JP2019075030A - Estimation device, estimation method and program - Google Patents

Abstract

To enable estimation of the quality of a display waiting time in which the influence of individual web pages has been eliminated.SOLUTION: An estimation device includes: a calculation unit that calculates a representative value of an index indicating a quality of a display waiting time, for each of a plurality of trials in which at least one of an area and a time zone is different, on the basis of measurement data of the display waiting time of a web page; a comparison unit that outputs a logical value indicating a result of comparison of the representative value associated with the trial with a threshold value for each of the trials; an extraction unit that extracts a parameter indicating a status of a communication traffic in the area and the time zone associated with the trial for each of the trials; a learning unit that causes an estimator to perform learning using the logical value of each of the trials as an objective variable and the parameter of each of the trials as an explanatory variable; and an estimation unit that inputs the parameters in a certain area and a certain time zone to the estimator to estimate the quality of the display waiting time of a web page in the area and the time zone.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an estimation apparatus, an estimation method, and a program.

  With the widespread use of the Internet, whether or not a web browser can be used comfortably occupies a large position as to whether the user can comfortably use the Internet.

  Therefore, it is important to perform quality control on display waiting time when a user browses a web page.

  In addition, in a mobile network generally represented by LTE (Long Term Evolution), the congestion situation differs between areas and time zones, so the quality of use differs depending on areas, and the quality also varies depending on time zones.

  As a method of grasping the quality for each area and time zone, there is a method (active measurement) in which content is actually accessed from an individual terminal and measurement is made (for example, Patent Document 1).

JP, 2017-92661, A

  However, performing active measurement in each area and time zone is not economical. In addition, web page servers are different from one another, and even under the influence of the same network quality degradation for each individual web page, there is a difference in the impact on display latency, etc. It is difficult to judge the quality (degree of deterioration).

  The present invention has been made in view of the above, and it is an object of the present invention to enable estimation of the quality of display waiting time in which the influence of individual web pages is removed.

  Therefore, in order to solve the above problem, the estimation apparatus determines the quality of the display waiting time based on measurement data of the display waiting time of the web page for each of a plurality of trials in which at least one of the area and the time zone is different. For each trial, a calculation unit that calculates a representative value of the indicator to be shown, a comparison unit that outputs a logical value indicating a comparison result with the threshold value for the representative value related to the trial, and for each trial The extraction unit which extracts the parameter which shows the condition of the communication traffic in each of the area and the time zone which concerns, the learning which makes the aforementioned logical value of each trial the objective variable and the parameter of the trial the explanatory variable. Section and the parameters of a certain area and time zone are input to the estimator to display the quality of the display latency of the web page in the area and time zone Having an estimation unit for estimating for.

  It is possible to enable estimation of the quality of display latency in which the influence of individual web pages has been removed.

It is a figure which shows the hardware structural example of the estimation apparatus 10 in embodiment of this invention. It is a figure which shows the function structural example of the estimation apparatus 10 in embodiment of this invention. It is a flowchart for demonstrating an example of the process sequence which the web degradation analysis part 11 performs. It is a flowchart for demonstrating an example of the process sequence which the traffic information extraction part 12 performs. It is a flowchart for demonstrating an example of the process sequence which the quality estimation part 13 performs in a learning stage. It is a flowchart for demonstrating an example of the process sequence which the estimation apparatus 10 performs in an estimation step.

  Hereinafter, embodiments of the present invention will be described based on the drawings. FIG. 1 is a diagram showing an example of a hardware configuration of an estimation device 10 according to an embodiment of the present invention. The estimation device 10 in FIG. 1 is a computer having a drive device 100, an auxiliary storage device 102, a memory device 103, a CPU 104, an interstage device 105, and the like mutually connected by a bus B.

  A program for realizing the processing in the estimation device 10 is provided by a recording medium 101 such as a CD-ROM. When the recording medium 101 storing the program is set in the drive apparatus 100, the program is installed from the recording medium 101 to the auxiliary storage apparatus 102 via the drive apparatus 100. However, the installation of the program does not necessarily have to be performed from the recording medium 101, and may be downloaded from another computer via a network. The auxiliary storage device 102 stores the installed program and also stores necessary files and data.

