User network quality identification method, device, electronic device and storage medium

Abstract

The present invention provides a user network quality identification method, device, electronic device, and storage medium, relating to the field of data analysis and processing technology. The method includes: obtaining network perception data of a user device at multiple times; determining the inflection point of a network perception data curve based on the network perception data and the time sequence corresponding to the network perception data; dividing the network perception data into multiple sets based on the inflection point, with the common point between two adjacent sets as the inflection point, fitting the network perception data in each set to obtain a network perception data curve; determining a target interval based on the left and right derivatives of the points corresponding to each time sequence in the network perception data curve; and outputting a user network quality identification result based on the target interval, wherein the user network quality identification result reflects the relationship between the network quality of the user device and the time period. The present invention can identify the relationship between network quality and time period.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis and processing, and in particular to a method, device, electronic device and storage medium for identifying user network quality. Background Art

[0002] There's a close relationship between network quality and user perception. User perception refers to a user's direct experience of network service quality. Poor network speed can significantly impact user experience, such as web browsing, online video playback, online gaming, and online shopping. Therefore, poor wireless network quality can have a significant negative impact on user perception, leading users to question the quality of network services and reducing their trust and satisfaction with them. To improve user satisfaction, network service providers need to identify periods of poor network quality, issue appropriate work orders, and implement measures to improve network stability, ensuring users enjoy high-quality network services.

[0003] In the existing technology, anomaly detection algorithms are often used to identify network quality. Anomaly detection determines outliers in network traffic by analyzing data distribution. This method cannot identify an interval with poor network quality, but can only identify several abnormal outliers based on the distribution. In other words, it is impossible to identify the relationship between network quality and time period. Summary of the Invention

[0004] The present invention provides a user network quality identification method, device, electronic device and storage medium, which are used to solve the defect in the prior art that the relationship between network quality and time period cannot be identified, and realize the identification of the relationship between network quality and time period.

[0005] The present invention provides a method for identifying user network quality, comprising:

[0006] Acquire network perception data of a user device at multiple times, where the network perception data reflects the amount of network downloaded data of the user device;

[0007] Determining an inflection point of a network perception data curve based on the network perception data and the time sequence number corresponding to the network perception data, wherein the network perception data curve is a curve showing changes in the network perception data as the time sequence number changes, and an interval between moments corresponding to two adjacent time sequence numbers is a preset duration;

[0008] Dividing the network perception data into multiple sets based on the inflection point, where a common point between two adjacent sets is the inflection point, and fitting the network perception data in each set to obtain the network perception data curve;

[0009] Determine a target interval based on the left derivative and the right derivative of each point corresponding to each time sequence number in the network perception data curve, wherein the start and end points of the target interval are points where the difference between the corresponding left derivative and the right derivative is greater than a preset difference standard;

[0010] A user network quality identification result is output based on the target interval, where the user network quality identification result reflects the relationship between the network quality of the user equipment and the time period.

[0011] According to a user network quality identification method provided by the present invention, determining the inflection point of a network perception data curve based on the network perception data and a time sequence number corresponding to the network perception data includes:

[0012] Normalizing the network perception data and the time sequence number corresponding to the network perception data to obtain normalized network perception data and a normalized time sequence number;

[0013] generating a normalized network perception data curve based on the normalized network perception data and the normalized time sequence number;

[0014] Dividing the normalized network perception data curve into at least one sub-curve based on the directionality and concavity of the normalized network perception data curve, each sub-curve corresponding to only one directionality and one concavity;

[0015] Updating the normalized network perception data based on the directionality and concavity of each sub-curve to obtain updated normalized network perception data, and obtaining an updated normalized network perception data curve based on the updated normalized network perception data, wherein the directionality of the updated normalized network perception data curve is increasing and the concavity is concave;

[0016] Generating a differential curve based on the updated normalized network perception data curve, wherein the value corresponding to each normalized time sequence number in the differential curve is the difference between the updated normalized network perception data and the normalized time sequence number;

[0017] The local maximum point in the differential curve is used as a reference point, the value corresponding to the local maximum point in the differential curve is greater than the values ​​corresponding to the two adjacent points on the left and right of the local maximum point in the differential curve, and the inflection point is determined based on the value corresponding to the reference point in the differential curve and the network perception data corresponding to the reference point.

