Network quality detection method and device, electronic equipment and storage medium

Abstract

The invention provides a network quality detection method and device, electronic equipment and a storage medium. The method comprises the following steps: collecting at least one network quality index when a target terminal uses a target network; obtaining a network quality index sequence according to the at least one network quality index; sequentially inputting the network quality index sequence into network experience detection models of N application scenes according to a sequence from small to large according to the network flow requirements of the application scenes until the network experience detection models output a first detection result, or the network experience detection model is the Nth network experience detection model; the first detection result is used for representing that the target terminal is stuck when using the target network in the application scene; according to the application scene represented by the first detection result, determining an experience score of using the target network by the target terminal; the experience score is used for representing the network quality of the target network. According to the invention, the accuracy of network quality detection is improved.

Description

Technical Field

[0001] The present application relates to network technology, and in particular to a network quality detection method, device, electronic device and storage medium. Background Art

[0002] With the development of network technology, wireless network application scenarios are increasing. Network quality has a great impact on user experience, so network quality detection is crucial. At present, the existing network quality detection methods are mainly based on the network quality indicator data of a single application scenario of the user terminal, and the network quality threshold set based on artificial experience corresponding to the network quality indicator data, to determine whether the current network is stuck.

[0003] However, the network is very complex and time-varying, and user terminals have different requirements for network quality in different application scenarios (such as watching videos, sending voice messages, video calls, etc.). Therefore, existing network quality detection methods have the problem of low accuracy. Summary of the invention

[0004] The present application provides a network quality detection method, device, electronic device and storage medium to overcome the problem of low accuracy of existing network quality detection methods.

[0005] In a first aspect, the present application provides a network quality detection method, the method comprising:

[0006] collecting at least one network quality indicator when the target terminal uses the target network;

[0007] Acquire a network quality indicator sequence according to the at least one network quality indicator;

[0008] Inputting the network quality indicator sequence into the network experience detection models of N application scenarios in order from small to large network traffic requirements of the application scenarios, until the network experience detection model outputs a first detection result, or the network experience detection model is the Nth network experience detection model; the first detection result is used to characterize that the target terminal is stuck when using the target network in the application scenario;

[0009] An experience score of the target terminal using the target network is determined according to the application scenario represented by the first detection result; the experience score is used to represent the network quality of the target network.

[0010] Optionally, determining, according to the application scenario represented by the first detection result, an experience score of the target terminal using the target network includes:

[0011] Determining, according to the application scenario represented by the first detection result, an experience level of the target terminal using the target network; the experience level is positively correlated with the network traffic demand of the application scenario;

[0012] The experience level of the target terminal using the target network is input into a preset experience score calculation formula to obtain an experience score of the target terminal using the target network.

[0013] Optionally, after obtaining the experience score of the target terminal using the target network, the method further includes:

[0014] According to the experience score of the target terminal using the target network, at least one of the experience score of the wireless access point providing the target network, the experience score of the area where the wireless access point is located, and the experience score of the entire network to which the area belongs is obtained and output.

[0015] Optionally, obtaining, according to the experience score of the target terminal using the target network, an experience score of a wireless access point providing the target network includes:

[0016] Acquire the number of target terminals connected to the wireless access point, the number of low experience level target terminals among the target terminals connected to the wireless access point, and the weight of the number of low experience level target terminals; the low experience level target terminals are target terminals whose experience level of using the target network is less than or equal to a preset experience level;

[0017] The experience score of the wireless access point is obtained according to the number of target terminals connected to the wireless access point, the number of target terminals with low experience levels, the weights, and the experience scores of the target terminals using the target network.

[0018] Optionally, obtaining the experience score of the area where the wireless access point is located according to the experience score of the target terminal using the target network includes:

[0019] Acquire the number of wireless access points included in the area where the wireless access point is located;

[0020] The experience score of the area where the wireless access point is located is obtained according to the number of the wireless access points and the experience score of each of the wireless access points.

[0021] Optionally, obtaining, according to the experience score of the target terminal using the target network, an experience score of the entire network to which the area belongs, includes:

[0022] For any area in the entire network, determine the experience score coefficient of the area according to the ratio of the number of target terminals with low experience level connected to all wireless access points in the area to the number of all target terminals connected to all wireless access points;

[0023] Obtain the experience score of the entire network to which the area belongs according to the experience score of the area and the experience score coefficient of the area.

[0024] Optionally, before sequentially inputting the network quality index sequence into the network experience detection models of N application scenarios in ascending order of the network traffic requirements of the application scenarios, the method further includes:

[0025] Judge whether the uplink and downlink traffic of the target terminal is greater than or equal to a preset uplink and downlink traffic threshold;

[0026] If so, sequentially input the network quality index sequence into the network experience detection models of N application scenarios in ascending order of the network traffic requirements of the application scenarios;

[0027] If not, determine that the experience level of the target terminal using the target network is the target second experience level; the target second experience level is lower than the experience level of the target terminal using the target network when the detection result output by any of the network experience detection models indicates that the target terminal is stuck in this application scenario.

[0028] In a second aspect, the present application provides a method for training a network experience detection model, and the method includes:

[0029] Obtain a sample data set of network quality indicators of a sample terminal in N application scenarios; where the sample data set corresponding to the jth application scenario includes: a positive sample network quality index sequence and a negative sample network quality index sequence in the jth application scenario, a negative sample network quality index sequence in the ith application scenario, a positive sample network quality index sequence in the kth application scenario, and the first labels of the negative sample network quality index sequences and the second labels of the positive sample network quality index sequences; i < j < k ≤ N; the network traffic requirements of the sample terminal in the ith application scenario, the jth application scenario, and the kth application scenario increase in sequence; the first label is used to represent that the network experience corresponding to the negative sample network quality index sequence in this application scenario is stuck; the second label is used to represent that the network experience corresponding to the positive sample network quality index sequence in this application scenario is not stuck;

[0030] For any of the above application scenarios, use the sample data set corresponding to the application scenario to train a preset model to obtain a network experience detection model for the application scenario; the network experience detection model is used to output a detection result indicating whether the target terminal lags when using the target network in this application scenario based on the network quality metric sequence when the target terminal uses the target network.

[0031] Optionally, the obtaining of the sample data set of the network quality metrics of the sample terminal in N application scenarios includes:

[0032] Obtain an initial sample data set of the network quality metrics of the sample terminal in the application scenario; the initial sample data set includes: a positive sample network quality metric sequence, a negative sample network quality metric sequence in the application scenario, and the first label and the second label;

[0033] Add the negative sample network quality metric sequence and the first label in the i-th application scenario to the initial sample data set corresponding to the j-th application scenario, and add the positive sample network quality metric sequence and the second label in the k-th application scenario to the initial sample data set corresponding to the j-th application scenario to obtain the sample data set corresponding to the j-th application scenario.

