JP2011040954A - System, method and program for evaluating network - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To estimate the influence of various fault instances exerted on a user including an instance in that a fault is not found and the influence of the fault exerted on the user is unclear, and to evaluate the reliability of a network. <P>SOLUTION: Network fault information, evaluation object information and an evaluation reference value are input (procedure 0). Fault instances included in the network fault information are classified into normal fault instances and the fault instances for which the degree of the influence by the fault is unclear (procedure 1). The degree of the influence of the fault of evaluation object equipment is calculated from the fault time of the network fault information and the number of influenced users (procedure 2), and a relation between the degree of the influence of the fault relating to the normal fault instances and the number of reports from the users relating to the respective fault instances is estimated (procedure 3). The degree of the influence is corrected using the estimated relation for the instance for which the degree of the influence by the fault is unclear in the classified fault instances (procedure 4), the reliability of the network is evaluated on the basis of the degree of the influence of the fault and the corrected degree of the influence of the fault (procedure 5), and an evaluated result is displayed (procedure 6). <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a communication network technology realized by a network protocol typified by IP (Internet Protocol), and in particular, a component constituting a network (a device for transferring a packet such as an IP packet, a link for connecting devices). Based on the failure information of the network equipment such as transmission lines and transmission line equipment to be configured) and the information on the number of reports from users affected by the failure, various types of failure cases occurring in the network will affect the user By estimating the impact and evaluating the reliability of the network, failure cases and network equipment that have a large impact on reliability degradation are selected, and the operation management of a highly reliable or desired network is automatically performed. Network evaluation that can be realized efficiently and accurately Stem and network evaluation methods, as well as a program therefor.

  Conventionally, the reliability of a communication network depends on the equipment that constitutes the target network (for example, network components such as network devices and links) and the equipment configuration (facility connection configuration or redundant configuration). It is created and calculated using the product or sum of failure probabilities (for example, failure rate and unavailability) of individual network equipment. Then, based on the failure probability required for the communication network, the failure probability required for each network facility is allocated, and calculated based on the allocated failure probability and the evaluation model. Has been evaluated by comparing the failure probability.

  As a conventional technique related to the calculation method of network reliability, “To understand the NTT communication network” (NTT Communication Network Study Group, 1994) “Chapter 8 Stable Quality”, pp. 314-329 (Non-Patent Document 1). The reliability evaluation technique disclosed in this document is a technique for supporting accurate judgment in implementing network planning, design, and construction.

  Also, Stanshine JA “Modeling silent failures in telecommunications systems” in Proceeding of RAMS1995, pp.261-264 (Non-Patent Document 2) theoretically calculates the effect of silent failures on network unavailability. A method has been proposed. Silent failure is a failure case in which when a device fails, it cannot be reported from the device to the operation management system, and the actual impact level on the user due to the failure is unknown. .

  In the occurrence of a silent failure, there is a risk of delaying the discovery of the failure, prolonging the service interruption time and having a great influence on the user. Non-Patent Document 2 discloses a method of calculating the influence on the entire system including a corresponding part as an average non-operation time when a silent failure occurs.

  As IP network technology has progressed, networking has been started by actively introducing commercial technologies and general-purpose products, so the development period of related devices, network construction period, and operation verification period tend to be shortened. It is in.

  In addition, a plurality of types of network devices that have been newly developed and have a low operational track record have been introduced, and a plurality of types of network devices have been mixed on the network.

  Therefore, not only the deterioration over time, but also the factors of equipment stoppage and failure (hereinafter, abbreviated as failure), such as initial failure of component parts and incompatibility of setting conditions, are diversified. The possibility of these device failure events occurring after the start of operation of the system has increased.

  However, when a device failure occurs, the device does not have a function to autonomously find and notify the failure event before the device starts operation for all failure events. In some cases, the occurrence of a failure cannot be detected immediately.

  Such failure cases are generally known as silent failures. For specific causes and examples of silent failures, see, for example, Ramana Rao Kompella, Jennifer Yates, Alex C. Snoeren, “Detection and Localization of Network Black Holes” in Proceeding of IEEE INFOCOM2007, pp. 2180-2188. Non-patent document 3).