[0018] According to a user network quality identification method provided by the present invention, the normalized network perception data is updated based on the directionality and concavity of each sub-curve to obtain the updated normalized network perception data, including:

[0019] When the directionality of the sub-curve is decreasing and the concavity is concave, updating each of the normalized network perception data in the sub-curve to a value that is reversely sorted according to the normalized time sequence number;

[0020] When the directionality of the sub-curve is decreasing and the concavity is convex, updating the normalized network perception data in the sub-curve to a target difference value, where the target difference value is a difference between a target maximum value and the normalized network perception data, and the target maximum value is a maximum value among the normalized network perception data in the sub-curve;

[0021] When the directionality of the sub-curve is increasing and the concavity is convex, updating the normalized network perception data in the sub-curve to a value obtained by reversely sorting the target difference values ​​according to the normalized time sequence number order;

[0022] When the directionality of the sub-curve is increasing and the concavity is concave, the normalized network-sensed data in the sub-curve is maintained unchanged.

[0023] According to a user network quality identification method provided by the present invention, fitting the network perception data in each of the sets to obtain the network perception data curve includes:

[0024] Obtaining a network perception data sub-curve corresponding to each of the sets, and connecting the network perception data sub-curves corresponding to each of the sets to obtain the network perception data curve;

[0025] Obtaining the network-sensing data sub-curve corresponding to the set includes:

[0026] Using the network perception data in the set as a dependent variable and the time sequence number as an independent variable, fitting multiple univariate polynomial regression models, each of the univariate polynomial regression models having a different number of terms;

[0027] The network-aware data sub-curve corresponding to the set is determined based on each of the univariate polynomial regression models.

[0028] According to a method for identifying user network quality provided by the present invention, determining a target interval based on the left derivative and the right derivative of a point corresponding to each time sequence number in the network perception data curve includes:

[0029] Setting the first target point as the starting point of the first interval and the second target point as the end point of the first interval;

[0030] Setting the third target point as the starting point of the second interval and the fourth target point as the end point of the second interval;

[0031] using the first interval and the second interval as the target interval;

[0032] Among them, the left derivative of the first target point is less than the first preset threshold, and the difference between the right derivative and 0 is less than the second preset threshold, the left derivative of the second target point is greater than the third preset threshold, and the difference between the right derivative and 0 is less than the second preset threshold, the difference between the left derivative of the third target point and 0 is less than the second preset threshold, and the right derivative is greater than the third preset threshold, and the difference between the left derivative of the fourth target point and 0 is less than the second preset threshold, and the right derivative is less than the first preset threshold.

[0033] According to a user network quality identification method provided by the present invention, outputting a user network quality identification result based on the target interval includes:

[0034] The user network quality identification result is output based on the target interval including the inflection point, wherein the user network quality in a time period corresponding to the target interval including the inflection point is lower than that in other time periods.

[0035] According to a user network quality identification method provided by the present invention, the network perception data is HTTP downstream traffic.

[0036] The present invention also provides a user network quality identification device, comprising:

[0037] A data acquisition module is used to obtain network perception data of the user equipment at multiple times, wherein the network perception data reflects the amount of network downloaded data of the user equipment;

[0038] an inflection point determination module, configured to determine an inflection point of a network perception data curve based on the network perception data and a time sequence number corresponding to the network perception data, wherein the network perception data curve is a curve showing changes in the network perception data as the time sequence number changes, and an interval between moments corresponding to two adjacent time sequence numbers is a preset duration;

[0039] a curve fitting module, configured to divide the network perception data into a plurality of sets based on the inflection point, wherein a common point between two adjacent sets is the inflection point, and to fit the network perception data in each set to obtain the network perception data curve;

[0040] an interval partitioning module, configured to determine a target interval based on the left derivative and the right derivative of each point corresponding to each time sequence number in the network perception data curve, wherein the start and end points of the target interval are points where the difference between the corresponding left derivative and the right derivative is greater than a preset difference standard;

[0041] The result generating module is configured to output a user network quality identification result based on the target interval, where the user network quality identification result reflects the relationship between the network quality of the user equipment and the time period.