[0034] In a third aspect, the present application provides a network experience detection model training device, and the device includes:

[0035] An obtaining module, configured to obtain a sample data set of network quality metrics of a sample terminal in N application scenarios; wherein, the sample data set corresponding to the j-th application scenario includes: a positive sample network quality metric sequence and a negative sample network quality metric sequence in the j-th application scenario, a negative sample network quality metric sequence in the i-th application scenario, a positive sample network quality metric sequence in the k-th application scenario, and the first label of each negative sample network quality metric sequence and the second label of each positive sample network quality metric sequence; i < j < k ≤ N; the network traffic requirements of the sample terminal in the i-th application scenario, the j-th application scenario, and the k-th application scenario increase in sequence; the first label is used to indicate that the network experience corresponding to the negative sample network quality metric sequence in this application scenario is lagging; the second label is used to indicate that the network experience corresponding to the positive sample network quality metric sequence in this application scenario is not lagging;

[0036] A training module is used to train a preset model for any of the application scenarios using a sample data set corresponding to the application scenario to obtain a network experience detection model for the application scenario; the network experience detection model is used to output a detection result that characterizes whether the target terminal experiences lag when using the target network in the application scenario based on a network quality indicator sequence when the target terminal uses the target network.

[0037] In a fourth aspect, the present application provides a network quality detection device, the device comprising:

[0038] A collection module, used to collect at least one network quality indicator when the target terminal uses the target network;

[0039] An acquisition module, configured to acquire a network quality indicator sequence according to the at least one network quality indicator;

[0040] A first processing module is used to input the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large according to the network traffic demand of the application scenario, until the network experience detection model outputs a first detection result, or the network experience detection model is the Nth network experience detection model; the first detection result is used to characterize that the target terminal is stuck when using the target network in the application scenario;

[0041] The second processing module is used to determine an experience score of the target terminal using the target network according to the application scenario represented by the first detection result; the experience score is used to represent the network quality of the target network.

[0042] In a fifth aspect, the present application provides an electronic device, the electronic device comprising a memory and a processor;

[0043] The memory stores a computer program;

[0044] The processor is configured to execute the method according to any one of the first aspect and / or the second aspect through the computer program.

[0045] In a sixth aspect, the present application provides a computer-readable storage medium having computer execution instructions stored thereon. When the computer execution instructions are executed by a processor, the method described in the first aspect and / or the second aspect is implemented.

[0046] In a seventh aspect, the present application provides a computer program product, comprising a computer program, which implements the method described in the first aspect and / or the second aspect when executed by a processor.

[0047] The network quality detection method, device, electronic device and storage medium provided by the present application do not need to determine the application scenario of the target terminal, and predict the network quality through at least one network quality indicator when the target terminal uses the target network. Through the above method, the network quality prediction method is not limited to the application scenario of the target terminal, and predicts the network quality based on the network experience detection model of a variety of different application scenarios, reflecting the experience of the user terminal in different application scenarios, and does not need to rely on manual experience to set the network quality threshold to determine the network quality, thereby improving the accuracy of network quality detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be 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 application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0049] Figure 1 A flowchart of a network experience detection model training method provided in this application;

[0050] Figure 2 A schematic diagram of a sample data set provided for this application;

[0051] Figure 3 A schematic diagram of network quality indicator collection provided for this application;

[0052] Figure 4 A flowchart of a network quality detection method provided by this application;

[0053] Figure 5 A schematic diagram of a process for determining an experience level provided for this application;

[0054] Figure 6 A schematic diagram of an application scenario of the network quality detection method provided in this application;

[0055] Figure 7 A schematic diagram of the structure of a network experience detection model training device provided in this application;

[0056] Figure 8 A schematic diagram of the structure of a network quality detection device provided in this application;

[0057] Fig. 9 A schematic diagram of the structure of an electronic device provided in this application.

[0058] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0060] With the development of network technology, wireless networks (such as Wi-Fi) are covering more and more scenarios. In some large and medium-sized wireless network coverage scenarios, channel planning, co-channel interference, network delay, etc. may affect the network quality of the wireless network.

[0061] At present, the existing network quality detection method mainly determines the network quality of the user terminal connection based on the network quality indicator data of a single application scenario of the user terminal. When using the existing network quality detection method to detect the network quality of the network connected to the user terminal, the server can obtain the network quality indicator data when the application scenario of the user terminal is watching videos (or other single application scenarios), such as packet loss rate, etc. Then, the server can determine whether the current network is stuck based on the network quality indicator data and the fixed network quality threshold corresponding to the network quality indicator data.

[0062] However, the network is very complex and time-varying, and user terminals have different requirements for network quality in different application scenarios (such as watching videos, sending voice messages, video calls, etc.). In other words, the above-mentioned network quality indicator data based on a single application scenario of the user terminal and the detection method of determining network quality with a fixed network quality threshold do not reflect the actual use of the wireless network and cannot be applied to the actual use of the network in multiple application scenarios. Therefore, the existing network quality detection method has the problem of low accuracy.

[0063] Taking into account the above-mentioned problems existing in the existing network quality detection methods, the present application proposes a network experience detection model based on different application scenarios, which detects whether the user terminal is stuck when using the target network in different application scenarios, and outputs the user experience score of the target network based on the result of whether the user terminal is stuck when using the target network in different application scenarios. The user experience score can be used to characterize the network quality of the target network. Through the above method, various application scenarios of the user terminal are covered, thereby reflecting the experience of the user terminal in different application scenarios, and there is no need to rely on manual experience to set the network quality threshold to determine the network quality, thereby improving the accuracy of network quality detection.

[0064] The network experience detection models corresponding to the above different application scenarios can be obtained by training based on sample network quality indicator data of different application scenarios. The execution subject of the network experience detection model training method can be any electronic device with processing capability, or cloud, or server cluster device, etc., and this application does not limit this.

[0065] The following first takes the execution subject of the network experience detection model training method as an electronic device as an example, and describes in detail the technical solution of "network experience detection model training" provided by this application in combination with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0066] Figure 1 A flowchart of a network experience detection model training method provided in this application. Figure 1 As shown, the method comprises the following steps:

[0067] S101. Obtain a sample data set of network quality indicators of sample terminals in N application scenarios.

[0068] Optionally, the present application does not limit the types of the above sample terminals, nor the number of sample terminals. The sample terminal can be any terminal device that can connect to a wireless network, such as a mobile phone, a tablet computer, a laptop computer, a desktop computer, etc. Optionally, the present application does not limit the above application scenarios. Exemplarily, the application scenarios may include, for example: viewing virtual reality (VR) information, viewing augmented reality (AR) information, video conferencing, live broadcast, watching videos, sending voice, viewing web pages, sending texts, etc.