  As a result, the silent failure is found in two ways: a case where it is discovered by a function provided outside the failure device (see Non-Patent Document 3) and a case where it is found based on an inquiry from a user who encounters the failure. There are cases. The former case can be realized for some devices in the network due to device specifications, cost constraints, etc., but is generally not applicable to all devices or device models.

  Therefore, as in the latter case, the occurrence of a failure is often found by an inquiry from the user, and in that case, there is a high possibility that the discovery of the failure of the device will be delayed and that the user will be greatly affected.

  Hereinafter, in the present specification, only a failure case that directly leads to a service stop / interruption due to the occurrence of a failure in a failure case in which the occurrence of the failure cannot be found immediately is defined as a silent failure.

“To understand the NTT communication network” (NTT Communication Network Study Group, 1994) “Chapter 8 Stable Quality”, pp. 314-329 Stanshine J. A. “Modeling silent failures in telecommunications systems” in Proceeding of RAMS1995, pp.261-264 Ramana Rao Kompella, Jennifer Yates, Alex C. Snoeren, `` Detection and Localization of Network Black Holes '' in Proceeding of IEEE INFOCOM2007, pp.2180-2188.

  The problem to be solved by the present invention is that the occurrence of a failure is not immediately detected and is not notified to the operation management system as in the case of the silent failure described above. For faults, we estimate the impact of these faults on network reliability and evaluate the magnitude of the impact.

  The technology disclosed in Non-Patent Document 1 described above calculates the failure probability of each component device from the device configuration of the network to be evaluated, the failure rate of each component device, and the average recovery time when each component device fails. The sum or product of the average failure probability of each component is calculated as the network failure probability. By comparing the calculated network failure probability with the reference value, the reliability of the network Is to evaluate.

  On the other hand, as described above, when a silent failure occurs, information on the date and time of occurrence of the failure is not obtained, and therefore the time that the failure actually continued (failure time) is unknown. Therefore, it is not possible to accurately evaluate the influence of various failure cases including silent failures on the reliability of the network.

  In addition, the technology disclosed in Non-Patent Document 2 described above is based on the assumption that a silent failure occurs in a redundant configuration location, and the service is not interrupted only by the failure of one device, and the other device The main subjects are failure cases in which service is interrupted by redundant failures. At this time, the silent failure (failure of one device in the redundant part) and the failure occurrence rate of the other device are used as parameters, and the influence of the silent failure on the system unavailability is calculated.

  However, since it is difficult to accurately determine the number of relevant points that cause silent failure before the start of operation of the device or immediately after the occurrence of the failure, the failure rate (failure rate) The instantaneous value at the time of occurrence cannot be calculated accurately.

  Further, since the average value of the failure rate is assumed and changed as a parameter, and the average characteristic is evaluated with respect to the reliability of the system, the influence of each failure event on the user cannot be evaluated.

  In addition, since average values are used as parameters for the time from the occurrence of silent failure to system recovery, the failure event is reported to the user when there are cases with extremely long failure times or instantaneous failures. There is a risk of underestimating or overestimating the impact on the system.

  The purpose of the present invention is to solve the above-mentioned problems. In the network to be evaluated, various failure cases including cases in which the occurrence of a failure such as a silent failure is not immediately discovered and the influence of the failure on the user is unknown. It is to provide a network evaluation system, a network evaluation method, and a program therefor, which can estimate the influence of a user on a user and evaluate the reliability of the network.

  In order to achieve the above object, the present invention adopts a configuration having the following functions.

(1) In the information on the failure history collected from the network, for the device to be evaluated, information on the failed device model and function, failure time from failure occurrence to recovery, number of users affected by failure, and discovery of failure Information related to the type and the number of inquiries from the user regarding each failure is extracted and stored.

(2) The extracted failure information is classified into a normal failure and a silent failure based on the failure discovery type.
If the fault discovery type is an alert from a faulty device, it is classified as a normal fault. In addition, if the failure discovery type is other than that (discovery by a failure discovery function outside the device or discovery by a report from the user), it is classified as a silent failure.