  The memory device 103 reads out the program from the auxiliary storage device 102 and stores it when there is an instruction to start the program. The CPU 104 executes the function related to the estimation device 10 in accordance with the program stored in the memory device 103. The inter-stage device 105 is used as an inter-stage for connecting to a network.

  FIG. 2 is a diagram showing an example of a functional configuration of the estimation device 10 according to the embodiment of the present invention. In FIG. 2, the estimation device 10 includes a web degradation analysis unit 11, a traffic information extraction unit 12, a quality estimation unit 13, and the like. These units are realized by processing that one or more programs installed in the estimation device 10 cause the CPU 104 to execute.

  The web degradation analysis unit 11 is a display waiting time in the terminal for one or more (K types) of web pages (web contents) for each of Z trials distinguished by a combination of area (region) and time zone. Terminal measurement data D1 which is a set of measurement results of The web degradation analysis unit 11 outputs, as a result of the analysis, a degree of degradation or the like which is an example of an index indicating the quality of the display waiting time.

  The traffic information extraction unit 12 extracts (acquires) a parameter or the like indicating the status of communication traffic from the traffic measurement data D2 of the network in the area and time zone of each trial.

  At the learning stage, the quality estimation unit 13 causes the one or more estimators to learn the information output from the web degradation analysis unit 11 and the information extracted by the traffic information extraction unit 12. Also, when traffic measurement data D2 in an arbitrary area and time zone is input at the estimation stage, quality estimation unit 13 sets the degree of deterioration of the display page display waiting time in the area and time zone to one or more. Estimate using an estimator.

  The estimator is, for example, a SVM (Support Vector Machine). However, other estimators may be used as long as one logical value can be estimated from the numerical vector.

  An example of a method of collecting the terminal measurement data D1 will be described. Here, URL_1, URL_2 ... URL_K: (K measurement target URLs; K is an arbitrary natural number of 1 or more (that is, one URL may be used)) is set as the URL of the measurement target web page. It is assumed that The URL is not limited to a specific one, but it is desirable to include various types of web pages that are often used.

  For each trial: As described above, the trial is a concept that is distinguished by the combination of the area to be measured and the time zone (period). Therefore, if at least one of the area and the time zone is different, the trial is different.

  (1) For the area / time zone corresponding to one trial, the display waiting time is measured P times using the terminal capable of displaying the web page for K measurement target URLs. That is, P × K measurements are performed for one trial. However, the value of P may not be common to all the URLs. Also, P may be 1. That is, measurement may not necessarily be performed multiple times for each URL. In order to obtain the representative value of the area and time zone, it is desirable that the measurement be performed at different places and times within the same area and time zone, but the measurement pattern (for example, the same area and time zone) The choice of location and time may be arbitrary, and the measurement method is not limited to a specific one. Also, the method for acquiring the display waiting time is not limited to a specific one, but a method using the time difference between navigationStart and loadEventEnd using the navigation timing API may be used. Note that it is assumed that the P measurements in each trial end within the totalized time width W.

  (2) Perform (1) for Z trials.

  The terminal measurement data D1 acquired as a result of the execution of (1) and (2) is a set of measurement values (display waiting time) represented as "L (z, y, x)" below. Here, L (z, y, x) is the display waiting time at the x-th (x: 1 to P) measurement of the y-th URL (y: 1 to K) of the z-th (z: 1 to Z) trial Indicates

  Next, a method of collecting traffic measurement data D2 will be described. Here, in order to evaluate communication traffic quality information for each area and time zone corresponding to each trial without individual active measurement, for example, all in the area or (if sampling) one A DPI (Deep Packet Inspection) device is connected to a monitoring port of a device such as a router that processes communication traffic, and traffic measurement data D2 is collected by monitoring the traffic using the DPI device. Therefore, the traffic measurement data D2 may be history information (log) of traffic monitored by the DPI device. The monitoring target is traffic in the area and time zone (time width W) corresponding to each of Z trials (however, the traffic is not limited to the traffic related to the display latency measurement). If the measurement of display latency is mobile or LTE, traffic in a specific base station (eNB) or a specific sector (ECI) is to be monitored, and if it is a fixed network, traffic in a specific accommodation router or the like is to be monitored. However, identification methods such as these base stations, sectors, and accommodating routers are not limited to predetermined ones. It is assumed that those who have the authority to monitor upstream of the network can be understood naturally.