[0069] Among them, the sample data set corresponding to the j-th application scenario may include: the positive sample network quality index sequence and the negative sample network quality index sequence in the j-th application scenario, the negative sample network quality index sequence in the i-th application scenario, the positive sample network quality index sequence in the k-th application scenario, and the first label of each of the above negative sample network quality index sequences and the second label of each positive sample network quality index sequence. Among them, i < j < k ≤ N. The network traffic requirements of the sample terminal in the i-th application scenario, the j-th application scenario, and the k-th application scenario increase in sequence. The above first label can be used to characterize that the network experience corresponding to the negative sample network quality index sequence in this application scenario is lagging. The above second label can be used to characterize that the network experience corresponding to the positive sample network quality index sequence in this application scenario is not lagging.

[0070] Optionally, the network quality index sequence may include the values of at least one network quality index. Exemplarily, the network quality index sequence may, for example, include the values of at least one of the following network quality indexes: such as packet loss rate, retransmission rate, channel occupancy ratio, bandwidth, delay, delay jitter, throughput, delay-bandwidth product, rate, round-trip time (RTT), uplink and downlink rate, effective data rate, background noise, etc.

[0071] Among them, the above positive sample network quality index sequence is a sequence composed of each network quality index collected when the sample terminal uses the network without lag in the corresponding application scenario. The above negative sample network quality index sequence is a sequence composed of each network quality index collected when the sample terminal uses the network with lag in the corresponding application scenario. It should be understood that this application does not limit the number of network quality indexes included in the positive sample network quality index sequence and the negative sample network quality index sequence.

[0072] Exemplarily, taking the above application scenario and the network traffic requirements of the application scenario from large to small as follows: watching VR / AR > video conferencing > live broadcast > watching video > sending voice > viewing web page > sending text, Figure 2 is a schematic diagram of a sample data set provided by this application. As Figure 2As shown, taking the j-th application scenario of watching videos as an example, the i-th application scenario can be at least one of sending voice, browsing web pages, and sending text, and the k-th application scenario can be at least one of live broadcast, video conferencing, and watching VR / AR. Then the sample data set corresponding to the video watching application scenario may include: a positive sample network quality indicator sequence and a negative sample network quality indicator sequence in the video watching application scenario, a negative sample network quality indicator sequence in at least one application scenario of sending voice, browsing web pages, and sending text, and a positive sample network quality indicator sequence in at least one application scenario of live broadcast, video conferencing, and watching VR / AR.

[0073] For the application scenario with the largest network traffic demand among the N application scenarios, that is, the Nth application scenario, optionally, the sample data set corresponding to the application scenario may include: a positive sample network quality indicator sequence and a negative sample network quality indicator sequence under the Nth application scenario, and a negative sample network quality indicator sequence under at least one other application scenario.

[0074] For the application scenario with the smallest network traffic demand among the N application scenarios, that is, the first application scenario, optionally, the sample data set corresponding to the application scenario may include: a positive sample network quality indicator sequence and a negative sample network quality indicator sequence under the first application scenario, and a positive sample network quality indicator sequence under at least one other application scenario.

[0075] S102: For any application scenario, use a sample data set corresponding to the application scenario to train a preset model to obtain a network experience detection model for the application scenario.

[0076] The above network experience detection model is used to output a detection result that characterizes whether the target terminal experiences lag when using the target network in the application scenario based on a network quality indicator sequence when the target terminal uses the target network.

[0077] Optionally, the preset models corresponding to different application scenarios may be the same or different, and the present application does not limit this. Exemplarily, the preset model may be any existing neural network model, and the present application does not limit this. Alternatively, the preset model may also be, for example, a decision tree model. Taking the preset model as a decision tree model as an example, the electronic device uses a sample data set corresponding to the application scenario to train the implementation method of the decision tree model, and can refer to any existing decision tree model training method to obtain a network experience detection model for the application scenario. In this implementation method, each node in the decision tree model can be used to characterize a network quality indicator in a network quality indicator sequence.

[0078] It should be understood that for any application scenario, this application does not limit how the electronic device uses the sample data set corresponding to the application scenario to train the preset model to obtain the network experience detection model of the application scenario. Optionally, any existing model training method can be referred to, and this application will not be repeated here.

[0079] In this embodiment, for any application scenario, the negative sample data in the sample data set of the application scenario includes: its own negative sample network quality indicator sequence, and the negative sample network quality indicator sequence of the application scenario with a smaller network traffic demand than its own; the positive sample data includes: its own positive sample network quality indicator sequence, and the positive sample network quality indicator sequence with a larger network traffic demand than its own. Through the above method, the sample data set of each application scenario can include a rich and comprehensive sample network quality indicator sequence in a stuck / non-stuck state, thereby improving the universality and accuracy of the network experience detection model trained based on the sample data set. Therefore, the accuracy of the subsequent use of the network experience detection model can be improved to predict the detection result used to characterize whether the target terminal is stuck when using the target network in the application scenario based on the network quality indicator sequence when the target terminal uses the target network.

[0080] The following is a detailed description of how the electronic device obtains a sample data set of network quality indicators of a sample terminal in N application scenarios:

[0081] As a possible implementation method, for any application scenario, the electronic device may first obtain an initial sample data set of network quality indicators of the sample terminal in the application scenario. The initial sample data set may include: a positive sample network quality indicator sequence and a negative sample network quality indicator sequence in the application scenario, as well as a first label of the negative sample network quality indicator sequence and a second label of the positive sample network quality indicator sequence.

[0082] Optionally, for any application scenario, the user may, for example, operate the sample terminal in the application scenario (eg, use the sample terminal to watch a video). Figure 3 A schematic diagram of network quality indicator collection provided by this application. Figure 3 As shown, for example, with the sample terminal (eg Figure 3 The wireless access point (Access Point, AP) connected to the mobile phone and computer shown in the figure can collect the network quality indicators of the sample terminal and send them to the cloud database through the access controller (Access Controller, AC) connected to the Ap.

[0083] Optionally, the electronic device can obtain the network quality indicator of the sample terminal in the application scenario from the database, and form a network quality indicator sequence based on at least one network instruction indicator. Then, the electronic device can receive the label for each network quality indicator sequence triggered by the user. If the label corresponding to the network quality indicator sequence is a first label (for example, 0), the network quality indicator sequence is a negative sample network quality indicator sequence in the application scenario; if the label corresponding to the network quality indicator sequence is a second label (for example, 1), the network quality indicator sequence is a positive sample network quality indicator sequence in the application scenario.