(3) The influence of each failure case on the user is calculated based on the failure duration of each failure case (failure time) and the number of affected users when the failure occurs.
a) For a normal failure, since the failure time is longer and the influence degree of the failure is larger as the number of affected users is larger, the influence degree of the failure is calculated from the information of the failure time and the number of affected users due to the failure. Here, the failure time is the time from failure occurrence to recovery, and the number of affected users is the total number of users who have suffered service interruption due to failure occurrence.

b) On the other hand, for silent failure, since the date and time of occurrence of the failure is unclear, the date and time of failure discovery is recorded as the date and time of failure occurrence, and the time from failure discovery to recovery is calculated as the failure time. Therefore, it is necessary to estimate the degree of influence on the user by correcting the calculated degree of influence of the failure in consideration of the time from occurrence of the failure to recovery.

  Therefore, in the case of a silent failure, the degree of influence due to the occurrence of the silent failure is corrected by using information on the number of reports (number of inquiries) related to the failure from the user affected by the failure. For normal failure cases, the number of user reports due to failure tends to increase as the influence of failure increases. Therefore, the correlation between the degree of influence of the failure and the number of reports from the user is clarified using the failure information of the normal failure case, and the relational expression is estimated. For the silent failure, the degree of influence of the failure is corrected using the estimated relational expression and the number of cases reported from the user regarding each failure case.

(4) The degree of influence of the failure on the user is evaluated together with the uncorrected case and the corrected case.
The reliability of the network is evaluated by comparing the reliability reference value desired by the network operation manager set in advance with the calculated and corrected failure influence level.

(5) Finally, the obtained reliability evaluation result is displayed.

  The present invention is characterized by having the following means as a specific configuration for realizing the above functions.

(A) Means for executing an input procedure (procedure 0) for inputting network failure information, evaluation object information, and evaluation reference values.

(B) A means for executing a failure case classification procedure (procedure 1) for classifying network failure information into a normal failure case and a silent failure case based on the failure discovery type.

(C) Means for executing an influence degree calculation procedure (procedure 2) for calculating an influence degree of a failure for each failure case from information on a failure time in the network failure information and the number of users affected by the failure.

(D) For normal failure cases, a relational expression estimation procedure for estimating the relationship between the failure impact level and the number of user reports based on the calculated failure impact level and the number of user reports on the failure ( Means for performing step 3).

(E) Means for executing an influence correction procedure (procedure 4) for correcting the influence degree of a failure by using the number of cases reported from the user and the estimated relational expression for silent failure.

(F) Means for executing the reliability evaluation procedure (procedure 5) for evaluating the reliability of the network by combining the uncorrected failure case and the corrected failure case with respect to the degree of influence of the failure.

(G) Means for executing an evaluation result display procedure (procedure 6) for displaying an evaluation result

In this invention, it has the following effects by employ | adopting the said structure.
(A) Since the influence degree of failure is estimated from the failure information of the network equipment and the information of the number of reports from the user regarding the failure, the devices and causes that are highly effective in reducing the number of reports from the user are clarified. It is possible.

(B) For cases where the degree of influence on the user is unknown, such as silent failure, it is possible to realize a more safe evaluation by estimating the degree of influence.

(C) By evaluating the degree of influence on the reliability of the network due to the occurrence of the failure using the evaluation reference value, the operation of the network having high reliability or the reliability desired by the network operation manager can be performed. It becomes possible.

(D) In addition, by comparing the degree of influence of each failure case, select a failure case whose cause should be preferentially analyzed and a device that should strengthen monitoring of the operation status based on quantitative judgment. Is possible.

(E) In addition, by quantitatively evaluating the reliability of the network, it is possible to determine in advance a quick recovery action in the event of a failure and to take measures against the failure to reduce the impact on the user during network operation management. An indicator for making a decision can be given.

It is a figure which shows an example of the processing flow (processing procedure) of the network evaluation system which concerns on this invention. It is a figure which shows one Embodiment of the structure of the network evaluation system which concerns on this invention, and the flow of its processing. It is a figure which shows an example of network failure information. It is an example of the network failure information after implementation of classification processing. It is an example of failure data after execution of calculation processing of the influence degree of each failure case. It is an example of the failure data after execution of the correction process of the influence degree of each failure case. It is an example of the network used as the implementation object of this invention. It is a figure which shows the scatter diagram of the influence degree of a failure and the number of reports from a user, and a linear regression line in embodiment of this invention. It is an example of the flowchart which implements reliability evaluation in the Example of this invention.