[0042] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any of the above-mentioned user network quality identification methods is implemented.

[0043] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for identifying user network quality as described above is implemented.

[0044] The user network quality identification method, device, electronic device and storage medium provided by the present invention divide the network perception data of the user device into multiple sets according to the inflection points of the network perception data curve of the user device, fit the network perception data in each set to obtain the network perception data curve, and then identify the target interval based on the derivative of the midpoint of the network perception data curve. Based on the target interval, a network quality identification result including the relationship between network quality and time period is obtained. This can realize the identification of the relationship between network quality and time period, and facilitate network service providers to determine the time period with poor network quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 This is a flow chart of a method for identifying user network quality provided by the present invention;

[0047] Figure 2 It is a line graph of HTTP downstream traffic in a method for identifying user network quality provided by the present invention;

[0048] Figure 3 It is a schematic diagram of an inflection point in a method for identifying user network quality provided by the present invention;

[0049] Figure 4 This is a schematic diagram of the start and end points of a target interval in a method for identifying user network quality provided by the present invention;

[0050] Figure 5 It is a schematic diagram of a network quality identification result in a user network quality identification method provided by the present invention;

[0051] Figure 6 It is a structural diagram of the user network quality identification device provided by the present invention;

[0052] Figure 7 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0053] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0054] The following combination Figure 1-Figure 5 Describe the user network quality identification method provided by the present invention, such as Figure 1 As shown, the method includes the steps of:

[0055] S110: Acquire network perception data of the user device at multiple times, where the network perception data reflects the amount of network downloaded data of the user device;

[0056] S120: Determine an inflection point of a network perception data curve based on the network perception data and the time sequence numbers corresponding to the network perception data, where the network perception data curve is a curve showing changes in the network perception data as the time sequence numbers change, and the interval between the time sequences corresponding to two adjacent time sequence numbers is a preset duration.

[0057] S130, dividing the network perception data into multiple sets based on inflection points, with the common point between two adjacent sets being the inflection point, and fitting the network perception data in each set to obtain a network perception data curve;

[0058] S140: Determine a target interval based on the left derivative and the right derivative of each point corresponding to each time sequence number in the network perception data curve, where the start and end points of the target interval are points where the difference between the corresponding left derivative and the right derivative is greater than a preset difference standard;

[0059] S150: Output a user network quality identification result based on the target interval, where the user network quality identification result reflects the relationship between the network quality of the user device and the time period.

[0060] The network perception data may be HTTP downstream traffic. In the method provided by the present invention, the network perception data of the same device is processed to obtain the user network quality identification result, thereby preventing the network device from affecting the result. The network perception data of the user device at multiple moments are obtained, and the interval between each moment may be a preset duration, such as 10 minutes, that is, the user perception data of the user device is collected every 10 minutes. When collecting user perception data, the corresponding fields may be collected, including time (which may include days, hours, minutes), device, sub-device, sub-device type, etc. Furthermore, when collecting user perception data, data with missing fields may be cleaned to remove data with missing fields.

[0061] like Figure 2 As shown in the figure, if the user's network speed in a day is plotted as a line graph, under normal circumstances, the user's network traffic should be a reminder or a rectangle. However, if there is a problem with the network service, the user's network traffic will fluctuate and decrease, that is, there will be a period of poor network quality, for example Figure 2 The first and fourth boxes in the figure can be regarded as time periods with good network quality, and the second and third boxes are time periods with poor network quality.

[0062] The network perception data curve is a curve whose independent variable is the time sequence number and whose dependent variable is the network perception data. The method provided by the present invention fits this curve and identifies time periods of poor network quality based on its derivative. Specifically, the inflection point of the network perception data curve is first identified.