[0084] Alternatively, the electronic device may also receive the initial sample data sets corresponding to each application scenario input by the user through an application programming interface (Application Programming Interface, API) or a graphical user interface (Graphical User Interface, GUI).

[0085] Then, the electronic device can add the negative sample network quality indicator sequence under the i-th application scenario and the first label of the negative sample network quality indicator sequence to the initial sample data set corresponding to the j-th application scenario, and add the positive sample network quality indicator sequence under the k-th application scenario and the second label of the positive sample network quality indicator sequence to the initial sample data set corresponding to the j-th application scenario to obtain the sample data set corresponding to the j-th application scenario.

[0086] Optionally, the electronic device may, for example, add all negative sample network quality indicator sequences in the i-th application scenario to the initial sample data set corresponding to the j-th application scenario. Alternatively, the electronic device may, for example, randomly extract a preset number of negative sample network quality indicator sequences from all negative sample network quality indicator sequences in the i-th application scenario, and add them to the initial sample data set corresponding to the j-th application scenario.

[0087] Optionally, the electronic device may, for example, add all positive sample network quality indicator sequences in the kth application scenario to the initial sample data set corresponding to the jth application scenario. Alternatively, the electronic device may, for example, randomly extract a preset number of positive sample network quality indicator sequences from all positive sample network quality indicator sequences in the kth application scenario, and add them to the initial sample data set corresponding to the jth application scenario.

[0088] In this embodiment, it is achieved that the sample data set corresponding to the j-th application scenario includes the negative sample network quality indicator sequence under the ith application scenario and the positive sample network quality indicator sequence under the k-th application scenario, thereby improving the richness of the sample data set corresponding to the j-th application scenario.

[0089] As another possible implementation manner, the electronic device may also receive, through an API or a GUI, a sample data set of network quality indicators of a sample terminal in N application scenarios input by a user.

[0090] The network experience detection model of each application scenario described above can be deployed in the cloud, so the cloud can use the network experience detection model of each application scenario to perform network quality detection. In some embodiments, the execution subject of the network quality detection method can also be any electronic device with processing capabilities. The electronic device and the execution subject of the aforementioned network experience detection model training method can be the same electronic device or different electronic devices.

[0091] The following takes the execution subject of the network quality detection method as the cloud as an example, and describes in detail the technical solution of "network quality detection" provided by this application in combination with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0092] Figure 4 A flow chart of a network quality detection method provided in this application. Figure 4 As shown, the method comprises the following steps:

[0093] S201. Collect at least one network quality indicator when a target terminal uses a target network.

[0094] Optionally, the present application does not limit the type of the above-mentioned target terminal. The target terminal can be any terminal device that can connect to a wireless network. In some embodiments, the target terminal can also be a sensor in a wireless network. The target network can be, for example, a mobile communication network such as wifi, the fifth generation mobile communication technology (5th Generation Mobile Communication Technology, 5G).

[0095] Optionally, the cloud may, for example, collect at least one network quality indicator when the target terminal uses a target network provided by the AP to which the target terminal is connected.

[0096] S202: Acquire a network quality indicator sequence according to at least one network quality indicator.

[0097] Optionally, the cloud may, for example, add the at least one network quality indicator to the same sequence as a network quality indicator sequence.

[0098] S203. Input the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large according to the network traffic demand of the application scenario, until the network experience detection model outputs the first detection result, or the network experience detection model is the Nth network experience detection model.

[0099] The first detection result is used to characterize that the target terminal is stuck when using the target network in the application scenario. The network experience detection models for the N application scenarios may be trained based on the network experience detection model training method described in any of the aforementioned embodiments.

[0100] Exemplarily, take N application scenarios, and the network traffic requirements of the N application scenarios are as follows from large to small: watching VR / AR> video conferencing> live broadcast> watching video> sending voice> watching web pages> sending text. The cloud can first input the above network quality indicator sequence into the network experience detection model of the text sending scene to detect whether the text sending scene is stuck. If the detection result is used to characterize the target terminal using the target network in the text sending scenario (hereinafter referred to as the detection result characterizing the stuck), the following step S204 can be executed. If there is no stuck, the above network quality indicator sequence is input into the network experience detection model of the web page watching scene.

[0101] If the network experience detection model for the webpage scenario determines that the detection result indicates a jam, the following step S204 may be performed. If not, the network quality indicator sequence is input into the network experience detection model for the voice sending scenario.

[0102] If the network experience detection model for the voice sending scenario determines that the detection result represents a jam, the following step S204 may be performed. If there is no jam, the network quality indicator sequence is input into the network experience detection model for the video watching scenario.

[0103] If the network experience detection model for watching a video scene determines that the detection result represents a freeze, the following step S204 may be performed. If there is no freeze, the above network quality indicator sequence is input into the network experience detection model for the live broadcast scene.

[0104] If the network experience detection model of the live broadcast scenario determines that the detection result represents a jam, the following step S204 may be performed. If not, the network quality indicator sequence is input into the network experience detection model of the video conference scenario.

[0105] If the network experience detection model of the video conferencing scene determines that the detection result represents a jam, the following step S204 may be performed. If there is no jam, the above network quality indicator sequence is input into the network experience detection model of the VR / AR scene.

[0106] If the network experience detection model of the VR / AR scene determines that the detection result represents lag, or no lag, because the network experience detection model of the VR / AR scene is the Nth network experience detection model, the cloud can execute the following step S204.

[0107] S204: Determine an experience score of the target terminal using the target network according to the application scenario represented by the first detection result.

[0108] The experience score can be used to characterize the network quality of the target network. For example, the higher the experience score of the target terminal using the target network, the better the network quality of the target network. The lower the experience score of the target terminal using the target network, the worse the network quality of the target network.

[0109] Optionally, the cloud can determine the experience score of the target network based on the application scenario represented by the first detection result and the mapping relationship between the application scenario and the experience score. The mapping relationship can be pre-stored in the cloud, for example. The experience score can be positively correlated with the "network traffic demand of the application scenario corresponding to the network experience detection model that outputs the first detection result".

[0110] Alternatively, the cloud may first determine the experience level of the target terminal using the target network based on the application scenario represented by the first detection result, for example. The experience level may be positively correlated with the network traffic demand of the application scenario. Then, the cloud may determine the experience score of the target terminal using the target network based on the experience level of the target terminal using the target network.

[0111] In this embodiment, it is not necessary to determine the application scenario of the target terminal, and the network quality is predicted by at least one network quality indicator when the target terminal uses the target network. Through the above method, the network quality prediction method is not limited to the application scenario of the target terminal, and the network quality is predicted based on the network experience detection model of a variety of different application scenarios, covering a variety of application scenarios from low traffic to high traffic intervals, reflecting the experience of the user terminal in different application scenarios, and does not need to rely on manual experience to set the network quality threshold to determine the network quality, thereby improving the accuracy of network quality detection. In addition, taking the network quality detection result as an example of providing data support for network operation and maintenance, the above method improves the accuracy of network quality detection, thereby improving the accuracy of active operation and maintenance based on the network quality.