(Outline of the present invention)
First, a network configuration targeted by the present invention will be described. FIG. 7 is a diagram showing a network configuration targeted by the present invention, and includes a control network 100, a core network 200, an access network 300, and an end user 400. It is assumed that the end user 400 uses the Internet access and various IP (Internet protocol) services.

  A failure in a configuration where the network element is not made redundant leads to service interruption, so it is highly likely that the service usage of the user will be affected. As a simple example, for example, the access network 300 is similar as shown in the figure. Or a component in the facility (device A 301, device B 302a, device B 302b, device C 303a, device C 303b, device C 303c, device C 303d) can be described as a tree-shaped model. .

  Although a failure in a redundant configuration location such as the core network 200 is unlikely to lead to service interruption, service interruption due to switching or service interruption due to switching function failure may occur.

  Therefore, in order to evaluate the reliability of the network, it is necessary to carry out detailed analysis of all the failure cases that caused service interruption. When data is collected, it is not always easy to classify individual failure locations and factors and perform comparative evaluation on all combinations.

  Therefore, the present invention takes into account the above points, and in order to grasp the failure characteristics and perform a quantitative evaluation after arranging all the failure cases side by side, the failure device model input and recorded in the network evaluation system Function name, failure occurrence date / time (when the failure occurrence date / time is not clear, failure discovery date / time), failure recovery date / time, failure time (failure duration = failure recovery date / time / failure occurrence date / time), number of affected users (fault affects Range (people)), failure discovery type, and the number of reports from users, comprehensively evaluate network reliability by estimating the degree of failure impact (= failure time x number of affected users) It is.

(Embodiment)
Hereinafter, embodiments of a network evaluation system according to the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a diagram showing an example of a processing flow (processing procedure) of the network evaluation system according to the present invention.

  As shown in the figure, the processing procedure in the present invention is classified into a normal failure case and a silent failure case, an input procedure (procedure 0) for inputting network failure information, evaluation target information, and evaluation reference value. Failure case classification procedure (procedure 1), impact degree calculation procedure (procedure 2) for calculating the impact degree of each failure case from the information of failure time and the number of affected users, and the relational expression between the failure degree and the number of reports from users Estimating relational expression estimation procedure (procedure 3), influence degree correcting procedure (procedure 4) for correcting the degree of influence of silent failure using the estimated relational expression, and reliability evaluation for estimating and evaluating network reliability It consists of a procedure (procedure 5) and a procedure (procedure 6) for displaying the evaluation result.

  FIG. 2 is a diagram showing an embodiment of the configuration of the network evaluation system according to the present invention and the processing flow thereof, and also describes (procedure 0) to (procedure 6) in FIG.

  In the figure, 1 is network failure information input to the network evaluation system in (procedure 0), 2 is input means for inputting evaluation object information and evaluation reference values, 10 is a network evaluation system according to the present invention, and 11 is A receiving unit (receiving unit) that receives network failure information, 12 is a network failure information DB (database) that stores network failure information, 13a to 13d are storage units (storage unit), and 14 is a failure that executes (procedure 1). A case classification unit (failure case classification means), 15 an influence calculation unit (impact calculation means) for executing (procedure 2), 16 a relational expression estimation unit (relational expression estimation means) for executing (procedure 3), Reference numeral 17 denotes an influence degree correcting unit (influence degree correcting means) for executing (Procedure 4), 18 denotes a reliability evaluating part (reliability evaluating means) for executing (Procedure 5), and 3 denotes (Procedure 6). A display unit for executing (display means), display the network qualification evaluated at (step 5).

  The timing of execution of the processing of the system in the present invention includes a case where a fixed period is set in advance and a case where an arbitrary period is input and executed, and in any case, the embodiment is the same. Therefore, in the following, the present invention is repeatedly implemented for each fixed period, and a case where the implementation timing is set in advance will be described as an example.