  Hereinafter, the process procedure which the estimation apparatus 10 performs is demonstrated. FIG. 3 is a flowchart for explaining an example of the processing procedure executed by the web degradation analysis unit 11.

In step S101, the web degradation analysis unit 11 calculates a reference value regarding display waiting time for each of the K URLs. For example, if it is a reference value for a certain URL, a median, a 10% value, a minimum value or the like of Z × P measurement values of the URL may be used as the reference value. Alternatively, the reference value may be preset by the user of the estimation device 10. Hereinafter, the reference values for URL_1, URL_2 ... URL_K are Wb_1, Wb_2,. . Assume Wb_K. In this case, if the median is used as the reference value, the reference value of Wb_1 is as follows.
Wb_1 = (median of display latency set of L (z, 1, x))
Subsequently, the web degradation analysis unit 11 calculates the “deterioration degree (Dw (z, y, x)”) for each of the K URLs as follows, every Z trials (S102).
Dw (z, y, x) = L (z, y, x) / Wb_y
That is, the deterioration degree Dw (z, y, x) is a ratio of the measured value to the reference value, and in the present embodiment, is an example of an index indicating the quality (degree of deterioration) of the display waiting time of the web page. . The degradation degree Dw (z, y, x) may be calculated by another method as long as the quality of the display waiting time of the web page can be grasped.

  Subsequently, the web degradation analysis unit 11 calculates the “deterioration degree representative value (D_z)” every time Z trials (S103).

  Specifically, in the web degradation analysis unit 11, a set in which i is fixed to z and j and l are all possible values of D_z = (Dw (i, j, l) (i.e. , A set including P × K elements) is extracted for all z, and values of d_th quantiles of the respective sets are calculated. d_th is a deterioration threshold given at a quantile and is set in advance. For example, if d_th is 0.8 (ie, 80th percentile), 80% value (80th percentile) of the set of Dw (z, j, l) is calculated. If K = 1 and P = 1, the degradation degree representative value (D_z) may be set to the degradation degree (Dw (z, 1, 1)).

  Subsequently, the web deterioration analysis unit 11 calculates “the deterioration degree logical value (Db_ (z, s))” every Z trials (S104). Here, s is a variable for the degradation determination threshold for D_z and the degradation determination threshold, and in the present embodiment, (−m, − (m−1),. ,..., N−1, n), multi-stage Tsh_s is set. That is, m and n are parameters that define the number (the number of steps) of the deterioration determination threshold Tsh_s, and are set in advance. Note that n may be m or n ≠ m. Also, if only one estimator is used, then m = n = 0 (i.e., Tsh_s may be one).

Therefore, the number of Db_ (z, s) calculated as follows for one trial is m + n + 1.
Db_ (z, s) = bool (D_z> Tsh_s)
That is, for each trial, D_z and the deterioration determination threshold Tsh_s are compared for each deterioration determination threshold Tsh_s, and a logical value indicating the comparison result is substituted into Db_ (z, s).

  Db_ (z, s) is a logical value group that represents “deterioration” if true, and “non-deterioration” if false, and represents different deterioration in intensity by the difference in threshold (every s). Note that the degradation degree representative value D_z is independent of s (type of threshold), and the degradation determination threshold Tsh_s is independent of trial (z).

  Subsequently, a processing procedure performed by the traffic information extraction unit 12 will be described. FIG. 4 is a flowchart for explaining an example of the processing procedure executed by the traffic information extraction unit 12.

  In step S201, the traffic information extraction unit 12 substitutes 1 into a variable i. The variable i is a variable for identifying each trial in the description of FIG. Hereinafter, the i-th trial is referred to as “trial i”.