[0112] In some embodiments, taking the aforementioned preset model as a decision tree model as an example, that is to say, the network experience detection model of the application scenario can be obtained by training based on the decision tree model. In this implementation, the cloud can also determine the target terminal's use of the target network in the application scenario based on the network quality indicator sequence when the target terminal uses the target network, and output the reason for the target network to be stuck in the application scenario, such as outputting the network quality indicator that causes the target network to be stuck in the application scenario. For example, the cloud can output the path from the root node to the leaf node of the decision tree model, and each node in the path includes a network quality indicator. The network quality indicator in the path is the network quality indicator that causes the target network to be stuck in the application scenario, so the cloud can use these network quality indicators as the cause of the failure of the target network, such as a large packet loss rate, high latency, etc. By outputting the cause of the failure of the target network, the user can quickly locate the failure, further improving the accuracy of active operation and maintenance based on the network quality, and further improving the user experience.

[0113] The following is a detailed description of how the cloud determines the experience score of the target terminal using the target network according to the application scenario represented by the first detection result:

[0114] As a possible implementation method, the cloud can determine the experience level of the target terminal using the target network based on the application scenario represented by the first detection result. The experience level is positively correlated with the network traffic demand of the application scenario. Then, the cloud can input the experience level of the target terminal using the target network into a preset experience score calculation formula to obtain the experience score of the target terminal using the target network.

[0115] Optionally, the cloud may determine the experience level of the target terminal using the target network based on the application scenario represented by the first detection result and the mapping relationship between the network experience detection model and the experience level.

[0116] For example, take N application scenarios, and the network traffic requirements of the N application scenarios are as follows from large to small: watching VR / AR> video conferencing> live broadcast> watching video> sending voice> watching web pages> sending text, Figure 5 A schematic diagram of the experience level determination process provided by this application. Figure 5 As shown, if the detection result of the network experience detection model for the text sending scenario is used to characterize that the target terminal uses the target network in the text sending scenario (text sending is stuck), then the experience level of the target terminal using the target network is determined to be experience level 1.

[0117] If the detection result of the network experience detection model for the web page opening scenario is used to characterize the lag when opening the web page, it is determined that the experience level of the target terminal using the target network is experience level 2.

[0118] If the detection result of the network experience detection model for the voice sending scenario is used to characterize the jamming of voice sending, it is determined that the experience level of the target terminal using the target network is experience level 3.

[0119] If the network experience detection model for the video watching scenario determines that the detection result is used to characterize video watching freezes, then the experience level of the target terminal using the target network is determined to be experience level 4.

[0120] If the network experience detection model for the live broadcast scenario determines that the detection result is used to characterize the live broadcast freeze, then the experience level of the target terminal using the target network is determined to be experience level 5.

[0121] If the network experience detection model of the video conferencing scenario determines that the detection result is used to characterize the video conferencing freeze, then the experience level of the target terminal using the target network is determined to be experience level 6.

[0122] If the network experience detection model of the VR / AR scenario determines that the detection result is used to characterize VR / AR lag, then the experience level of the target terminal using the target network is determined to be experience level 7.

[0123] Exemplarily, the above preset experience score calculation formula may be as shown in the following formula (1):

[0124]

[0125] Wherein, score represents the experience score of the target terminal using the target network. L represents the experience level of the target terminal using the target network. pred can be, for example, a preset value between 0 and 1.

[0126] In some embodiments, if the network experience detection model is the Nth network experience detection model, and the detection result output by the network experience detection model is used to characterize that the target terminal uses the target network without lag in the application scenario, then the target terminal can determine that the experience level of the target network is the target first experience level.

[0127] Because the Nth network experience detection model is the network experience detection model corresponding to the application scenario with the highest network traffic demand among the network experience detection models corresponding to N application scenarios, if the detection result output by the network experience detection model is used to characterize that the target terminal uses the target network without lag in the application scenario, it means that the network quality of the target network is good. Therefore, the target first experience level can be higher than the detection result output by the Nth network experience detection model, which is used to characterize the experience level of the target terminal using the target network when the target terminal lags when using the target network in the application scenario.

[0128] For example, Figure 5For example, assuming that the Nth network experience detection model is a network experience detection model for a VR / AR scenario, the detection result output by the network experience detection model for the VR / AR scenario is used to characterize that when the target terminal uses the target network without lag in the application scenario, the experience level of the target network can be experience level 8.

[0129] In this embodiment, when the network quality of the target network is good, the experience level of the target network is high, and when the network quality of the target network is poor, the experience level of the target network is low, thereby improving the accuracy of determining the experience level of the target network, and further improving the accuracy of determining the experience score used to characterize the network quality based on the experience level.

[0130] Furthermore, as a possible implementation method, after obtaining the experience score of the target terminal using the target network, the cloud can also obtain and output at least one of the experience score of the wireless access point providing the target network, the experience score of the area where the wireless access point is located, and the experience score of the entire network to which the area belongs, based on the experience score of the target terminal using the target network.

[0131] It should be understood that the area where the wireless access point is located can be predetermined by the user. A region can include one or more ACs. Each AC can be connected to at least one AP. The area where the at least one AP is located is the region. It should be understood that the entire network to which the region belongs can refer to the entire local area network to which the region belongs. For example, an internal local area network of an enterprise can be called the entire network, and the network of a building or a floor of the enterprise can be called a region.

[0132] For example, Figure 6 A schematic diagram of an application scenario of the network quality detection method provided in this application. Figure 6 As shown, terminals 1 to 5 can all be the above-mentioned target terminals. Among them, terminals 1 and 2 are connected to the same AP (ap1), terminal 4 is connected to ap2, and terminals 3 and 5 are connected to the same AP (ap3). Ap1 and ap2 are in the same area (area 1), and ap3 is in area 2. Area 1 and area 2 are in the same entire network.

[0133] Exemplarily, the cloud can output to the target electronic device at least one of the experience score of the wireless access point providing the target network, the experience score of the area where the wireless access point is located, and the experience score of the entire network to which the area belongs. Then, the target electronic device can, for example, display the experience score of the wireless access point providing the target network, the experience score of the area where the wireless access point is located, and the experience score of the entire network to which the area belongs, so that the user can view the experience scores of the above different dimensions. It should be understood that the present application does not limit the operations performed by the user based on the experience scores of the different dimensions. For example, the user can determine whether the target network needs to be optimized based on the experience scores of the different dimensions.