Hereinafter, processing in each procedure will be described in detail with reference to the drawings.
(Procedure 0) <Input of network failure information, evaluation target information, evaluation reference value>
This procedure (procedure 0) is automatically executed in this network evaluation system, or is executed in advance by the network operation manager, and is described separately from the subsequent procedures (procedure 1 to procedure 6). .

  The generated network failure information 1 is taken into the network evaluation system, and the evaluation target information is input from the input means 2 by the network operation manager from the outside. The network failure information 1 is stored in a database (DB) (not shown) outside the network evaluation system 10 and is input to the receiving unit 11 of the network evaluation system 10.

  As shown in FIG. 3, the network failure information 1 includes a failure facility, that is, a failure device (failed device model), a failure occurrence date / time (a failure discovery date / time when the failure occurrence date / time is not clear), a failure recovery date / time, The information includes information on the failure time (minutes), the number of affected users (people), the failure discovery type, the number of cases reported from the user for each failure case, and the failure location / cause.

  Here, the failure time (minutes) is the time during which communication is interrupted due to a failure, is defined by the difference (in minutes in this example) from the failure occurrence date and time to the failure recovery date and time, and is input by the network operation manager . However, as for silent failure, since the date and time when the failure actually occurred is not clear, the date and time when the failure was found is usually input. In addition, the number of affected users (persons) is usually defined by the number of failed devices accommodated. The network failure information 1 input to the receiving unit 11 is stored in the network failure DB 12.

  In addition, the network operation manager inputs information related to equipment (generic name including equipment / devices) to be subjected to reliability evaluation as evaluation target information to the network evaluation system 10 via the input unit 2. . The evaluation target information input from the input unit 2 is temporarily stored in the storage unit 13c.

  With reference to the evaluation object information stored in the storage unit 13c, the failure information of the equipment to be evaluated is set as the evaluation object for the network failure information 1 input in the network failure information DB 12. Other failure information is excluded as information that is not subject to subsequent procedures.

(Procedure 1) <Process for classifying network failure information into normal failure cases and silent failure cases>
In (Procedure 1), the network failure information 1 set as the evaluation target in (Procedure 0) is classified into two types of normal failure and silent failure by the failure case classification unit 14.

  Normally, when a failure occurs, the operation system side is notified of the discovery of the failure by an alert from the failed device. Therefore, if the fault discovery type entered in the fault information is an alert from the faulty device, it is classified as a normal fault.

  In addition, if the failure discovery type is other than that (discovery by a failure discovery function outside the device or discovery by a report from the user), it is classified as a silent failure.

  FIG. 4 is a diagram illustrating an example of network failure information after performing the classification process, and a column of failure case classification indicating whether a normal (general) failure or a silent failure is added to the network failure information illustrated in FIG. ing.

(Procedure 2) <Process for calculating the degree of influence of each failure case from information on the failure time in network failure information and the number of users affected by the failure>
In (Procedure 2), for the failure cases for which the classification process has been performed in (Procedure 1), the influence calculation unit 15 shows the influence of each failure case on the user, the failure time of each failure case, and the effect at the time of failure occurrence. Based on the number of users, the impact level of the failure is calculated. For silent failure, the time from failure detection to recovery is calculated as the failure time. The failure impact level is calculated by defining the product of the failure time and the number of affected users as the failure impact level.

  Numbers i are assigned to failure cases in order of the time and date of failure occurrence. The failure time (minutes) of each failure case is di, the number of users affected by the failure (people) is ri, and the failure influence degree (minutes × people) hi for the i-th failure case is the failure time (minutes) di and the affected users. The product of the number (person) ri, that is, di × ri is calculated (hi = di × ri).

  FIG. 5 is a diagram illustrating an example of the calculation result of the failure influence degree when the product of the failure time and the number of affected users is defined as the failure influence degree. The network failure information after the classification process illustrated in FIG. A column of the degree of influence of failure is added.

  Usually, in a network operation management system, failure time information is collected in units of minutes, and information on the number of affected users is collected in units of people. Therefore, the degree of influence of failure is calculated in units of [minutes / people]. The calculated failure influence degree is stored in the storage unit 13b.