  Subsequently, the traffic information extraction unit 12 acquires log data corresponding to the trial i from the traffic measurement data D2 (S202). Specifically, log data corresponding to the area corresponding to the trial i and the time zone (totaling time width W) is acquired. Hereinafter, communication traffic related to the acquired log data is referred to as “target traffic”. In the area corresponding to trial i, the area is in base station units or sector units, and when the IP address or CID of the eNB is added to the log data, the IP address of the eNB in the area corresponding to trial i is used. The log data of the target traffic can be narrowed down based on the CID of the sector of the area or the area. On the other hand, when the IP address or CID of the eNB is not assigned to the log data, the range that can be observed by the monitoring port of the area corresponding to the trial i is the log data regarding the area corresponding to the trial i.

  In the subsequent steps S203 to S207, the traffic information extraction unit 12 extracts a parameter indicating the status of the communication traffic related to the trial i from the log data related to the target traffic.

  In step S203, the traffic information extraction unit 12 relates to the trial i by executing the following steps S203-1 to S203-3 based on the in-time throughput and the total transfer amount related to the specific port number of the target traffic. Calculate low throughput representative values.

In step S203-1, the traffic information extraction unit 12 extracts, from the target traffic, a TCP flow that satisfies the following conditions (1) and (2).
(1) The TCP port is thpt_port. Here, thpt_port is one of the preset parameters, and indicates the TCP throughput port number to be aggregated. In the present embodiment, the value of thpt_port is “80”.
(2) Furthermore, if th_thpt_L and th_thpt_U are set as pre-set parameters, the TCP flow having a total transfer amount of th_thpt_L or more and th_thpt_U or less among the TCP flows corresponding to (1). Here, th_thpt_L is a lower limit flow size to be extracted, and th_thpt_U indicates an upper limit flow size to be extracted. When 0 is set for both th_thpt_L and th_thpt_U, the condition (2) is invalidated.

  In step S203-2, the traffic information extraction unit 12 extracts the throughput value (within time) of each extracted TCP flow for each aggregation unit time τ for dividing (segmenting) the time zone of the trial i (aggregation time width W). Transfer amount / τ) is calculated, and the value of thpt_p1 quantile of the set of throughput values is calculated.

  Here, thpt_p1 and the following thpt_p2 are one of preset parameters, and are extraction threshold values of throughput values designated by quantiles (percentiles). Therefore, for example, when thpt_p1 is 0.2, the 20th percentile of the set of aggregated throughput values is calculated. Also, τ is one of preset parameters, and W = τ * κ (κ is an integer). Thus, step S203-2 is repeated κ times to obtain a set of κ thpt_p1 quantile values.

  In step S203-3, the traffic information extraction unit 12 calculates the value of the thpt_p2 (for example, 0.2 or the like) quantile in the set of 個 values. The value of thpt_p2 quantile point calculated in step S203-3 is a representative value of throughput (hereinafter, referred to as “Th1_z” (z is 1 .. Z)).

  Subsequent to step S203, the traffic information extraction unit 12 executes the following steps S204-1 to S204-3 based on the RTT sample calculated from the ack response of the TCP flow in the target traffic, etc. The representative value of the high RTT with respect to is calculated (S204). The method of calculating the RTT (Round Trip Time) from the TCP packet string is not limited to a predetermined one. As an example of the existing method, tcp. analysis. There is a method to read by ack_rtt.

  In step S204-1, the traffic information extraction unit 12 extracts an RTT data group for the target traffic for each aggregation unit time τ of trial i.

  In step S204-2, the traffic information extraction unit 12 calculates the value of the rtt_p1 quantile of the extracted RTT data group for each aggregation unit time τ of trial i. In step S204-1, κ RTT data groups are extracted. Therefore, in step S204-2, a set of values of κ rtt_p1 quantiles is calculated. Note that rtt_p1 and the following thpt_p2 are one of preset parameters, and are values for specifying a throughput extraction threshold with a quantile (for example, 0.1, 0.2, etc.).

  In step S204-3, the traffic information extraction unit 12 calculates the value of the rtt_p2 quantile of the values of 個 rtt_p1 quantiles. The value of the rtt_p2 quantile point calculated in step S204-3 is a representative value of the high RTT (hereinafter, referred to as "Rtt1_z" (z is 1 .. Z)).

