WO2024045761A1 - 通信预测方法、服务器、预测装置、终端设备和存储介质 - Google Patents

通信预测方法、服务器、预测装置、终端设备和存储介质 Download PDF

Info

Publication number
WO2024045761A1
WO2024045761A1 PCT/CN2023/099477 CN2023099477W WO2024045761A1 WO 2024045761 A1 WO2024045761 A1 WO 2024045761A1 CN 2023099477 W CN2023099477 W CN 2023099477W WO 2024045761 A1 WO2024045761 A1 WO 2024045761A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
communication quality
target
application
preset
Prior art date
Application number
PCT/CN2023/099477
Other languages
English (en)
French (fr)
Inventor
吴咸樾
薛学儒
王健伟
索传奇
石仟华
王毛
刘得煌
何坚
Original Assignee
Oppo广东移动通信有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Oppo广东移动通信有限公司 filed Critical Oppo广东移动通信有限公司
Publication of WO2024045761A1 publication Critical patent/WO2024045761A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/52—Network services specially adapted for the location of the user terminal
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/06—Testing, supervising or monitoring using simulated traffic
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/08—Testing, supervising or monitoring using real traffic
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the present application relates to the field of communication technology, and in particular, to a communication prediction method, server, prediction device, terminal equipment and storage medium.
  • the terminal devices can obtain various functions from clients with various functions through the network. information, such as watching videos, listening to music, reading files, etc.
  • the terminal device needs to be connected to the network to use the client. Due to the small coverage of the base station, or due to reasons such as being indoors or traveling by subway, the network signal may be weak or no network signal. Therefore, when the terminal device enters an area with weak or no network signal, the network is basically unavailable, and the client cannot be used normally.
  • the terminal device by pre-collecting and analyzing the changes in network signal quality (such as air interface transmission delay, real-time air interface bandwidth, real-time air interface rate and packet loss rate, etc.) when the terminal device moves on a fixed travel route, it can learn that the terminal device is about to enter
  • the time of the weak signal area and the time of leaving the weak signal area are the process of constructing a communication quality map for the terminal device.
  • Embodiments of the present application provide a communication prediction method, server, prediction device, terminal equipment and storage medium, which improves the versatility of the communication quality map.
  • embodiments of the present application provide a communication prediction method.
  • the method includes: obtaining multiple communication quality data in multiple attribute dimensions; each of the communication quality data corresponds to attributes in the multiple attribute dimensions.
  • Information the plurality of attribute dimensions represent dimensions of different scenarios or applications; determine target communication quality data and target time information corresponding to the target communication quality data according to the plurality of communication quality data and preset parameter threshold values;
  • a communication quality map is determined; wherein the communication quality map is used by the terminal device to communicate with different applications in different scenarios. Quality prediction.
  • embodiments of the present application provide another communication prediction method.
  • the method includes: obtaining the current location and a communication quality map sent by the server; the communication quality map is used to provide information when the communication quality is lower than the preset communication quality. , attribute information and time information corresponding to different applications in different scenarios; determine a communication prediction strategy based on the communication quality map and the current location; send the communication prediction strategy to the application module in the terminal device for the The application module performs communication prediction on the running application according to the communication prediction strategy.
  • inventions of the present application provide a server.
  • the server includes: a first acquisition part configured to acquire multiple communication quality data in multiple attribute dimensions; each of the communication quality data corresponds to the multiple communication quality data. Attribute information under multiple attribute dimensions; the multiple attribute dimensions represent dimensions of different scenarios or applications; the first determination part is configured to determine the target communication quality based on a plurality of the communication quality data and preset parameter threshold values The target time information corresponding to the data and the target communication quality data; determine the communication quality map according to the target attribute information of the target communication quality data in the multiple attribute dimensions and the target time information; wherein, the communication quality The map is configured for terminal devices to predict communication quality for different applications in different scenarios.
  • inventions of the present application provide a prediction device.
  • the prediction device includes: a second acquisition part configured to acquire the current location and a communication quality map sent by the server; the communication quality map is configured to provide When the communication quality is lower than the preset communication quality, the corresponding attribute information and time information of different applications in different scenarios; the second determination part is configured to determine the communication quality according to the communication quality. map and the current location to determine the communication prediction strategy; the sending part is configured to send the communication prediction strategy to the application module in the terminal device, so that the application module can perform communication prediction on the running application according to the communication prediction strategy.