(Procedure 3) <Process for estimating the relational expression between the degree of influence of failure and the number of reports from the user>
In (Procedure 3), the relational expression estimation unit 16 estimates the relationship between the degree of influence of the failure and the number of reports for the failure from the user.

  For only normal failure cases, the impact level of each failure calculated in (Procedure 2) and stored in the storage unit 13b and the number of reports from the user for the corresponding failure (inquiries and complaints regarding failures) The number of occurrences).

  At this time, since the number of reports from the user increases as the degree of influence of the failure increases, the increasing tendency of the number of reports from the user with respect to the degree of influence of the failure may be estimated.

  Here, let ni be the number of reports from the user for the i-th failure case. The relationship of the number of reports ni to the failure influence level hi is expressed using a proportional relationship: ni = a × hi, and the value of the coefficient a is estimated.

  Specifically, the point (hi, ni) consisting of the failure impact level hi and the number of reported cases ni is plotted as a scatter diagram on the coordinates with the horizontal axis representing the failure impact level and the vertical axis representing the number of reported cases. A linear regression line may be applied using a method such as multiplication to estimate the value of the coefficient a. Here, the coefficient a indicates the slope of the straight line.

  FIG. 8 is a scatter diagram in which the horizontal axis represents the failure impact level and the vertical axis represents the number of reported cases, and the linear regression line for the points on the scatter diagram, regarding the failure degree and the number of reports from the user. FIG.

  FIG. 6 is a diagram illustrating an example of failure data after performing the correction process of the influence degree of each failure case, and shows that the influence degree of the failure in the failure data illustrated in FIG. 5 is partially corrected. .

(Procedure 4) <Process for correcting the influence level of failure for silent failure>
In (Procedure 4), the influence correction unit 17 performs correction processing on the failure influence degree of the silent failure calculated in (Procedure 2) and stored in the storage unit 13b.

  For the silent failure, since the failure discovery date is recorded as the failure occurrence date, the failure time is calculated to be smaller than the actual failure duration, and as a result, the failure influence degree is also calculated to be small. Therefore, it is necessary to correct the failure time or the influence degree of the failure. Here, an example of performing the correction of the failure influence level will be described.

  For the failure information classified as silent failure in (Procedure 1), the failure impact level hi of each failure case calculated in (Procedure 2) and the number of reports ni from the user included in the network failure information The combination (hi, ni) is extracted.

Next, the degree of influence of the failure is corrected using an estimation formula that employs the coefficient a estimated in the above (procedure 3). Since it is expressed as hi = ni / a by the estimation formula, the influence level of the failure is corrected by the following formula.
hi = max (ni / a, hi)
Here, “max (x, y)” is an arithmetic symbol that represents the smaller value of x and y.

(Procedure 5) <Process for Evaluating Network Reliability Using Failure Impact Level>
In (Procedure 5), the reliability evaluation unit 18 evaluates the degree of influence of the failure. Evaluation is performed by combining the failure cases in which the influence value is corrected in (Procedure 4) and the failure cases calculated in (Procedure 2) (the correction process is not performed).

  There are two types of evaluation methods: a method for evaluating the impact level for each failure case (evaluation method 1 and evaluation method 2) and a method for evaluating the impact level for each device model / function (evaluation method 3). There is a way.

(Evaluation method 1)
The evaluation method 1 is an evaluation method by comparing the degree of influence hi of each failure case with the evaluation reference value input from the input means 2, and is evaluated as follows, for example.

(A) hi ≧ evaluation reference value If a failure case satisfying this relationship exists, it is evaluated that the reliability of the network is low.
(B) hi <evaluation reference value When this relationship is satisfied, it is evaluated that the reliability of the network is high.

  As described above, the network operation manager sets an arbitrary value in advance using the input unit 2 as described above. For example, a value of about 100,000 is set (for example, a failure time of 10 minutes, If the number of affected users is 10,000, the value of the degree of influence of the failure is 100,000).

(Evaluation method 2)
The evaluation method 2 is an evaluation method that uses the ratio of the influence degree hi of each failure case to the influence degree of all the failure cases, and satisfies, for example, “hi / (Σhi) ≧ evaluation reference value” ( The i-th) failure case is evaluated as a “failure case having a great influence on a decrease in the reliability of the network”.