[0134] By outputting the experience scores in different dimensions, the user can perform subsequent operations based on the experience scores in different dimensions, thereby improving the user experience and the practicability of the network quality detection method.

[0135] In some embodiments, the cloud may further output the experience score of the target terminal using the target network after obtaining the experience score of the target terminal using the target network.

[0136] Through the above method, the experience scores of AP, region and whole network can be determined, which characterizes the network quality in each dimension. Taking the network quality detection before and after network RF optimization as an example, the experience scores of different dimensions can well reflect the quality of RF optimization strategy.

[0137] Taking the example of the cloud obtaining the experience score of the wireless access point providing the target network according to the experience score of the target terminal using the target network, optionally, the cloud may first obtain the number of target terminals connected to the wireless access point, the number of target terminals with low experience levels among the target terminals connected to the wireless access point, and the weight of the number of target terminals with low experience levels. Then, the cloud may obtain the experience score of the wireless access point according to the number of target terminals connected to the wireless access point, the number of target terminals with low experience levels, the weight, and the experience score of the target terminal using the target network.

[0138] The above weights can increase the impact of the number of low experience level target terminals on the experience score of the wireless access point. The low experience level target terminal is a target terminal whose experience level of using the target network is less than or equal to the preset experience level. In other words, the low experience level target terminal is a target terminal whose experience level of using the target network is within the preset level range. For example, Figure 5 Taking the experience level 1 to experience level 8 shown in as an example, the preset level range may be, for example, experience level [1, 2, 3].

[0139] Optionally, the cloud can determine the number of target terminals connected to each wireless access point by, for example, querying a database. Alternatively, the cloud can receive the number of target terminals connected to the wireless access point from the wireless access point. Optionally, the cloud can also determine the identifiers of the target terminals connected to the wireless access point by, for example, querying a database. Then, based on the identifiers of the target terminals connected to the wireless access point, the cloud can determine the low experience level target terminals among all the low experience level target terminals connected to the wireless access point, and determine the number of the low experience level target terminals. Optionally, the weight of the number of the low experience level target terminals can be, for example, pre-stored in the cloud.

[0140] Exemplarily, the cloud can obtain the experience score of the wireless access point based on the number of target terminals connected to the wireless access point, the number of target terminals with low experience levels, weights, and the experience score of the target terminal using the target network, for example, by the following formula (2):

[0141]

[0142] Among them, score ap Indicates the experience score of the wireless access point, device num Indicates the number of target terminals connected to the wireless access point, score i Indicates the experience score of each target terminal using the target network. bad num represents the number of target terminals with low experience level. coef represents the weight (or weight coefficient) of the number of target terminals with low experience level, and the weight coefficient can better represent the influence of target terminals with low experience level on the experience score of the wireless access point.

[0143] Through the above method, the experience score of the wireless access point dimension is obtained based on the number of target terminals connected to the wireless access point, the number of target terminals with low experience levels, the weights, and the experience score of the target terminals using the target network, and network quality detection in the wireless access point (AP) dimension is realized, which improves the flexibility and scope of application of the network quality detection method and improves the detection dimension of the network quality detection.

[0144] Taking the example of obtaining the experience score of the area where the wireless access point is located according to the experience score of the target terminal using the target network, optionally, the cloud can first obtain the number of wireless access points included in the area where the wireless access point is located. Then, the cloud can obtain the experience score of the area where the wireless access point is located according to the number of wireless access points and the experience score of each wireless access point.

[0145] Exemplarily, the cloud may determine the area where each wireless access point is located, and determine the number of wireless access points included in the area where each wireless access point is located, for example, by querying a database.

[0146] Exemplarily, the cloud can obtain the experience score of the area where the wireless access point is located based on the number of wireless access points and the experience score of each wireless access point by using the following formula (3):

[0147]

[0148] Among them, score area Indicates the experience score of the area where the wireless access point is located. num Indicates the number of wireless access points in the area. Indicates the experience score of each wireless access point in the area.

[0149] Through the above method, the experience score of the area where the wireless access point is located is obtained based on the number of wireless access points included in the area where the wireless access point is located and the experience score of each wireless access point, thereby realizing network quality detection in the regional dimension, further improving the flexibility and scope of application of the network quality detection method, and further improving the detection dimension of the network quality detection.

[0150] Through the above method, the experience score of each dimension is determined based on the number of target terminals with low experience levels, so that the final experience score of each dimension can have a certain gap even when the total number of target terminals is the same, reflecting the network quality differences between regions and APs.

[0151] Taking the experience score of the entire network to which the above-mentioned area belongs as an example, based on the experience score of the target terminal using the target network, optionally, the cloud can first determine the experience score coefficient of any area in the entire network based on the ratio of the number of low-experience-level target terminals connected to all wireless access points included in the area to the number of all target terminals connected to all wireless access points.

[0152] The experience score coefficient of the area may be negatively correlated with the above ratio, for example. In some embodiments, the cloud may determine the experience score coefficient of the area based on the above ratio, and the mapping relationship between the ratio range and the experience score coefficient of the area. Exemplarily, the mapping relationship between the ratio range and the experience score coefficient of the area may be shown in Table 1 below:

[0153] Table 1

[0154] Ratio range Experience score coefficient of the region Greater than or equal to 1 / 10 0.8 Less than 1 / 10 and greater than 1 / 50 0.9 Less than or equal to 1 / 50 1

[0155] Assuming that the ratio of the number of target terminals with low experience levels connected to all wireless access points in the area to the number of all target terminals connected to all wireless access points is 1 / 10, the cloud can determine that the experience score coefficient of the area is 0.8.

[0156] Then, the cloud can obtain the experience score of the entire network to which the area belongs based on the experience score of the area and the experience score coefficient of the above area, for example.

[0157] Exemplarily, the cloud can implement the experience score based on the region and the experience score coefficient of the region through the following formula (4) to obtain the experience score of the entire network to which the region belongs:

[0158]

[0159] Among them, area num Indicates the number of regions in the entire network. represents the experience score of the ith region, coef i Represents the experience score coefficient of the i-th area.

[0160] Through the above method, the experience score coefficient of each area is determined based on the ratio of the number of low experience level target terminals connected to all wireless access points included in each area to the number of all target terminals connected to all wireless access points, and the experience score of the entire network to which the area belongs is obtained based on the experience score of the area and the experience score coefficient of the above area, thereby realizing network quality detection in the dimension of the entire network, further improving the flexibility and scope of application of the network quality detection method, and further improving the detection dimension of network quality detection.

[0161] As a possible implementation method, before the cloud inputs the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large according to the network traffic requirements of the application scenarios, it can also determine whether the uplink and downlink traffic of the target terminal is greater than or equal to the preset uplink and downlink traffic thresholds.

