JP2009071684A - Network test planning device, network test planning method, network test planning program and network monitoring system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To permit simultaneous execution of a plurality of testings in parallel under a restriction that the loading conditions of testing terminals are constant. <P>SOLUTION: The network test planning device works out the plan of communication quality test executed with respect to all combinations among (n) sets of bases in total. The device is provided with an origination side base determining means for determining the base of origination side of the test in accordance with a parameter (i) which specifies the order of parallel tests, a parameter (j) which specifies either one test in the parallel tests and a parameter (n), a destination side base determining means for determining the base of destination side of the test in accordance with the parameters (i), (j), (n), a test adding means for adding a pair of determined origination side base and destination side base to a test planning list, a parameter changing means in parallel test for sequentially changing the parameter (j) among a plurality of kinds of values during a period wherein the parameter (i) keeps the same value and a parameter changing means in parallel test for sequentially changing the parameter (i) among a plurality of kinds of values. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a network test planning device, a network test planning method, a network test planning program, and a network monitoring system. For example, voice quality (hereinafter referred to as voice quality) in voice communication over an IP network to which VoIP (Voice over Internet Protocol) is applied. It can be applied to regular monitoring of sound quality.

  In network monitoring, as described in Japanese Patent Application Laid-Open No. H10-228707, there is one that monitors the status of each device (for example, a packet switch) and collects the status of each device.

In VoIP, the voice quality is ITU-T G. Evaluation is performed according to the evaluation value (R value) in the evaluation model recommended in 107 (see Non-Patent Document 1).
JP-A-7-162420 Oizumi Osamu, “All About VoIP with Illustrations”, Nihon Jitsugyo Publishing Co., Ltd., 2003, pp. 353-354

  However, the monitoring method described in Patent Document 1 is not for a communication system using an IP network. In the network targeted by the monitoring system described in Patent Document 1, there is almost one route between the packet switches. In an IP network, a specific route changes depending on a network condition at the time of outgoing call, and even a packet related to the same call changes depending on the packet. Therefore, the monitoring method described in Patent Document 1 that is not intended for a communication system using an IP network cannot be applied to a communication system using an IP network.

  ITU-T G. The R value defined in 107 evaluates a signal that has passed through a route such as a transmitter-terminal-IP network-terminal-receiver, in other words, represents a sound quality between two points. A sound quality evaluation between two points cannot be a sound quality evaluation of an IP network.

  In order to evaluate the sound quality of an IP network, a sound quality evaluation between two points must be performed on a combination of a plurality of two points. If the voice quality of a combination of a plurality of two points is obtained while traffic is generated, it is not possible to obtain a measurement value that correctly reflects the influence of network traffic that changes daily or from moment to moment. Therefore, it is preferable to obtain the voice quality of a combination between a plurality of two points by a test. However, if a test for evaluating sound quality is performed on a combination between two points without any plan, for example, a certain device (for example, IP-PBX) on the network is overloaded, May adversely affect communications.

  Therefore, a network test planning device, a network test planning method, and a network test planning method that can obtain a quality evaluation value that correctly reflects the traffic of the IP network that changes daily or from time to time without causing an overloaded device due to the test, A network test planning program is desired, and a network monitoring system to which such a network test planning device, a network test planning method, and a network test planning program are applied is also desired.

  The first aspect of the present invention is a network test planning apparatus for formulating a communication quality test plan to be executed for all combinations among all n sites where test terminals are provided, and (1) a sequence of parallel tests Of the test in accordance with a parameter i defining one of a plurality of tests in a parallel test, and a parameter n taking n when n is an even number and n + 1 when n is an odd number. (2) Calling site determination means for determining the receiving site of the test according to parameter i, parameter j, and parameter n, and (3) Determination of the calling site And a test addition means for adding a pair of a caller base and a callee base determined by the callee base determination means to the test plan list, and (4) parameter i takes the same value. Parameter change in parallel test to change parameter j sequentially in n / 2 values when n is even and (n + 1) / 2 values when n is odd And (5) parallel test parameter changing means for sequentially changing the parameter i among (n-1) types of values.