[0063] Determining an inflection point of a network perception data curve based on the network perception data and a time corresponding to the network perception data includes:

[0064] Normalizing the network perception data and the time sequence number corresponding to the network perception data to obtain normalized network perception data and a normalized time sequence number;

[0065] generating a normalized network perception data curve based on the normalized network perception data and the normalized time sequence number;

[0066] The normalized network perception data curve is divided into at least one sub-curve based on the directionality and concavity of the normalized network perception data curve, and each sub-curve corresponds to only one directionality and one concavity;

[0067] The normalized network perception data is updated based on the directionality and the concavity and convexity of each sub-curve to obtain updated normalized network perception data, and an updated normalized network perception data curve is obtained based on the updated normalized network perception data, wherein the directionality of the updated normalized network perception data curve is increasing and the concavity and convexity is concave;

[0068] Generate a differential curve based on the updated normalized network perception data curve, where the differential curve summarizes the value corresponding to each normalized time sequence number as the difference between the updated normalized network perception data and the normalized time sequence number;

[0069] The local maximum point in the differential curve is used as the reference point. The value corresponding to the local maximum point in the differential curve is greater than the values ​​corresponding to the two adjacent points on the left and right of the local maximum point in the differential curve. The inflection point is determined based on the value corresponding to the reference point in the differential curve and the network perception data corresponding to the reference point.

[0070] Taking the network perception data as the dependent variable y and the time sequence number as the independent variable x, first normalize x and y to obtain the normalized time sequence number x_normalize and the normalized network perception data y_normalize. Based on x_normalize and the corresponding y_normalize, a normalized network perception curve can be generated. The normalized network perception curve is a curve composed of the individual broken lines obtained by sequentially connecting (x_normalize, y_normalize). Next, y_normalize is converted based on the directionality and concavity of the normalized network perception curve, converting y_normalize into data with increasing directionality and concave concavity. This updates the normalized network perception curve. The updated normalized network perception curve has increasing directionality and concave concavity.

[0071] Specifically, the normalized network perception data is updated based on the directionality and concavity of each sub-curve to obtain the updated normalized network perception data, including:

[0072] When the directionality of the sub-curve is decreasing and the concavity is concave, each normalized network perception data in the sub-curve is updated to a value that is sorted in reverse order according to the normalized time sequence number;

[0073] When the directionality of the sub-curve is decreasing and the concavity is convex, the normalized network perception data in the sub-curve is updated to the target difference, where the target difference is the difference between the target maximum value and the normalized network perception data, and the target maximum value is the largest value in the normalized network perception data in the sub-curve;

[0074] When the directionality of the sub-curve is increasing and the concavity is convex, the normalized network perception data in the sub-curve is updated to the value obtained by sorting the target difference in reverse order according to the normalized time sequence number;

[0075] When the directionality of the sub-curve is increasing and the concavity is concave, the normalized network-sensed data in the sub-curve is maintained unchanged.

[0076] That is, if the directionality of y_normalize is decreasing and the concavity is concave, y_normalize is reversed in sequence; if the directionality of y_normalize is decreasing and the concavity is convex, each bit of y_normalize is changed to the difference between the maximum value of y_normalize and y_normalize; if the directionality of y_normalize is increasing and the concavity is convex, each bit of y_normalize is changed to the difference between the maximum value of y_normalize and y_normalize, and then reversed in sequence. When reversing the order, the normalized network perception data are sorted in reverse order according to the time sequence corresponding to the normalized network perception data. That is, assuming that according to the time sequence, the values ​​of the five normalized network perception data before the update in the sub-curve are ABCDE, then the values ​​of the five normalized network perception data after the update in the sub-curve are EDCBA, that is, the value of the first normalized network perception data in the sub-curve is updated to the value of the m-th normalized network perception data in the sub-curve, the value of the second normalized network perception data is updated to the value of the m-1-th normalized network perception data in the sub-curve, and so on, where m is the number of all normalized network perception data included in the sub-curve.

[0077] After the conversion, the updated normalized network perception data is subtracted from its corresponding normalized time sequence. This is done by subtracting x_normalize from y_normalize to obtain the difference curve. The local maximum points of the difference curve are determined. These points indicate when the rate of increase of y begins to decrease. The local maximum points of the difference curve correspond to local maxima, which are values ​​greater than the two adjacent values ​​on the left and right.