[0162] The uplink and downlink traffic may be a network quality indicator. That is, after obtaining at least one network quality indicator including uplink and downlink traffic when the target terminal uses the target network, the cloud may determine whether the uplink and downlink traffic of the target terminal is greater than or equal to a preset uplink and downlink traffic threshold.

[0163] Exemplarily, the preset uplink and downlink traffic thresholds may be pre-stored in the cloud, for example.

[0164] If the uplink and downlink traffic of the target terminal is greater than or equal to the preset uplink and downlink traffic threshold, it means that the target terminal is more active in using the target network. Then the cloud can perform the operation of "inputting the network quality indicator sequence into the network experience detection model of N application scenarios in the order of small to large network traffic requirements of the application scenarios". Figure 5 For example, when the cloud determines through the above activity judgment that the target terminal is active (that is, determines that the uplink and downlink traffic of the target terminal is greater than or equal to the preset uplink and downlink traffic threshold), it executes the subsequent step of inputting the network quality indicator sequence into the network experience detection model corresponding to the text sending scenario.

[0165] If the uplink and downlink traffic of the target terminal is less than the preset uplink and downlink traffic threshold, it means that the target terminal is not active in using the target network. Then the cloud can determine that the experience level of the target terminal using the target network is the target second experience level. Among them, the target second experience level is lower than "the detection result output by any network experience detection model is used to characterize the experience level of the target terminal using the target network when the target terminal is stuck in using the target network in this application scenario".

[0166] For example, Figure 5 For example, the cloud can determine that the target terminal is inactive through activity judgment (that is, it is determined that the uplink and downlink traffic of the target terminal is less than the preset uplink and downlink traffic threshold), and then determine that the experience level of the target network is 0.

[0167] Through the above method, the cloud can execute the subsequent network quality detection method when it is determined that the target terminal is actively using the target network, ensuring that the network quality detection result is determined when the target terminal uses the target network, improving the authenticity of the network quality detection result, and thus improving the accuracy of network quality detection. When it is determined that the target terminal is inactive, the experience level of the target network can be directly determined as the target second experience level, improving the efficiency of determining the experience level of the target network.

[0168] Figure 7 A schematic diagram of the structure of a network experience detection model training device provided in this application. Figure 7 As shown, the device includes: an acquisition module 31 and a training module 32. Among them,

[0169] The acquisition module 31 is used to acquire sample data sets of network quality indicators of sample terminals in N application scenarios.

[0170] Among them, the sample data set corresponding to the j-th application scenario includes: the positive sample network quality index sequence and the negative sample network quality index sequence in the j-th application scenario, the negative sample network quality index sequence in the i-th application scenario, the positive sample network quality index sequence in the k-th application scenario, and the first label of each negative sample network quality index sequence and the second label of each positive sample network quality index sequence; i < j < k ≤ N; the network traffic demands of the sample terminal in the i-th application scenario, the j-th application scenario, and the k-th application scenario increase in sequence; the first label is used to represent that the network experience corresponding to the negative sample network quality index sequence in this application scenario is stuck; the second label is used to represent that the network experience corresponding to the positive sample network quality index sequence in this application scenario is not stuck.

[0171] The training module 32 is used to train a preset model for any one of the application scenarios using the sample data set corresponding to the application scenario, and obtain a network experience detection model for the application scenario. Among them, the network experience detection model is used to output a detection result for representing whether the target terminal is stuck when using the target network in this application scenario based on the network quality index sequence when the target terminal uses the target network.

[0172] Optionally, the obtaining module 31 is specifically configured to obtain an initial sample data set of the network quality index of the sample terminal in the application scenario; add the negative sample network quality index sequence and the first label in the i-th application scenario to the initial sample data set corresponding to the j-th application scenario, and add the positive sample network quality index sequence and the second label in the k-th application scenario to the initial sample data set corresponding to the j-th application scenario, to obtain the sample data set corresponding to the j-th application scenario. Among them, the initial sample data set includes: the positive sample network quality index sequence, the negative sample network quality index sequence, and the first label and the second label in this application scenario.

[0173] The network experience detection model training device provided by the present application is used to execute the foregoing embodiments of the network experience detection model training method, and its implementation principle and technical effects are similar, which will not be elaborated here.

[0174] Figure 8 This is a schematic structural diagram of a network quality detection device provided by the present application. As Figure 8 shown, the device includes: a collection module 41, an acquisition module 42, a first processing module 43, and a second processing module 44. Among them,

[0175] The collection module 41 is used to collect at least one network quality index when the target terminal uses the target network.

[0176] The acquisition module 42 is configured to acquire a network quality indicator sequence according to the at least one network quality indicator.

[0177] The first processing module 43 is used to input the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large network traffic demand of the application scenario, until the network experience detection model outputs a first detection result, or the network experience detection model is the Nth network experience detection model. The first detection result is used to characterize the target terminal's use of the target network in the application scenario; the network experience detection model of the N application scenarios is trained based on the network experience detection model training method described in any of the aforementioned embodiments.

[0178] The second processing module 44 is configured to determine an experience score of the target terminal using the target network according to the application scenario represented by the first detection result, wherein the experience score is used to represent the network quality of the target network.

[0179] Optionally, the second processing module 44 is specifically configured to determine the experience level of the target terminal using the target network according to the application scenario represented by the first detection result; and input the experience level of the target terminal using the target network into a preset experience score calculation formula to obtain the experience score of the target terminal using the target network. The experience level is positively correlated with the network traffic demand of the application scenario.

[0180] Optionally, the second processing module 44 is also used to obtain and output at least one of the experience score of the wireless access point providing the target network, the experience score of the area where the wireless access point is located, and the experience score of the entire network to which the area belongs, after obtaining the experience score of the target terminal using the target network.

[0181] Optionally, the second processing module 44 is specifically configured to obtain the number of target terminals connected to the wireless access point, the number of low-experience-level target terminals among the target terminals connected to the wireless access point, and the weight of the number of low-experience-level target terminals; and obtain the experience score of the wireless access point according to the number of target terminals connected to the wireless access point, the number of low-experience-level target terminals, the weight, and the experience score of the target terminal using the target network. The low-experience-level target terminal is a target terminal whose experience level using the target network is less than or equal to a preset experience level.

[0182] Optionally, the second processing module 44 is specifically configured to obtain the number of wireless access points included in the area where the wireless access point is located; and obtain the experience score of the area where the wireless access point is located according to the number of wireless access points and the experience score of each of the wireless access points.