  Subsequent to step S204, the traffic information extraction unit 12 performs the following steps S205-based on the TCP retransmission occurrence ratio (ratio of retransmission packets in the TCP packet) for each τ calculated from the analysis of the TCP flow in the target traffic. By executing 1 to S205-2, the representative value of the retransmission rate for trial i is calculated (S205). Note that the method of detecting retransmission from a TCP packet string is not limited to a predetermined one. As an example of the existing method, tcp. analysis. There is a method of evaluating from parameters such as retransmission.

  In step S205-1, the traffic information extraction unit 12 extracts the retransmission rate for the target traffic for each aggregation unit time τ. Thus, a set of κ retransmission rate values is obtained.

  In step S205-2, the traffic information extraction unit 12 calculates the value of the ret_p quantile of κ retransmission rates. Note that ret_p is one of preset parameters, and is a value for specifying a retransmission rate extraction threshold with a quantile (for example, 0.9, 0.8, etc.). The value of the ret_p quantile is a representative value of the retransmission rate (hereinafter, referred to as “Ret1_z” (z is 1 .. Z)).

  Subsequent to step S205, the traffic information extraction unit 12 executes the following steps S206-1 to S206-3 to calculate a representative value of DNS response time (S206). DNS response time is the time from DNS search to DNS response calculated from analysis of DNS search in the target traffic.

  In step S206-1, the traffic information extraction unit 12 extracts the DNS-RTT data set for the target traffic for each aggregation unit time τ of the trial i.

  In step S206-2, the traffic information extraction unit 12 calculates the value of the dnst_p1 quantile point for the extracted DNS-RTT data set for each aggregation unit time τ of the trial i. In step S206-1, κ DNS-RTT data sets are extracted. Therefore, in step S206-2, a set of values of κ dnst_p1 quantiles is calculated. Note that dnst_p1 and the following dnst_p2 are one of preset parameters, and are values for specifying a DNS response time extraction threshold with a quantile point (for example, 0.9, 0.8, etc.).

  In step S206-3, the traffic information extraction unit 12 calculates the value of the dnst_p2 quantile of the values of dn dnst_p1 quantiles. The value of the dnst_p2 quantile is a representative value of the DNS response time (DNS-RTT) (hereinafter, referred to as “Drt1_z” (z is 1 .. Z)).

  Subsequent to step S206, the traffic information extraction unit 12 executes the following steps S207-1 to S207-3 to calculate a representative value of the DNS response rate (S207). The DNS response rate is the ratio of responses to DNS searches calculated from analysis of DNS searches in monitored traffic.

  In step S207-1, the traffic information extraction unit 12 extracts the number of DNS searches for the target traffic and the number of DNS responses for the search for each aggregation unit time τ of the trial i. In addition, since a response may become time after (tau), it is counted until dns_timeout progress. dns_timeout is one of preset parameters, and is a timeout time at the time of DNS response rate calculation.

  In step S207-2, the traffic information extraction unit 12 calculates the response rate (the number of DNS responses / the number of DNS searches) for each aggregation unit time τ of the trial i. There is no value for τ with a search number of 0. In step S207-2, κ response rates are calculated. Therefore, a set of κ response rates is obtained.

  In step S207-3, the traffic information extraction unit 12 excludes “no value” from the set of κ response rates, and dnsr_p of values of ((may be excluded less than κ). Calculate the quantile value. dnsr_p is one of preset parameters, and is a value for specifying a DNS response rate extraction threshold with a quantile (for example, 0.2 or the like). The value of the dnsr_p quantile is a representative value of the DNS response rate (hereinafter, referred to as “Drp1_z” (z is 1 .. Z)).

Steps S202 to S207 are performed for all the trials Z times (S208, S209). As a result, the following vectors are obtained as the throughput representative value Th1.
Th1 = (Th_1, Th1_2, Th1_3 ... Th1_z)
A plurality of sets of thpt_p1 and thpt_p2 may be set. In that case, vectors such as Th2_z and Th3_z are obtained for each set.