  A second aspect of the present invention is a network test plan method for formulating a communication quality test plan to be executed for all combinations among all n sites where test terminals are provided, comprising: A side base determination means, a test addition means, a parallel test parameter change means, and a parallel test parameter change means. (1) The transmission side base determination means includes a parameter i that defines the order of parallel tests, (2) The incoming call is determined according to a parameter j that defines one of a plurality of tests and a parameter n that takes n when n is an even number and takes n + 1 when n is an odd number. The side base determining means determines the base of the test receiving side according to the parameter i, the parameter j, and the parameter n. (3) The test adding means determines the transmitting side base And the pair of the caller base and the callee base determined by the callee base determination means are added to the test plan list. (4) In the parallel test parameter changing means, the parameter i has the same value. In the period, the parameter j is sequentially changed in n / 2 types of values when n is an even number, and (n + 1) / 2 types of values when n is an odd number. The test parameter changing means is characterized in that the parameter i is sequentially changed among (n-1) kinds of values.

  According to a third aspect of the present invention, there is provided a network test plan program for formulating a communication quality test plan to be executed for all combinations among all n sites where test terminals are provided, and (1) a parallel test Of the test, a parameter j that defines one of a plurality of tests in the parallel test, and a parameter n that takes n when n is even and takes n + 1 when n is odd. A transmitting site determining means for determining a transmitting site, (2) a receiving site determining means for determining a test receiving site according to parameter i, parameter j, and parameter n, and (3) the transmitting side. A test addition means for adding a pair of a transmission-side base and a reception-side base determined by the base determination means and the reception-side base determination means to the test plan list; In the period when i takes the same value, the parameter j is set to n / 2 types of values when n is an even number, and (n + 1) / 2 types of values when n is an odd number, A parallel test parameter changing unit that sequentially changes, and (5) a parameter i that functions as a parallel test parameter changing unit that sequentially changes the parameter i among (n-1) types of values.

  A network monitoring system according to a fourth aspect of the present invention includes: (1) a test terminal provided at each site; (2) a network test plan apparatus according to the first aspect of the present invention; and (3) a test in the network test plan apparatus. A test instruction means for instructing test communication from the test terminal at each of the transmitting bases to the test terminal at the receiving base in a plurality of tests in the parallel test according to the plan list; and (4) a test terminal at the receiving base. And information collecting means for collecting and storing information related to communication quality.

  According to the present invention, it is possible to simultaneously perform a plurality of tests while satisfying the constraint that the load state of the test terminal is constant.

(A) Main Embodiment Hereinafter, an embodiment of a network test planning device, a network test planning method, a network test planning program, and a network monitoring system according to the present invention will be described in detail with reference to the drawings. In this embodiment, a network of a VoIP communication system (for example, an IP telephone system) is targeted for monitoring.

(A-1) Configuration of Embodiment FIG. 1 is a block diagram showing a configuration of a network monitoring system according to this embodiment.

  1, the network monitoring system 10 according to the embodiment includes a test instruction terminal 11, a voice information collection server 12, a plurality (n pieces in FIG. 1) of pseudo call processing terminals 13-1 to 13-n, and a database 14. ing. Note that a plurality (m pieces in FIG. 1) of IP-PBXs 15-1 to 15-m exist in the monitored IP network.

  The test instruction terminal 11 is a terminal that can be operated by an operator equipped with a Web browser and the like, and, for example, bears the following functions (a) to (c). (A) Setting of URI (SIP-URI) of pseudo call processing terminals 13-1 to 13-n, (b) Collection of test instructions and test results of matrix test and one call test, (c) Test results It is responsible for reference.

