i = 1, ..., N, j = 1, ..., P is the number of the system being monitored (KS 1 or its section 6, 7, 8 , 9), k is the number of measurement of parameters (k = 1, ..., T).

, k=1,…, T, где n_o,k - общее число отказавших систем по всем параметрам при k-том замере.With a limited test time, an additional parameter x _{N + 1} is introduced, which is defined as the probability of a system failure by the end of the test time. The values of the parameter x _{N + 1} for each measurement k are

, k = 1, ..., T, where n _{o, k} is the total number of failed systems in all parameters at the k-th measurement.

The algorithm for evaluating the information content and priority of the parameters of the TS of the enterprise CS is presented in the form of a sequential implementation of the following steps (figure 3).

1-2 - the interaction of the analyzer 2 with the computer network 1 is performed (sending a signal on the network channel and receiving feedback from the analyzer). The feedback signal received by the analyzer is processed, and the values of the measured characteristics are estimated.

3 - saving the obtained values in the storage unit of the measured parameters 3.

4 - processing of accumulated information. The administrator of the CS performing its diagnostics enters the data stored in the third stage into the block 4 prepared in advance for calculations according to the claimed method, in which:

,

, i=1,…, N, j=1,…, N, N - количество измеренных характеристик.4.1 - based on the calculation of the elements of the covariance matrix, a correlation matrix is constructed

,

, i = 1, ..., N, j = 1, ..., N, N is the number of measured characteristics.

4.2 - a new parameter is calculated - the probability that the network will not complete the tasks by the end of the test (measurement) time, the average values of the parameter and the calculation of the correlation coefficients are calculated, the last row and the last column of the obtained correlation probability matrix are added as the last row and column to the correlation matrix of the TS parameters CS calculated in block 4.1.

, i=1,…, N; j=1,…N;

.As C _{i, N + 1} (C _{N + 1, j} ), the correlation between the probabilities of failure to go beyond the tolerances of the parameter x _i and the failure at the end of the test x _{N + 1 is} calculated. From the resulting matrix, auxiliary ones are distinguished:

, i = 1, ..., N; j = 1, ... N;

.

4.3 - the transformation α is constructed that describes the internal connections of the measured parameters of the controlled network:

- столбцы координатной матрицы

, C_ij=(e_i, e_j), то координаты α_ij можно вычислить по рекуррентным соотношениям (i=1,…, N+1, j=1,…, N+1):If

- columns of the coordinate matrix

, C _ij = (e _i , e _j ), then the coordinates α _ij can be calculated from the recurrence relations (i = 1, ..., N + 1, j = 1, ..., N + 1):

The calculation of the priority coefficients a _i parameters of the TS CS is made by solving the system (3):

The algorithm for solving the system are recurrence relations

4.4 - comparison of the priority coefficients by their absolute value, rearrangement and reindexing of the initial parameters by increasing absolute value a _i .

. Вывод значений коэффициентов приоритетности и информативности исходной системы параметров ТС КС на экран (блок 5.1).4.5 - assessment of the information content of the original system of parameters of the TS CS according to the formula

. The output of the values of the coefficients of priority and information content of the original system of parameters of the TS CS on the screen (block 5.1).

4.6 - parameters with the lowest priority coefficient are sequentially excluded; the information content of the system of the remaining parameters is calculated; The information loss of the resulting set of indicators is estimated in comparison with the initial one.

The calculation of the information functions I ₁ , I ₂ , ... is carried out according to the formulas:

where I _j is the information content of the system parameters when discarding the first j parameters, j = 1, 2 ... N-1; A _j and C _j are matrices obtained from matrices A and C by deleting the first j-rows and j-columns corresponding to the discarded j parameters.

Information loss is calculated by the formula:

5.1 - output of the values of the priority coefficients, the information content of the obtained system of parameters of the CS CS and the level of information loss after excluding the parameter on the screen; making decisions on the further exclusion of unimportant characteristics.

The parameters with the lowest priority are excluded in accordance with the rule: if the obtained level of information loss is satisfactory for the user (for example, does not exceed 30%), the next highest priority parameter is excluded (in order of increasing importance), the information content of the reduced set of characteristics is monitored and information loss at this stage. If the obtained level of information loss exceeds the limit set initially by the administrator of the CS, the process of excluding the parameters is terminated, and the set of indicators selected in the previous step is the desired one sufficient to effectively control the CS of the CS.

5.2 - display of the final set of characteristics on the screen.

The proposed method for evaluating the information content and priority parameters of the technical state of a computer network or its section will increase the efficiency of diagnostics of the CS of the CS, improve the technical readiness of the CS without interrupting its operation, reduce the time and financial costs of diagnosing and further monitoring the technical state of the CS due to the reasonable level of information loss reduce the number of monitored parameters and control only the most significant characteristics. Monitoring of the status of the CS is carried out in generalized control points, providing the user with the maximum information content about the diagnosed object, which is achieved by calculating additional probabilities of the parameters of the CS CS beyond the permissible values and the probabilities of network failures. The method can be applied to any set of indicators of technical condition.

Claims (1)

A method for evaluating the information content and priority of the parameters of the technical state of a computer network, consisting of sending a signal to an analyzer measuring the characteristics of the technical state of a computer network, storing values in a data storage unit, recording the measured values in a data processing unit, in which the average values of each parameter are successively calculated for all monitored systems similar in purpose and operating conditions to sections of a computer network, calculate the total average value of the parameter According to the number of measurements taken, the elements of the covariance matrix are calculated and the correlation matrix is built on their basis, an additional parameter is calculated - the probability of the network not meeting the tasks by the end of the test, measurement, average values of this parameter and the correlation coefficients are calculated by them, add the last row and last column the obtained correlation matrix of probabilities as the last row and column to the correlation matrix of parameters of the technical state of the computer network, I form t the transformation describing the internal relationships of the considered parameters of the controlled network or its section, the product of the columns of which gives the corresponding element of the resulting final correlation matrix, calculate the priority coefficients of each indicator by solving the system of equations obtained on the basis of the elements of the constructed transformation, evaluate the information content of the original set of characteristics using the function Shannon information measures, drop parameters with the lowest coefficient prioritize In this case, they determine the information content of the reduced set of parameters, calculate the information loss of the system of the remaining parameters of the technical condition in comparison with the initial system of indicators, make the final selection of the set of the most important parameters of the technical state in accordance with the user-specified maximum level of information loss, the calculation of additionally evaluated parameters is carried out by the probability of exit beyond the permissible values for each parameter and probability not fulfilled by the network of tasks at the end of the test time, characterized in that the additional parameter is calculated in the data processing unit after calculating the correlation matrix of the controlled parameters of the technical state of the computer network, the probability of exceeding the permissible values for each parameter of the technical state of the computer network and the probability of the network not meeting the tasks are calculated by the end of the test time according to the number of network failures, the calculation of an additional parameter is carried out on the basis of average probabilities failure rates for all measurements for each characteristic of the technical condition of the computer network and the average probability of network failures for all characteristics for each measurement.

Images