Also, the following vector is obtained as the high RTT representative value Rtt1.
Rtt1 = (Rtt1_1, Rtt1_2, ... Rtt1_z)
A plurality of sets of rtt_p1 and rtt_p2 may be set. In that case, vectors such as Rtt2_z and Rtt3_z are obtained for each set.

Also, the following vector is obtained as the retransmission rate representative value Ret1.
Ret1 = (Ret1_1, Ret1_2, ... Ret1_z)
A plurality of values may be set for ret_p. In that case, vectors such as Ret2_z and Ret3_z are obtained for each value.

Also, the following vector is obtained as the DNS response time representative value Drt1.
Drt1 = (Drt1_1, Drt1_2, ... Drt1_z)
A plurality of sets of dnst_p1 and dnst_p2 may be set. In that case, vectors such as Drt2_z and Drt3_z are obtained for each set.

Furthermore, the following vector is obtained as the DNS response rate Drp1.
Drp1 = (Drp1_1, Drp1_2, ... Drp1_z)
A plurality of values may be set for dnsr_p. In that case, vectors such as Drp2_z and Drp3_z are obtained for each value.

Subsequently, the traffic information extraction unit 12 outputs the following DPIvec as a set (matrix) of vectors of the respective representative values each having the length Z (S210).
DPIvec = (Th1 (, Th2, Th3 ..), Rtt1 (, Rtt2, Rtt3 ..), Ret1 (, Ret2, Ret3 ..), Drt1 (, Drt2, Drt3), Drp1 (, Drp2, Drp3 ..) )
Note that (, Th2, Th3 ..), (, Rtt2, Rtt3 ..), (, Ret2, Ret3 ..), (, Drt2, Drt3) and (, Drp2, Drp3 ..) are used as preset parameters. It is shown that the number of sets of vectors is changed by the number of sets of threshold values (quantile points) of.

  Subsequently, a processing procedure performed by the quality estimation unit 13 in the learning stage will be described. FIG. 5 is a flowchart for explaining an example of the processing procedure performed by the quality estimation unit 13 in the learning stage.

In step S301, the quality estimation unit 13 inputs the degradation degree logic value Db_ (z, s), which is the following logic value vector of m + n + 1 lengths of Z calculated by the web degradation analysis unit 11.
Db_ (z, s) = (Db_ (z, -m), Db_ (z, -m + 1), ... Db_ (z, n-1), Db_ (z, n))
Note that Db_ (z, s) is a vector of Z element number when s is fixed. Since s takes m + n + 1 values, Db_ (z, s) is a vector having m + n + 1 Z vectors as elements.

Subsequently, the quality estimation unit 13 inputs the following set (vector) of vectors of representative values output by the traffic information extraction unit 12 (S302).
DPIvec = (Th1 (, Th2, Th3 ..), Rtt1 (, Rtt2, Rtt3 ..), Ret1 (, Ret2, Ret3 ..), Drt1 (, Drt2, Drt3), Drp1 (, Drp2, Drp3 ..) )
Subsequently, the quality estimation unit 13 generates m + n + 1 (ie, one or more) estimator instances (S303). That is, the same number of estimators (for each degradation determination threshold Tsh_s) as the number of degradation determination thresholds Tsh_s is prepared. Each estimator is denoted "S_s". Here, s corresponds to m + n + 1 types of Tsh_s.

  Subsequently, the quality estimation unit 13 learns each S_s with Db_ (z, s) as an objective variable and DPIvec as an explanatory variable (S304). Each S_s learns about Db_ (z, s) common to s. In this way, S_s has been learned and can be used to estimate the quality of the display page waiting time.

  Then, the process sequence which the estimation apparatus 10 performs in an estimation step is demonstrated. FIG. 6 is a flowchart for explaining an example of the processing procedure performed by the estimation device 10 in the estimation stage.

  As a premise of the processing procedure of FIG. 6, it is assumed that traffic measurement data D2 (for example, log data of DPI) corresponding to the area and time zone to be estimated are collected. The traffic measurement data D2 is hereinafter referred to as "estimated target measurement data". It is not necessary to measure the display waiting time of the web page. This is because the degree of deterioration of the display waiting time is the target variable (target of estimation).