  Here, the “matrix test” is a test that is performed for periodic fixed-point observation, and the “1 call test” is a test that is performed between incoming and outgoing pseudo call processing terminals designated by interruption. This embodiment has features relating to the matrix test, and the following description will be given mainly from the aspect of the matrix test. In each test, a pair of pseudo call processing terminals is a test target, and in the matrix test, the pair of pseudo call processing terminals change sequentially, and in one call test, the pair of pseudo call processing terminals is designated. There is only one pair.

  The voice information collection server 12 basically collects test instructions from the test target pseudo call processing terminal 13-N (N is 1 to n) and collects test results from the test target pseudo call processing terminal. The test results are stored in the database 14. The audio information collection server 12 is realized by, for example, a CPU and a program executed by the CPU except for the communication configuration. Functionally, the voice information collection server 12 includes a test planning unit 12A, a test instruction unit 12B, and a test result receiving unit 12C. Have. The test plan unit 12A corresponds to the network test plan apparatus of the embodiment.

  The test plan unit 12A is for formulating a test plan for a matrix test, and is a major feature of this embodiment. Details of the test plan unit 12A will be clarified in the description of the operation.

  The test instruction unit 12B instructs the pseudo call processing terminal 13-N to perform a test while performing time control in accordance with the test plan for the matrix test formulated by the test plan unit 12A. Note that the test instruction unit 12B appropriately gives instructions for the one-call test.

  The test result receiving unit 12 </ b> C receives the test result from the pseudo call processing terminal 13 -N and stores the result in the database 14.

  Each pseudo call processing terminal 13-N is arranged in a different network base (for example, Tokyo, Osaka, Fukuoka, etc.). Each pseudo call processing terminal 13-N behaves as a terminal (IP telephone terminal) of a VoIP communication system (for example, IP telephone system) in accordance with a test instruction from the test instruction unit 12B and executes a test. Each pseudo call processing terminal 13-N records voice information (R value, packet delay, number of transmitted / received packets, etc.) in the IP phone, and the pseudo call processing terminal 13-N performs a test at the end of the IP phone. The result is transmitted to the test result receiving unit 12C.

  The database 14 stores the test result given from the test result receiving unit 12C. The database 14 may be integrated with the audio information collection server 12 or may be separate. Although FIG. 1 shows that the test instruction terminal 11 can execute the reference of the test result of the database 14 via the voice information collection server 12, the test instruction terminal 11 passes through the voice information collection server 12. Alternatively, the test results in the database 14 may be directly referred to. Further, FIG. 1 shows that the test result obtained by the pseudo call processing terminal 13-N is given to the database 14 via the voice information collection server 12, but the pseudo call processing terminal 13-N The test result may be directly given to the database 14 without going through the collection server 12.

  As described above, there are a plurality of IP-PBXs 15-1 to 15-m in the IP network, and each IP-PBX 15-M (M is 1 to m) accommodates the pseudo call processing terminal 13 accommodated therein. It also performs call control of pseudo calls related to -N.

(A-2) Operation of Embodiment Next, the operation of the network monitoring system according to the embodiment, particularly the planning operation of the matrix test will be described.

  Network traffic is changed day by day or from moment to moment, and even if there is no problem in sound quality at the time of measurement (test), it does not cause a permanent problem. Therefore, in order to maintain a certain quality in the VoIP communication system, periodic and fixed-point observations are required at each base (location) considering the network configuration. That is, it is necessary to install observation points (pseudo call processing terminals 13-1 to 13-n) at locations (bases) having different network configurations and monitor the results. Since the connection between the bases is an interconnection at n points, there are n × (n−1) number.

  For fixed-point observation, tests under the same conditions are required. For this reason, the load status of each pseudo call for performing a pseudo voice call by fixed point observation must be the same. In other words, the pseudo call processing terminals 13-1 to 13-n must not simultaneously make and receive a plurality of IP phones. In the matrix test plan, it is necessary to plan so that the pseudo call processing terminals 13-1 to 13-n do not make and receive IP telephones at the same time.