[0078] The inflection point is determined by using the local maximum point as a reference point. Specifically, the inflection point is determined based on a value corresponding to the reference point in the difference curve and network perception data corresponding to the reference point, including:

[0079] If the value corresponding to the next point of the reference point i in the difference curve is less than the threshold Then determine the directionality and concavity of the network perception data corresponding to the reference point;

[0080] If the directionality of the network perception data corresponding to the reference point i is decreasing and the concavity is convex, or the directionality of the network perception data corresponding to the reference point is increasing and the concavity is concave, then the point corresponding to the time sequence number x′ is taken as the inflection point, where x′ is the time sequence number corresponding to the reference point;

[0081] If the directionality of the network perception data corresponding to the reference point is increasing and the concavity is convex, or the directionality of the network perception data corresponding to the reference point is decreasing and the concavity is concave, then the point corresponding to the time sequence number -x′+n is determined as the inflection point, where n is the total number of all time sequence numbers;

[0082] If the value corresponding to the next point of the reference point i in the difference curve is not less than the threshold Then skip reference point i and proceed to judge the next reference point i.

[0083] The directionality and concavity of the network perception data corresponding to the reference point are consistent with the directionality and concavity of the sub-curve where the time sequence number corresponding to the reference point is located in the normalized network perception data curve. The calculation formula is:

[0084]

[0085] in, is the local maximum value corresponding to the reference point i, S is the preset sensitivity parameter, which is a constant. is the i+1th value in the normalized time sequence; is the i-th value in the normalized time sequence; n is the total number of time sequences. The smaller S is, the faster the inflection points can be found (i.e., the more inflection points can be found).

[0086] The schematic diagram of the inflection point is as follows Figure 3 shown.

[0087] After determining the inflection point, the network perception data is divided into multiple sets based on the inflection point. The common point between two adjacent sets is the inflection point, that is, the end point of the previous set is the starting point of the next set. The network perception data in each set is fitted to obtain the network perception data curve, including:

[0088] Obtain the network perception data sub-curve corresponding to each set, and connect the network perception data sub-curves corresponding to each set to obtain a network perception data curve;

[0089] The generation process of the network-aware data sub-curve corresponding to the set includes:

[0090] The network perception data in the collection is used as the dependent variable and the time sequence number is used as the independent variable to fit multiple univariate polynomial regression models, each with a different number of terms.

[0091] The network perception data sub-curve corresponding to the set is determined based on each univariate polynomial regression model.

[0092] Based on the inflection points obtained previously, the wireless network perception data can be divided into several parts, each of which is a set containing at least one wireless network perception data value. For each set, a univariate m-order polynomial regression model is trained using the wireless network perception data as the dependent variable and the time series as the independent variable. m ranges from 1 to 5. For each value of m, a model is trained and the mean square error (MSE) between the predicted wireless network perception data obtained by the model and the actual collected wireless network perception data is calculated. The curve corresponding to the univariate m-order polynomial regression model with the minimum MSE is selected as the network perception data subcurve.

[0093] After fitting the network perception data curve, the target interval is determined based on the left derivative and right derivative of the point corresponding to each time sequence number in the network perception data curve, including:

[0094] Set the first target point as the starting point of the first interval and the second target point as the end point of the first interval;

[0095] The third target point is set as the starting point of the second interval, and the fourth target point is set as the end point of the second interval;

[0096] The first interval and the second interval are taken as target intervals;

[0097] Among them, the left derivative of the first target point is less than the first preset threshold, and the difference between the right derivative and 0 is less than the second preset threshold, the left derivative of the second target point is greater than the third preset threshold, and the difference between the right derivative and 0 is less than the second preset threshold, the difference between the left derivative of the third target point and 0 is less than the second preset threshold, and the right derivative is greater than the third preset threshold, and the difference between the left derivative of the fourth target point and 0 is less than the second preset threshold, and the right derivative is less than the first preset threshold.