  Here, Σhi represents the sum of the influence levels for all the failure cases. The evaluation reference value is an arbitrary value set by the network operation manager from the input means 2, and is set to about 0.1, for example. At this time, the degree of influence of the corresponding failure case corresponds to 10% of the total.

(Evaluation method 3)
The evaluation method 3 is a method of comparing and evaluating the degree of influence of failure for each device model. As a comparison method, there are methods such as comparing the total value of the degree of influence of failure for each apparatus model, or comparing the average value of the degree of influence of failure. The device model having the largest total failure influence value or the average failure influence average value is evaluated as “the device model having the greatest influence on the decrease in network reliability”. In this case, the evaluation can be performed without setting the evaluation reference value.

(Procedure 6) <Display of evaluation results>
In (Procedure 6), the evaluation result output in (Procedure 5) is displayed on the display unit 3.
The evaluation results displayed on the display unit 3 include the items shown in the following (a) to (e), and any of these items or a combination of some items may be used.

(A) Whether the reliability of the network is high or low (whether it is within the evaluation reference value) is displayed on the display unit 3.
B) In the case of (a) in the evaluation method (1), “x” is displayed.
B) In the case of (b) of the evaluation method (1), “◯” is displayed.

(B) The detailed information of the failure case having a large influence on the reliability reduction of the network is displayed on the display unit 3.
A) In the case of (a) in the evaluation method (1), “device model, failure occurrence date / time, failure time, number of affected users” of failure cases satisfying (a) is displayed.

(C) In the case of (a) in the evaluation method (1), the display unit 3 displays whether the network reliability degradation factor is a normal failure or a silent failure.
B) In the case of a silent failure, “Silent failure” is displayed.
B) “Normal failure” is displayed for normal failure cases.

(D) In the case of the evaluation method (2), failure cases that have a large influence on reliability degradation are displayed on the display unit 3.

(E) In the case of the evaluation method (3), a device model having a large influence on the reliability reduction is displayed on the display unit 3.

  FIG. 9 is a flowchart showing a procedure for performing reliability evaluation performed in the network evaluation system 10 according to the present invention. Step S1 is a failure performed by the failure case classification unit (failure case classification means) 14 of FIG. Corresponding to the case classification process (procedure 1 in FIG. 1), step S2 corresponds to the influence degree calculating process (procedure 2 in FIG. 1) performed by the influence degree calculating unit (influence degree calculating means) 15 in FIG. Step S3 corresponds to the relational expression estimating process (procedure 3 in FIG. 1) performed by the relational expression estimating unit (relational expression estimating means) 16 in FIG. 2, and step S4 is the influence degree correcting part (influence degree in FIG. 2). Corresponding to the influence correction process (procedure 4 in FIG. 1) performed by the correction unit) 17, step S5 is a reliability evaluation process (FIG. 1) performed by the reliability evaluation unit (reliability evaluation unit) 18 of FIG. Hands in Which corresponds to 5). Detailed processing in each step is as described above. When it is determined in the reliability evaluation process in step S5 whether or not the failure influence level is within the evaluation reference value, if the failure influence level is within the evaluation reference value (step S5: Y), the process ends. If the failure influence level is not within the evaluation reference value (step S5: N), the process returns to step S1.

  Note that the processing and functions performed by each unit (each unit) of the network evaluation system shown in FIG. 2 are performed using each unit (each unit) using hardware resources such as a CPU and a memory built in the computer constituting the network evaluation system. This is realized by executing a program corresponding to the processing executed in (Means). Further, the program can be distributed to the market via a recording medium such as FD, CD-ROM, DVD, or a network such as the Internet.