  The test plan by the test plan unit 12A of this embodiment is such that each pseudo call processing terminal 13-1 to 13-n does not make and receive an IP phone at the same time even when the matrix test and the one-call test are mixed. It is a method to plan in the next. The one-call test is performed for the purpose of measuring whether or not a quality problem in an IP phone has been found and solved, and may be a one-way test. In this test, each site makes a voice call to each other. However, since it is necessary to change the transmitting side and the receiving side, two tests are performed in which the transmitting side and the receiving side are interchanged even in the same two sites. To do.

  In the test plan unit 12A, the test order of the matrix test is determined. The network test plan program P shown in FIG. 2 is a program for calculating the test order of the matrix test that constitutes the main part of the test plan unit 12A, and is described in the object-oriented language Java (registered trademark).

  2 in the program P of FIG. 2 indicates the number of pseudo call processing terminals (fixed-point observation points), and the operator inputs using the test instruction terminal 11 or the operator uses the test instruction terminal 11. The total number of terminals is obtained from the information of the input pseudo call processing terminals 13-1 to 13-n.

  The processing and functions of each line in the network test plan program P are as follows.

  First line: An initial value of the number (n) of pseudo call processing terminals is set. In the case of the example of FIG. 2, eight are set as the number of pseudo call processing terminals.

  Second line: An initial value n of the number of pseudo call processing terminals is stored as a parameter orgN. The parameter orgN is used for condition determination on the 13th line described later.

  Third line: n is increased by 1 only when the number n of pseudo call processing terminals is an odd number. “%” Represents modulo, and the judgment formula in the first half means whether n is an even number or not. Then, n is maintained when n is an even number, and n is increased by 1 only when n is an odd number.

  Fourth line: A test plan list testList is created.

  5th and 20th lines: While changing the parameter i from 0 to 1, the process from the 6th line to the 19th line is repeated, and when i reaches n−1, the process from the 6th line to the 19th line is not executed. It stipulates. It is specified that the number of times of formulation is (n-1) times if the number of formulation of the test sequence of two times by simply replacing the originating UAC and the terminating UAS is counted as one.

  6th line: A list object sameExecutableList1 of the first half test in one formulation is generated.

  7th line: A list object nameExecutableList2 of the latter half test in one formulation is generated.

  8th line: The list object sameExecutableList1 of the first half test is set (added) to the test plan list testList.

  9th line: The list object nameExecutableList2 of the second half test is set (added) to the test plan list testList. By the processes on the 8th and 9th lines, the test plan list testList is set (added) with two test sequences in which the originating UAC and the terminating UAS are replaced.

  10th and 19th lines: While changing the parameter j from 0 to 1, the process from the 11th line to the 18th line is repeated. When j reaches n / 2, the process from the 11th line to the 18th line is not executed. It stipulates. At the timing of the first half test or the second half test, it is possible to simultaneously test a plurality of sets (n / 2 sets) in which the transmission side and the reception side are different. In this embodiment, the test is executed on a plurality of sets at the same time. By repeating the processing of the 11th to 18th lines, where the 10th line defines the repetition, the set of tests to be executed simultaneously is determined.

  11th line: The pseudo call processing terminal number uac on the caller side in the first half test in one formulation is calculated. The function c (i, j, n) used for the calculation follows the function defined in the 21st to 23rd lines to be described later.

  12th line: The pseudo call processing terminal number uas on the called side in the first half test in one formulation is calculated. The function c (i, n-1-j, n) used for the calculation follows the function defined in the 21st to 23rd lines to be described later.

  13th and 14th lines: When the set number n of pseudo call processing terminals is an odd number, the terminal numbers uac and uas calculated in the 11th or 12th line are actually assigned by increasing n by 1. If it is larger than the maximum number of possible terminals, the set consisting of the calculated terminal numbers uac and uas is excluded from the matrix test target.

  15th and 16th lines: When a test of a set in which the terminal having the terminal number uac calculated in the 11th line is the calling side and the terminal having the terminal number uas being calculated in the 12th line is executable, Add the set of tests to the list object sameExecutableList1 of the first half test.