  In step S401, the traffic information extraction unit 12 inputs estimation target measurement data. In the subsequent steps S402 to S406, the traffic information extraction unit 12 executes the same processing as in steps S203 to S207 of FIG. 4 to calculate (extract) each representative value from the estimation target measurement data. However, since there is one area and time zone to be estimated, Z = 1.

  As a result, the traffic information extraction unit 12 obtains a matrix of the same format as DPIvec, which has a length Z = 1, as in step S210 of FIG. Hereinafter, the matrix is referred to as "DPIvec_new".

Subsequently, the quality estimation unit 13 inputs a set of DPIvec_new to each S_s (each estimator) as described below to set a set of logical values (whether or not to be degraded) according to each degradation determination threshold Tsh_s. Get Est_s which is a vector) (S407)
Est_s = S_s (DPIvec_new) (for multiple s)
Note that Est_s is a set (vector) of logical values of length m + n + 1.

  Typically, when the value of Tsh_s is true (deterioration) for the maximum s, and s is changed so that Tsh_s becomes smaller in order, it becomes false (non-deterioration) when some s is θ, In the following Tsh_s, the pattern is all false (non-degraded) (although all false, all true, or the order may be broken).

  The quality estimation unit 13 searches for such θ from Est_s, and outputs Tsh_θ corresponding to the searched θ as “the estimated degree of deterioration” (estimated value of the degree of deterioration for the area and time zone to be estimated). (S409). Specifically, for example, if all false (non-deteriorated), the quality estimation unit 13 outputs “non-deteriorated”. In this case, the degradation is less than (minimum) Tsh_s. If all are true (deteriorated), the quality estimation unit 13 outputs “(maximum) Tsh_s or more”. Also, if all is not false and all is not true, even if false and true are not separated by a certain Tsh_s, the quality estimation unit 13 checks the values in order from the side with the largest Tsh_s, Output "Tsh_θ or more" using Tsh_θ that first becomes false. This Tsh_θ is an estimation result of the degree of deterioration of the display waiting time of the web page (which does not depend on only the individual web page).

  In this embodiment, a low throughput representative value, a high RTT representative value, a retransmission rate representative value, a DNS response time representative value, and a DNS response rate representative value as an example of a parameter for extracting a parameter indicating the status of communication traffic. Although the example which uses is demonstrated, these all may not necessarily be used and some (for example, any one) of these may be used. In addition, parameters other than these may be used as long as the parameters can grasp the status of communication traffic.

  As described above, according to the present embodiment, the estimation unit learns the relationship between the measured value of the display waiting time of the web page and the information on the communication traffic corresponding to the area and time zone of the measurement of the display waiting time. The degree of degradation of the latency of the web page can be estimated (independent of only the individual page) by inputting to the estimator the information on the communication traffic of unknown data to be estimated. Therefore, according to the present embodiment, it is possible to estimate the quality of display waiting time in which the influence of the individual web page is removed.

  Moreover, in the present embodiment, a plurality of web page servers are used for learning, and a general characteristic of "deterioration of the display page waiting time" (for reasons such as web page side not network origination) is partially By using the concept of “deterioration degree” normalized on the basis of each web page, and a method of extracting data by quantile points of multiple measurement data, in order to extract excluding the influence of things that only degrade Absorb the difference in waiting time for each web page to catch the deterioration of the waiting time for web pages. Therefore, it is possible to solve the problem that web pages and servers are different from one another, and even if they are used at the same time and place, the waiting time differs depending on each page.