[0098] Perform slope analysis on the network perception data curve, and for the point (x i ,y i ), x i -0.001 and x i +0.001 is input into the corresponding curve expression, and we get and Calculate point x using the following formula i The left derivative y′ left and the right derivative y′ right :

[0099]

[0100] When the difference between the derivative and 0 is less than the second preset threshold, the derivative is considered close to 0. The points where the left derivative is a negative number less than the threshold θ1 and the right derivative is close to 0 are recorded as candidate interval starting points, denoted as λ1-type starting points. The points where the left derivative is a positive number greater than the threshold λ2 and the right derivative is close to 0 are recorded as candidate interval endpoints, denoted as η1-type endpoints. The points where the left derivative is close to 0 and the right derivative is a positive number greater than the threshold θ2 are recorded as candidate interval starting points, denoted as λ2-type starting points. The points where the left derivative is close to 0 and the right derivative is a negative number less than the threshold θ1 are recorded as candidate interval endpoints, denoted as η2-type endpoints. The values ​​of the first preset threshold θ1, the second preset threshold, and the third preset threshold θ2 can be determined through multiple experiments. For example, based on experiments and expert verification, the value of θ1 can be -10, the value of θ2 can be 16, and the value of the second preset threshold can be 0.1.

[0101] like Figure 4 As shown in Figure 1, the data is segmented according to the time points λ1, η1, λ2, and η2 to obtain several intervals as target intervals. Based on the target intervals, the user network quality identification results are output, including:

[0102] The user network quality identification result is output based on the target interval including the inflection point, and the user network quality identification result is summarized. The user network quality in the time period corresponding to the rated target interval including the inflection point is lower than that in other time periods.

[0103] If there is an inflection point within the interval, then the interval is a poor quality interval for the user's wireless network, such as Figure 5 Returns the start and end times of all intervals with poor wireless network quality.

[0104] The user network quality identification device provided by the present invention is described below. The user network quality identification device described below and the user network quality identification method described above can be referred to each other. Figure 6 As shown, the user network quality identification device provided by the present invention includes:

[0105] The data collection module 610 is used to obtain network perception data of the user equipment at multiple times, where the network perception data reflects the amount of network downloaded data of the user equipment;

[0106] An inflection point determination module 620 is configured to determine an inflection point of a network perception data curve based on the network perception data and the time sequence number corresponding to the network perception data, wherein the network perception data curve is a curve showing the network perception data changing with the time sequence number, and the interval between the time sequences corresponding to two adjacent time sequence numbers is a preset duration;

[0107] A curve fitting module 630 is configured to divide the network sensing data into multiple sets based on inflection points, where the common point between two adjacent sets is the inflection point, and to fit the network sensing data in each set to obtain a network sensing data curve;

[0108] An interval partitioning module 640 is configured to determine a target interval based on the left derivative and the right derivative of each point corresponding to each time sequence number in the network sensing data curve, where the start and end points of the target interval are points where the difference between the corresponding left derivative and the right derivative exceeds a preset difference standard;

[0109] The result generating module 650 is configured to output a user network quality identification result based on the target interval, where the user network quality identification result reflects the relationship between the network quality of the user equipment and the time period.

[0110] Figure 7 An example of a physical structure diagram of an electronic device is shown below. Figure 7 As shown, the electronic device may include: a processor (processor) 710, a communication interface (Communications Interface) 720, a memory (memory) 730 and a communication bus 740, wherein the processor 710, the communication interface 720, and the memory 730 communicate with each other through the communication bus 740. The processor 710 may call logic instructions in the memory 730 to execute a user network quality identification method, the method comprising: obtaining network perception data of a user device at multiple time points, the network perception data reflecting the amount of network downloaded data of the user device; determining an inflection point of a network perception data curve based on the network perception data and a time sequence number corresponding to the network perception data, the network perception data curve being a curve showing the network perception data changing with the time sequence number, wherein the interval between moments corresponding to two adjacent time sequence numbers is a preset duration; dividing the network perception data into multiple sets based on the inflection point, wherein a common point between two adjacent sets is the inflection point, and fitting the network perception data in each set to obtain the network perception data curve; determining a target interval based on a left derivative and a right derivative of a point corresponding to each time sequence number in the network perception data curve, wherein the start and end points of the target interval are points where the difference between the corresponding left derivative and right derivative is greater than a preset difference standard; and outputting a user network quality identification result based on the target interval, the user network quality identification result reflecting the relationship between the network quality of the user device and the time period.