  17th and 18th lines: When a test of a set in which the terminal of the terminal number uas calculated in the 12th line is a caller side and the terminal of the terminal number uac calculated in the 11th line is a callee side is executable, Add the set of tests to the list object sameExecutableList2 of the second half test.

  21st line to 23rd line: Defines a function c (int r, int i, int n) used in the calculation of terminal numbers uac and uas. The function c (int r, int i, int n) outputs (replies) the value of the parameter i when the second parameter i is equal to a number smaller by 1 than the third parameter n, otherwise, This function outputs (replies) the remainder obtained by dividing (r + i) by (n-1).

  FIG. 3 shows an example of the order of matrix tests calculated by the network test planning program P. Numbers 0 to 7 in FIG. 3 indicate allocation numbers on the test plan of the pseudo call processing terminals 13-1 to 13-n. UAC indicates the originating side and UAS indicates the terminating side.

  If the identification numbers of the pseudo call processing terminals 13-1 to 13-n are not serial numbers, it is necessary to convert them into assigned numbers that are serial numbers. In this case, the correspondence between the identification numbers and the assigned numbers is managed. It is necessary to keep. If the identification numbers of the pseudo call processing terminals 13-1 to 13-n are not serial numbers, for example, the identification numbers of the pseudo call processing terminals 13-1 to 13-n are sorted in ascending or descending order, What is necessary is just to make it match | combine with a certain allocation number.

  In the case of the example of FIG. 3, eight are set as the number (n) of pseudo call processing terminals. Since it is an even number, it is applied as it is.

  When the parameter i in the fifth row is 0, the rows of the test order “1” and “2” in FIG. 3 are formed. Writing to the row of the test order “1” is performed by the processing of the 15th and 16th rows, and writing to the row of the test order “2” is performed by the processing of the 17th and 18th rows. “Test 1” in the simultaneous test related to the same row in FIG. 3 is determined when the parameter j is 0, and similarly, “test 2” to “test 4” have the parameter j of 1, 2 and 3 are determined.

  In the case of the example in FIG. 3, it is shown that four tests can be performed in parallel (in parallel), and all the tests between two sites (matrix tests) are performed by repeating the parallel test 14 times. It shows that it can be implemented. That is, all the tests between the two sites can be performed by executing the tests according to the order of the matrix test plan shown in FIG.

  In the matrix test, it is ideal to test all combinations between pseudo call processing terminals. However, in some cases, there is a section in which telephone calls (session establishment) are hardly performed between bases. In the case of this embodiment, the test between pseudo call processing terminals in such a section can be excluded from the matrix test plan.

  FIG. 4 shows an inter-pseudo call processing terminal test that the test plan unit 12A presents to the operator via the test instruction terminal 11 and is excluded from the matrix test plan (in reverse terms, the pseudo call processing terminal included in the matrix test plan) The example of a display screen for designating an interim test) is shown. In the example of the display screen of FIG. 4, even if the identification number of the pseudo call processing terminal is displayed, it is difficult for the operator to understand immediately. Therefore, the place name where the pseudo call processing terminal is provided is used instead of the identification number of the pseudo call processing terminal. Is displayed. In addition, the display screen example of FIG. 4 shows that the operator adds a mark to the check box at the intersection of the site name on the calling side (calling side) and the site name on the called side (called side), so that the matrix test plan By instructing the inter-pseudo call processing terminal test to be included and not adding a mark to the check box, the pseudo call processing inter-terminal test to be excluded from the matrix test plan is instructed.

  For example, the test plan unit 12A is instructed in the test plan mode and inputs information on each of the pseudo call processing terminals 13-1 to 13-n. In the state before starting the program P shown in FIG. 4 is displayed to instruct a pseudo call processing inter-terminal test to be excluded from the matrix test plan. Further, for example, the test plan unit 12A executes the program P shown in FIG. 2 and asks the operator whether there is a pseudo-call processing terminal test to be excluded from the matrix test plan in a state where the matrix test plan is obtained. However, when it is input from the test instruction terminal 11 that there is an inter-pseudo call processing terminal test to be excluded, the display screen example of FIG. 4 is displayed to instruct the pseudo call processing inter-terminal test to be excluded from the matrix test plan. You may do it.