  Further, in the present embodiment, against the instability of the contents of the passive measurement data (the communication data analysis log (user, communication partner, purpose is also disjointed from the communication data analysis log (DPI log)), Such instability compared to the case of extracting an average value or the like by extracting a quantile point of traffic data that has good correspondence with deterioration information, etc., and is unlikely to be influenced by an extreme abnormal value. Can respond appropriately to

In addition, although the appearance of the influence on the deterioration of the display waiting time is different for each individual web page even under the same NW deterioration influence, in the present embodiment, a plurality of target web pages are treated collectively and evaluated by the statistics By doing this, it can be quantified as "the typical way of causing deterioration". Also, in general, deterioration determination is performed using two values (logical values). Although the present embodiment also uses a logical value estimator, in the present embodiment a quantitative value is used as the “degree of degradation”, so when converting into a logical value vector at the time of learning (corresponding to a plurality of degradation levels) It is also possible to realize the determination of the degree of deterioration by preparing a plurality of estimators and preparing different levels for each. Note that, in the present embodiment, the web degradation analyzer 11 includes a calculator and a comparator. An example of The traffic information extraction unit 12 is an example of an acquisition unit. The quality estimation unit 13 is an example of a learning unit and an estimation unit.

  Although the embodiments of the present invention have been described above in detail, the present invention is not limited to such specific embodiments, and various modifications may be made within the scope of the present invention as set forth in the claims. Modifications and changes are possible.

DESCRIPTION OF REFERENCE NUMERALS 10 estimation device 11 web degradation analysis unit 12 traffic information extraction unit 13 quality estimation unit 100 drive device 101 recording medium 102 auxiliary storage device 103 memory device 104 CPU
105 Inter-stage device B bus

Claims (7)

A calculation unit that calculates a representative value of an indicator indicating the quality of the display waiting time based on measurement data of the display waiting time of the web page for each of a plurality of trials in which at least one of the area and the time zone is different;
A comparison unit that outputs, for each trial, a logical value indicating a comparison result with a threshold for the representative value related to the trial;
An acquisition unit configured to acquire, for each of the trials, a parameter indicating a state of communication traffic in each of a region and a time zone related to the trial;
A learning unit that uses the logical value of each trial as a target variable and the parameter of each trial as an explanatory variable to learn an estimator;
An estimation unit that inputs the parameters in a certain area and time zone to the estimator to estimate the quality of display latency of web pages in the area and time zone;
An estimation apparatus characterized by having: The calculation unit calculates a representative value of an indicator indicating the quality of the display waiting time of the web page based on the measurement data of the display waiting time of each of the plurality of web pages for each trial.
The estimation device according to claim 1, characterized in that: The comparison unit outputs, for each trial, a logical value group indicating comparison results with a plurality of threshold values for the representative value related to the trial;
The learning unit learns a plurality of estimators corresponding to each of the plurality of threshold values, with the logical value group of each trial as a target variable and the parameter of each trial as an explanatory variable.
The estimation unit inputs the parameters in a certain area and a time zone to the estimator corresponding to each of the plurality of threshold values to estimate the quality of display waiting time of the web page in the area and the time zone.
The estimation device according to claim 1 or 2, characterized in that: The acquisition unit acquires the parameter for each of a plurality of unit times dividing the time zone, and calculates a representative value of the acquired parameter for each unit time.
The learning unit learns the estimator using the logical value of each trial as a target variable and the representative value of the parameter of each trial as an explanatory variable.
The estimation apparatus according to any one of claims 1 to 3, characterized in that: The calculation unit calculates a representative value of an indicator indicating the quality of the display waiting time of the web page based on a plurality of measurement data of the display waiting time of the web page for each trial.
The estimation apparatus according to any one of claims 1 to 4, characterized in that: A calculation procedure of calculating a representative value of an indicator indicating the quality of the display waiting time based on measurement data of the display waiting time of the web page for each of a plurality of trials in which at least one of the region and the time zone is different;
A comparison procedure for outputting, for each trial, a logical value indicating a comparison result with a threshold for the representative value related to the trial;
An acquisition procedure for acquiring, for each of the trials, a parameter indicating a state of communication traffic in each of the region and the time zone related to the trial;
A learning procedure in which the logical value of each trial is used as an objective variable, and the parameter of each trial is used as an explanatory variable to learn an estimator;
An estimation procedure for inputting the parameters in a certain area and time zone into the estimator to estimate the quality of display latency of web pages in the area and time zone;
Estimation method characterized in that the computer executes. A program which causes a computer to function as each part according to any one of claims 1 to 5.

Images