[0111] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0112] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the user network quality identification method provided by the above methods, the method including: obtaining network perception data of the user device at multiple times, the network perception data reflecting the network download data volume of the user device; determining the inflection point of the network perception data curve based on the network perception data and the time sequence number corresponding to the network perception data, the network perception data curve is a curve of the network perception data changing with the time sequence number, and two adjacent The interval between the moments corresponding to the moment numbers is a preset duration; based on the inflection point, the network perception data is divided into multiple sets, the common point between two adjacent sets is the inflection point, and the network perception data in each set is fitted to obtain the network perception data curve; based on the left derivative and the right derivative of the point corresponding to each moment number in the network perception data curve, the target interval is determined, and the starting and ending points of the target interval are the points where the difference between the corresponding left derivative and right derivative is greater than the preset difference standard; based on the target interval, the user network quality identification result is output, and the user network quality identification result reflects the relationship between the network quality of the user device and the time period.

[0113] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the user network quality identification method provided by each of the above methods, the method comprising: obtaining network perception data of a user device at multiple time points, the network perception data reflecting the amount of network downloaded data of the user device; determining an inflection point of a network perception data curve based on the network perception data and the time sequence number corresponding to the network perception data, the network perception data curve being a curve showing the network perception data changing with the time sequence number, the interval between moments corresponding to two adjacent time sequence numbers being a preset duration; dividing the network perception data into multiple sets based on the inflection point, the common point between two adjacent sets being the inflection point, fitting the network perception data in each set to obtain the network perception data curve; determining a target interval based on the left derivative and the right derivative of each point corresponding to the time sequence number in the network perception data curve, the start and end points of the target interval being the points where the difference between the corresponding left derivative and right derivative is greater than a preset difference standard; and outputting a user network quality identification result based on the target interval, the user network quality identification result reflecting the relationship between the network quality of the user device and the time period.

[0114] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0115] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for identifying user network quality, characterized in that: include: Acquire network perception data of a user device at multiple times, where the network perception data reflects the amount of network downloaded data of the user device; Determining an inflection point of a network perception data curve based on the network perception data and the time sequence number corresponding to the network perception data, wherein the network perception data curve is a curve showing changes in the network perception data as the time sequence number changes, and an interval between moments corresponding to two adjacent time sequence numbers is a preset duration; Dividing the network perception data into multiple sets based on the inflection point, where a common point between two adjacent sets is the inflection point, and fitting the network perception data in each set to obtain the network perception data curve; Determine a target interval based on the left derivative and the right derivative of each point corresponding to each time sequence number in the network perception data curve, wherein the start and end points of the target interval are points where the difference between the corresponding left derivative and the right derivative is greater than a preset difference standard; A user network quality identification result is output based on the target interval, where the user network quality identification result reflects the relationship between the network quality of the user equipment and the time period.

2. The method for identifying user network quality according to claim 1, wherein: The determining the inflection point of the network perception data curve based on the network perception data and the time sequence number corresponding to the network perception data includes: Normalizing the network perception data and the time sequence number corresponding to the network perception data to obtain normalized network perception data and a normalized time sequence number; generating a normalized network perception data curve based on the normalized network perception data and the normalized time sequence number; Dividing the normalized network perception data curve into at least one sub-curve based on the directionality and concavity of the normalized network perception data curve, each sub-curve corresponding to only one directionality and one concavity; Updating the normalized network perception data based on the directionality and concavity of each sub-curve to obtain updated normalized network perception data, and obtaining an updated normalized network perception data curve based on the updated normalized network perception data, wherein the directionality of the updated normalized network perception data curve is increasing and the concavity is concave; Generating a differential curve based on the updated normalized network perception data curve, wherein the value corresponding to each normalized time sequence number in the differential curve is the difference between the updated normalized network perception data and the normalized time sequence number; The local maximum point in the differential curve is used as a reference point, the value corresponding to the local maximum point in the differential curve is greater than the values ​​corresponding to the two adjacent points on the left and right of the local maximum point in the differential curve, and the inflection point is determined based on the value corresponding to the reference point in the differential curve and the network perception data corresponding to the reference point.