  2 and 3 are for the case where the number of pseudo call processing terminals (number of bases) is 8, FIG. 4 shows the number of pseudo call processing terminals (number of bases) for the sake of simplicity of explanation. Shows the case of three. FIG. 4 shows the case where the test between Tokyo and Fukuoka and the test between Fukuoka and Tokyo are excluded from the tests between the pseudo call processing terminals installed at three locations in Tokyo, Osaka and Fukuoka. ing.

  For example, after executing the program P as shown in FIG. 2 and obtaining the result of the matrix test plan as shown in FIG. Thus, the instructed pseudo call processing inter-terminal test may be excluded (the corresponding column in FIG. 3 is left blank). Also, for example, if the instruction timing of the pseudo call processing inter-terminal test to be excluded is before the execution of the program P, the calculated sending side and receiving side set for the 15th or 17th line of the program P When determining whether or not the test can be executed, it may be confirmed whether or not the pseudo call processing inter-terminal test is excluded.

In order to measure the voice quality, a voice signal having a certain length or more must be transmitted. In addition, a test in which a plurality of calls are made simultaneously with the same pseudo call processing terminal 13-N is not desirable. From the above conditions, the upper limit T of the number of tests that can be operated in one day can be obtained by the following equation (1) (T
If it has a value after the decimal point, the integer part is the upper limit of the number of tests). Note that C (seconds) in the equation (1) is a test time (= test instruction + call time + test result transmission time).

T = 24 × 60 × 60 ÷ C (1)
Here, since the matrix test includes tests that can be executed in parallel as described above, the number of times that one call test can be executed per day is O times (upper limit), and there are n pseudo call processing terminals. In this case, the number of repetitions (maximum number) R that can be repeated in one day of the matrix test can be obtained by the following equation (2).

R = (T−O) ÷ ((M−1) × 2)
When n is an odd number, M = n + 1
When n is an even number, M = n (2)
FIG. 5 shows a matrix test calculated by the number n of pseudo call processing terminals according to the equation (2) when the test time C is 60 seconds and the upper limit number O of one call test operable in one day is 50 tests. The maximum number of repetitions. With such a calculation, it is possible to present the upper limit of the number of times of the matrix test operating in one day or the test interval to the user. For example, if there are 8 pseudo call processing terminals, the matrix test is 99.29 times, that is, the upper limit that can be executed in a day is 99 times, and the minimum test interval is an interval of 14 minutes (24 × 60 ÷ 99 integer) Part).

  The test plan unit 12A sets the test plan result of the matrix test (see FIG. 3) obtained by executing the program P shown in FIG. 2 in the low priority queue of the test instruction unit 12B at each test interval. Here, the test instruction unit 12B implements a low priority queue and a high priority queue, and the test order (test plan result) of the matrix test created by the test plan unit 12A is put in the low priority queue. 1 call test indication is placed in the high priority queue.

  FIG. 6 is a flowchart showing a test instruction operation by the test instruction unit 12B. The test instruction unit 12B starts a task (routine) shown in FIG. 6 at each test interval.

  When the test instructing unit 12B starts the task shown in FIG. 6, first, it tries to take out one call test from the high priority queue (step 21), and determines whether or not one call test exists (step 22).

  If there is a one-call test, the test instruction unit 12B instructs the one-call test (step 23). For example, the test instruction unit 12B, for the pseudo call processing terminal 13-i on the transmission side of the one-call test, receives the SIP-URI of the reception destination, the SIP server to be used (for example, IP-PBX), and the transmission destination of the test result. Instructs the one-call test with the specified address. Thereafter, the sleep time is calculated and sleep is performed for that time (step 29). The sleep time is determined such that the time from the timing when the processing of FIG. 6 is activated this time until the timing when the processing of FIG.