[0183] Optionally, the second processing module 44 is specifically used to determine, for any area in the entire network, the experience score coefficient of the area based on the ratio of the number of target terminals with low experience levels connected to all wireless access points included in the area to the number of all target terminals connected to all wireless access points; and obtain the experience score of the entire network to which the area belongs based on the experience score of the area and the experience score coefficient of the area.

[0184] Optionally, the first processing module 43 is also used to determine whether the uplink and downlink traffic of the target terminal is greater than or equal to a preset uplink and downlink traffic threshold before inputting the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large according to the network traffic demand of the application scenario; when the uplink and downlink traffic of the target terminal is greater than or equal to the preset uplink and downlink traffic threshold, input the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large according to the network traffic demand of the application scenario; when the uplink and downlink traffic of the target terminal is less than the preset uplink and downlink traffic threshold, determine that the experience level of the target terminal using the target network is the target second experience level. The target second experience level is lower than the detection result output by any of the network experience detection models, which is used to characterize the experience level of the target terminal using the target network when the target terminal freezes when using the target network in the application scenario.

[0185] The network quality detection device provided in the present application is used to execute the aforementioned network quality detection method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0186] Fig. 9 This is a schematic diagram of the structure of an electronic device provided in this application. Fig. 9 As shown, the electronic device 500 may include: at least one processor 501 and a memory 502 .

[0187] The memory 502 is used to store programs. Specifically, the programs may include program codes, and the program codes include computer operation instructions.

[0188] The memory 502 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory.

[0189] The processor 501 is used to execute the computer-executable instructions stored in the memory 502 to implement the network experience detection model training and / or network quality detection method described in the aforementioned method embodiment. The processor 501 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0190] Optionally, the electronic device 500 may further include a communication interface 503. In a specific implementation, if the communication interface 503, the memory 502 and the processor 501 are implemented independently, the communication interface 503, the memory 502 and the processor 501 may be interconnected through a bus and communicate with each other. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc., but it does not mean that there is only one bus or one type of bus.

[0191] Optionally, in a specific implementation, if the communication interface 503, the memory 502 and the processor 501 are integrated on a chip, the communication interface 503, the memory 502 and the processor 501 can communicate through an internal interface.

[0192] The present application also provides a computer-readable storage medium, which may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, and other media that can store program codes. Specifically, the computer-readable storage medium stores program instructions, and the program instructions are used for the methods in the above embodiments.

[0193] The present application also provides a program product, which includes an execution instruction, which is stored in a readable storage medium. At least one processor of the electronic device can read the execution instruction from the readable storage medium, and at least one processor executes the execution instruction so that the electronic device implements the network experience detection model training and / or network quality detection method provided by the above various embodiments.

[0194] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application 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 replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A network quality detection method, characterized in that: The method comprises: collecting at least one network quality indicator when the target terminal uses the target network; Acquire a network quality indicator sequence according to the at least one network quality indicator; Inputting the network quality indicator sequence into the network experience detection models of N application scenarios in order from small to large network traffic requirements of the application scenarios, until the network experience detection model outputs a first detection result, or the network experience detection model is the Nth network experience detection model; the first detection result is used to characterize that the target terminal is stuck when using the target network in the application scenario; An experience score of the target terminal using the target network is determined according to the application scenario represented by the first detection result; the experience score is used to represent the network quality of the target network.

2. The method according to claim 1, characterized in that The determining, according to the application scenario represented by the first detection result, an experience score of the target terminal using the target network, includes: Determining, according to the application scenario represented by the first detection result, an experience level of the target terminal using the target network; the experience level is positively correlated with the network traffic demand of the application scenario; The experience level of the target terminal using the target network is input into a preset experience score calculation formula to obtain an experience score of the target terminal using the target network.

3. The method according to claim 2, characterized in that After obtaining the experience score of the target terminal using the target network, the method further includes: According to the experience score of the target terminal using the target network, at least one of the experience score of the wireless access point providing the target network, the experience score of the area where the wireless access point is located, and the experience score of the entire network to which the area belongs is obtained and output.

4. The method according to claim 3, characterized in that The obtaining, according to the experience score of the target terminal using the target network, an experience score of a wireless access point providing the target network, comprises: Acquire the number of target terminals connected to the wireless access point, the number of low experience level target terminals among the target terminals connected to the wireless access point, and the weight of the number of low experience level target terminals; the low experience level target terminals are target terminals whose experience level of using the target network is less than or equal to a preset experience level; The experience score of the wireless access point is obtained according to the number of target terminals connected to the wireless access point, the number of target terminals with low experience levels, the weights, and the experience scores of the target terminals using the target network.

5. The method according to claim 4, characterized in that The obtaining, according to the experience score of the target terminal using the target network, the experience score of the area where the wireless access point is located, includes: Acquire the number of wireless access points included in the area where the wireless access point is located; The experience score of the area where the wireless access point is located is obtained according to the number of the wireless access points and the experience score of each of the wireless access points.

6. The method according to claim 5, characterized in that The obtaining, according to the experience score of the target terminal using the target network, an experience score of the entire network to which the area belongs, includes: For any area in the entire network, determine the experience score coefficient of the area according to the ratio of the number of target terminals with low experience levels connected to all wireless access points included in the area to the number of all target terminals connected to all wireless access points; The experience score of the entire network to which the area belongs is obtained according to the experience score of the area and the experience score coefficient of the area.

7. The method according to any one of claims 2 to 6, characterized in that: Before inputting the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large network traffic demand of the application scenarios, the method further includes: Determine whether the uplink and downlink traffic of the target terminal is greater than or equal to a preset uplink and downlink traffic threshold; If yes, the network quality indicator sequence is input into the network experience detection model of N application scenarios in order from small to large network traffic demand of the application scenarios; If not, it is determined that the experience level of the target terminal using the target network is the target second experience level; the target second experience level is lower than the detection result output by any of the network experience detection models, which is used to characterize the experience level of the target terminal using the target network when the target terminal encounters a freeze when using the target network in this application scenario.

8. A network quality detection device, characterized in that: The device comprises: A collection module, used to collect at least one network quality indicator when the target terminal uses the target network; An acquisition module, configured to acquire a network quality indicator sequence according to the at least one network quality indicator; A first processing module is used to input the network quality indicator sequence into the network experience detection model of N application scenarios in order from small to large according to the network traffic demand of the application scenario, until the network experience detection model outputs a first detection result, or the network experience detection model is the Nth network experience detection model; the first detection result is used to characterize that the target terminal is stuck when using the target network in the application scenario; The second processing module is used to determine an experience score of the target terminal using the target network according to the application scenario represented by the first detection result; the experience score is used to represent the network quality of the target network.

9. An electronic device, characterized in that: The electronic device comprises a memory and a processor; The memory stores a computer program; The processor is configured to execute the method of any one of claims 1 to 7 through the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images