3. The method for identifying user network quality according to claim 2, wherein: The updating of the normalized network perception data based on the directionality and concavity of each sub-curve to obtain updated normalized network perception data includes: When the directionality of the sub-curve is decreasing and the concavity is concave, updating each of the normalized network perception data in the sub-curve to a value that is reversely sorted according to the normalized time sequence number; When the directionality of the sub-curve is decreasing and the concavity is convex, updating the normalized network perception data in the sub-curve to a target difference value, where the target difference value is a difference between a target maximum value and the normalized network perception data, and the target maximum value is a maximum value among the normalized network perception data in the sub-curve; When the directionality of the sub-curve is increasing and the concavity is convex, updating the normalized network perception data in the sub-curve to a value obtained by reversely sorting the target difference values ​​according to the normalized time sequence number order; When the directionality of the sub-curve is increasing and the concavity is concave, the normalized network-sensed data in the sub-curve is maintained unchanged.

4. The method for identifying user network quality according to claim 1, wherein: The fitting of the network perception data in each of the sets to obtain the network perception data curve includes: Obtaining a network perception data sub-curve corresponding to each of the sets, and connecting the network perception data sub-curves corresponding to each of the sets to obtain the network perception data curve; Obtaining the network-sensing data sub-curve corresponding to the set includes: Using the network perception data in the set as a dependent variable and the time sequence number as an independent variable, fitting multiple univariate polynomial regression models, each of the univariate polynomial regression models having a different number of terms; The network-aware data sub-curve corresponding to the set is determined based on each of the univariate polynomial regression models.

5. The method for identifying user network quality according to claim 1, wherein: The determining of the target interval based on the left derivative and the right derivative of the point corresponding to each time sequence number in the network perception data curve includes: Setting the first target point as the starting point of the first interval and the second target point as the end point of the first interval; Setting the third target point as the starting point of the second interval and the fourth target point as the end point of the second interval; using the first interval and the second interval as the target interval; Among them, the left derivative of the first target point is less than the first preset threshold, and the difference between the right derivative and 0 is less than the second preset threshold, the left derivative of the second target point is greater than the third preset threshold, and the difference between the right derivative and 0 is less than the second preset threshold, the difference between the left derivative of the third target point and 0 is less than the second preset threshold, and the right derivative is greater than the third preset threshold, and the difference between the left derivative of the fourth target point and 0 is less than the second preset threshold, and the right derivative is less than the first preset threshold.

6. The method for identifying user network quality according to claim 1, wherein: Outputting the user network quality identification result based on the target interval includes: The user network quality identification result is output based on the target interval including the inflection point, wherein the user network quality in a time period corresponding to the target interval including the inflection point is lower than that in other time periods.

7. The method for identifying user network quality according to any one of claims 1 to 6, characterized in that: The network perception data is HTTP downstream traffic.

8. A user network quality identification device, characterized in that: include: A data acquisition module is used to obtain network perception data of the user equipment at multiple times, wherein the network perception data reflects the amount of network downloaded data of the user equipment; an inflection point determination module, configured to determine an inflection point of a network perception data curve based on the network perception data and a time sequence number corresponding to the network perception data, wherein the network perception data curve is a curve showing changes in the network perception data as the time sequence number changes, and an interval between moments corresponding to two adjacent time sequence numbers is a preset duration; a curve fitting module, configured to divide the network perception data into a plurality of sets based on the inflection point, wherein a common point between two adjacent sets is the inflection point, and to fit the network perception data in each set to obtain the network perception data curve; an interval partitioning module, configured to determine a target interval based on the left derivative and the right derivative of each point corresponding to each time sequence number in the network perception data curve, wherein the start and end points of the target interval are points where the difference between the corresponding left derivative and the right derivative is greater than a preset difference standard; The result generating module is configured to output a user network quality identification result based on the target interval, where the user network quality identification result reflects the relationship between the network quality of the user equipment and the time period.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the user network quality identification method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying user network quality according to any one of claims 1 to 7 is implemented.

Images