  In addition, when the transmission side and the reception side of the test are specified, the SIP-URI of the reception destination, the SIP server to be used, and the like are specified, and information for the specification is used as the voice information collection server. 12 is built-in.

  If it is determined in step 22 that the result is that there is no one-call test, the test instruction unit 12B extracts the matrix test plan result from the low priority queue (step 24). Then, in the matrix test plan result, it is determined whether or not the data of the parallel test that has not been executed remains (step 25). If the data remains, information on the test that can be executed in parallel from the head side is determined. Is acquired (step 26). The test instruction unit 12B gives a test instruction specifying the SIP-URI of the reception destination, the SIP server to be used, and the transmission destination address of the test result to each of the transmission side pseudo call processing terminals of the acquired parallel test. Perform (step 27). Thereafter, the test information taken out in step 26 is deleted from the list (step 28), and the process returns to step 25.

  When a test instruction or the like that can be executed in parallel (loop processing in steps 25 to 28) is repeatedly executed and the process returns to step 25 after completion of all the parallel test instructions, data of the parallel test that is not being executed is stored. It is determined that it does not remain. At this time, the test instruction unit 12B calculates the sleep time and sleeps for that time (step 29).

(A-3) Effect of Embodiment According to the above-described embodiment, a plurality of tests can be performed in parallel while satisfying the constraint that the load status of the pseudo call processing terminal is constant by creating a matrix test plan.

  When there are n pseudo call processing terminals, there are (n−1) × n tests. If these tests are run in parallel, it takes (n-1) × n × C seconds (C is the test time). When the method of the above embodiment is used, it takes (n−1) × 2 × C seconds when n is an even number, and only n × 2 × C seconds when n is an odd number. Since the number n of pseudo call processing terminals is 2 or more, all speeds are increased except when n is 2. For example, when n is 8 and C is 60 seconds, (8-1) × 8 × 60 = 56 minutes (3360 seconds) in non-parallel processing, and (8-1) × 2 × 60 = 14 minutes in parallel processing ( 840 seconds), and the parallel processing is 1/4, and one matrix test is completed.

  Matrix test observations clarify the time before and after the problem and the network location. Changes in factors such as network changes (including office layout changes) and traffic changes (including wired and wireless noise) at that time It becomes a source of information for grasping the relationship.

(B) Other Embodiments The above embodiment has been applied to regular monitoring of voice quality of an IP network to which VoIP is applied. However, the present invention can also be applied to monitoring of other networks. .

  In the above embodiment, the voice information collection server 12 has the test plan unit 12A, the test instruction unit 12B, and the test result reception unit 12C. However, the test plan unit 12A, the test instruction unit 12B, and the test result reception are shown. The unit 12C may be distributed and provided in a plurality of servers.

  Furthermore, in the above embodiment, the pseudo call processing terminal has been shown to be either a caller side or a callee side. However, a matrix test is performed by providing a dedicated caller terminal and a dedicated caller terminal at each site. You may make it do. In such a case, for example, the tests of the test orders “1” and “2” in FIG. 3 can be processed in parallel.

  Furthermore, in the above embodiment, although one call test is not executed in parallel, a plurality of one call tests are queued in the high priority queue, and the calling terminal and the receiving terminal overlap. If not, parallel processing may also be executed for the one-call test.

It is a block diagram which shows the structure of the network monitoring system which concerns on embodiment. It is explanatory drawing which shows the network test plan program of embodiment. It is explanatory drawing which shows the example of a matrix test plan order planned by execution of the network test plan program of FIG. It is explanatory drawing which shows the example of a display screen for designating the test between pseudo call processing terminals excluded from the matrix test plan of embodiment. It is explanatory drawing which shows the relationship between the number of pseudo call processing terminals and the maximum number of repetitions of a matrix test in embodiment. It is a flowchart which shows the test instruction | indication operation | movement by the test instruction | indication part of embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Network monitoring system, 11 ... Test instruction | indication terminal, 12 ... Audio | voice information collection server, 12A ... Test plan part (network test plan apparatus), 12B ... Test instruction part, 12C ... Test result receiving part, 13-1-13- n ... pseudo call processing terminal, 14 ... database, 15-1 to 15-m ... IP-PBX.