1: Network failure information 2: Input means (input unit)
3: Display unit for displaying the determination result (display means)
10: Network evaluation system 11: Receiver (reception means)
12: Network failure information DB (database)
13a to 13d: Storage unit 14: Failure case classification unit (failure case classification means)
15: Influence calculation unit (impact calculation means)
16: Relational expression estimating unit (relational expression estimating means)
17: Influence degree correcting unit (influence degree correcting means)
18: Reliability evaluation section (reliability evaluation means)
19: Storage unit for storing determination target information 100: Control network 200: Core network 300: Access network 301: Apparatus A
302a to 302b: Device B
303a to 303d: Device C
400: End user

Claims (9)

A network evaluation system for evaluating a network having a plurality of facilities as a component,
Means for inputting information relating to network failure information including failure time for each facility and the number of affected users;
Means for inputting evaluation object information on equipment to be subjected to reliability evaluation;
Means for inputting an evaluation standard value for reliability evaluation;
Storage means for storing the inputted network failure information;
Storage means for storing the input evaluation object information;
Storage means for storing the inputted evaluation reference value;
A failure case classification means for classifying the failure cases included in the network failure information into normal failure cases and failure cases whose degree of influence due to failure is not clear;
An influence degree calculating means for calculating the influence degree of the failure of the equipment to be evaluated from the failure time of the network failure information and the number of affected users;
In the failure cases classified by the failure case classification means, an influence correction means for correcting the influence degree for cases where the influence degree due to the failure is not clear, and
Reliability evaluation means for evaluating the reliability of the network based on the influence degree of the failure calculated by the influence degree calculation means and the influence degree of the failure corrected by the influence degree correction means;
Display means for displaying an evaluation result evaluated by the reliability evaluation means;
A network evaluation system comprising: The network evaluation system according to claim 1,
The network evaluation system according to claim 1, wherein the failure case classification means is means for classifying failure cases by identifying the necessity of correction based on information relating to a failure discovery type. The network evaluation system according to claim 1 or 2,
The influence degree correcting means estimates the relationship between the degree of influence of a failure relating to a normal failure case and the number of reports from the user relating to each failure case, and based on the estimation result, the degree of failure for an incorrect failure case is determined. A network evaluation system which is a means for correcting an influence degree. The network evaluation system according to claim 3,
The estimation of the relationship between the degree of influence of a failure relating to the normal failure case and the number of cases reported from the user regarding each failure case is a scatter diagram of the degree of influence of the failure relating to the normal case and the number of cases reported from the user relating to each case. And a network evaluation system using a linear regression line for points on the scatter diagram. The network evaluation system according to claim 3 or 4, wherein:
The influence degree evaluation means compares the influence degree of the failure calculated by the influence degree calculation means and the influence degree of the failure corrected by the influence degree correction means with the evaluation reference value, thereby improving the reliability of the network. A network evaluation system characterized by being a means for evaluation. The network evaluation system according to claim 3 or 4, wherein:
The influence degree evaluation means includes the ratio of the influence degree of the failure to the total of the influence degree of the failure calculated by the influence degree calculation means and the influence degree of the failure corrected by the influence degree correction means, and the evaluation. A network evaluation system which is a means for evaluating the reliability of a network by comparing with a reference value. The network evaluation system according to claim 3 or 4, wherein:
The network evaluation system is characterized in that the influence degree evaluation means is a means for comparing the total or average value of the influence degree of failure of each device type and evaluating the reliability of the network based on the comparison result.   The program for functioning a computer as each means in the network evaluation system in any one of Claim 1 to 7. A network evaluation method using a computer for evaluating a network having a plurality of facilities as a component,
A procedure for inputting information related to network failure information including failure time and the number of affected users for each facility,
A procedure for entering the evaluation object information on the equipment to be subjected to reliability evaluation;
A procedure for entering the evaluation standard value for reliability evaluation,
A storage procedure for storing the inputted network failure information;
A storage procedure for storing the input evaluation target information;
A storage procedure for storing the inputted evaluation reference value;
A failure case classification procedure for classifying failure cases included in the network failure information into normal failure cases and failure cases in which the degree of influence due to the failure is not clear;
An impact calculation procedure for calculating the impact of the failure of the equipment to be evaluated from the failure time of the network failure information and the number of affected users;
In the failure cases classified by the failure case classification procedure, an impact correction procedure for correcting the impact for cases where the impact due to the failure is not clear, and
A reliability evaluation procedure for evaluating the reliability of the network based on the impact level of the failure calculated in the impact level calculation procedure and the impact level of the fault corrected in the impact level correction procedure;
A display procedure for displaying the evaluation result evaluated in the reliability evaluation procedure;
A network evaluation method comprising:

Images