Claims (6)

A network test planning device for formulating a communication quality test plan to be executed for all combinations among all n sites where test terminals are provided,
According to a parameter i that defines the order of parallel tests, a parameter j that defines any of a plurality of tests in the parallel test, and a parameter n that takes n when n is even and takes n + 1 when n is odd, A sending site determination means for determining a test sending site;
A receiving site determination means for determining a test receiving site according to the parameter i, the parameter j, and the parameter n;
A test addition means for adding a pair of a caller base and a callee base determined by the caller base determination means and the callee base determination means to a test plan list;
In the period in which the parameter i takes the same value, the parameter j is set to n / 2 types of values when n is an even number, and (n + 1) / 2 types of values when n is an odd number. Means for changing parameters in parallel tests to change sequentially;
A network test planning device comprising: parallel test parameter changing means for sequentially changing the parameter i among (n-1) types of values.   Reverse test addition that adds a pair with the calling site determined by the calling site determination unit as the receiving site and the receiving site determined by the receiving site determination unit as the calling site is added to the test plan list The network test planning apparatus according to claim 1, further comprising means. A network test plan method for formulating a communication quality test plan to be executed for all combinations among all n sites where test terminals are provided,
A transmission side base determination means, a reception side base determination means, a test addition means, a parallel test parameter change means and a parallel test parameter change means,
The originating site determination means includes a parameter i that defines the order of parallel tests, a parameter j that defines one of a plurality of tests in the parallel test, and n when n is an even number and n is an odd number. According to the parameter n taking n + 1, the base of the test originator is determined,
The destination-side base determination means determines the base of the test receiver side according to the parameter i, the parameter j, and the parameter n,
The test adding means adds a pair of the calling site and the called site determined by the calling site determining unit and the called site determining unit to the test plan list,
In the parallel test parameter changing means, the parameter i is a period in which the parameter i takes the same value, and the parameter j is set to n / 2 types of values when n is an even number, and (n + 1) when n is an odd number. / Change sequentially between the two types of values,
The parallel test parameter changing means sequentially changes the parameter i among (n-1) types of values.   The test plan list includes a pair in which the reverse direction test adding means sets the calling base determined by the calling base determining means as the receiving base and the receiving base determined by the receiving base determination means as the calling base. The network test planning method according to claim 3, further comprising: There is a network test plan program for formulating a communication quality test plan to be executed for all combinations among all n sites where test terminals are provided,
Computer
According to a parameter i that defines the order of parallel tests, a parameter j that defines any of a plurality of tests in the parallel test, and a parameter n that takes n when n is even and takes n + 1 when n is odd, A sending site determination means for determining a test sending site;
A receiving site determination means for determining a test receiving site according to the parameter i, the parameter j, and the parameter n;
A test addition means for adding a pair of a caller base and a callee base determined by the caller base determination means and the callee base determination means to a test plan list;
In the period in which the parameter i takes the same value, the parameter j is set to n / 2 types of values when n is an even number, and (n + 1) / 2 types of values when n is an odd number. Means for changing parameters in parallel tests to change sequentially;
A network test planning program that functions as a parallel test parameter changing means for sequentially changing the parameter i among (n-1) types of values. Test terminals installed at each site,
The network test planning device according to claim 1 or 2,
According to the test plan list in the network test plan device, a test instruction means for instructing test communication from the test terminal at each of the originating bases in a plurality of tests in the parallel test to the test terminal at the terminating base,
A network monitoring system comprising: information collection means for collecting and storing information related to communication quality from a test terminal at a receiving site.

Images