  • embodiments of the present application provide a terminal device, which includes: a prediction device and an application module as described in the fourth aspect; the application module is configured to predict running data according to the communication prediction policy. Applications for communication prediction.
  • embodiments of the present application provide a server, the server including: a first memory for storing an executable computer program; a first processor for executing the executable computer program stored in the first memory When , the communication prediction method of the above first aspect is implemented.
  • inventions of the present application provide a terminal device.
  • the terminal device includes: a second memory for storing an executable computer program; a second processor for executing the executable computer program stored in the second memory.
  • the communication prediction method of the second aspect is implemented.
  • embodiments of the present application provide a computer-readable storage medium storing a computer program for implementing the communication prediction method of the first aspect or the second aspect when executed by a processor.
  • Embodiments of the present application provide a communication prediction method, server, prediction device, terminal equipment and storage medium.
  • multiple communication quality data in multiple attribute dimensions are obtained; each communication quality data corresponds to attribute information in multiple attribute dimensions; the multiple attribute dimensions represent the dimensions of different scenarios or applications; according to Multiple communication quality data and preset parameter threshold values are used to determine target communication quality data and target time information corresponding to the target communication quality data.
  • the area corresponding to the target communication quality data represents the weak signal area.
  • the weak signal areas that need attention are filtered out according to the preset parameter threshold values.
  • the relevant characteristics carried by the weak signal area (for example, application, spatial location, weak signal time information, etc.) , can adapt to different applications and different scenarios.
  • the communication quality map is determined; among them, the communication quality map is used by the terminal device to predict communication quality for different applications in different scenarios.
  • Figure 1 is an optional step flow chart of a communication prediction method provided by an embodiment of the present application
  • Figure 2 is an exemplary flow chart of a communication prediction method based on QoS map construction provided by an embodiment of the present application
  • Figure 3 is an optional step flow chart of another communication prediction method provided by an embodiment of the present application.
  • Figure 4 is an exemplary flow chart of a low QoS information extraction algorithm provided by an embodiment of the present application.
  • Figure 5 is an exemplary schematic diagram of a low QoS area provided by an embodiment of the present application.
  • Figure 6 is an exemplary diagram of an exemplary QoS map update based on feedback information from a third-party application provided by an embodiment of the present application
  • Figure 7 is an optional step flow chart of yet another communication prediction method provided by an embodiment of the present application.
  • Figure 8 is an optional step flow chart of yet another communication prediction method provided by an embodiment of the present application.
  • FIG. 9 is an exemplary schematic diagram of the stuck feature of an application provided by the embodiment of the present application.
  • Figure 10 is an exemplary flow chart of another communication prediction method based on QoS map construction provided by the embodiment of the present application.
  • Figure 11 is an exemplary schematic diagram of the interaction between a Qos prediction service provided by an embodiment of the present application and a third-party application;
  • Figure 12 is an exemplary schematic diagram of another Qos prediction service provided by an embodiment of the present application interacting with a third-party application;
  • Figure 13 is an optional structural schematic diagram of a server provided by an embodiment of the present application.
  • Figure 14 is an optional structural schematic diagram of a prediction device provided by an embodiment of the present application.
  • Figure 15 is an optional structural schematic diagram of a terminal device provided by an embodiment of the present application.
  • Figure 16 is a schematic structural diagram of a server provided by an embodiment of the present application.
  • Figure 17 is a schematic structural diagram of a terminal device provided by an embodiment of the present application.
  • the network signal quality considered is measured based on air interface-based signal quality, which fails to reflect the impact of communication service quality caused by human flow factors.
  • areas with poor communication service quality usually occur during peak hours of human flow (for example, commuting to and from get off work). In this scenario, weak signal areas are easily missed.
  • the communication quality (Quality of Service, QoS) map constructed in this solution has low attribute dimensions. For example, it does not take into account scenarios, applications and other dimensions, and the prediction effect is limited. It does not take into account the suitability of the QoS map for different scenarios and different applications. Matching nature.
  • QoS refers to the client's description or measurement of the overall performance of communication services.
  • several related indicators are usually considered, such as packet loss, bit rate, throughput, transmission delay, availability, jitter, etc.
  • the application data preloading method provided in the related art is, for example, the mobile terminal obtains the positioning position in real time through the Global Positioning System (GPS), and determines whether it is about to arrive based on the current location and the weak signal map constructed based on the GPS. Entering a weak signal area. If it is determined that a weak signal area is about to be entered, the time required to pass through the weak signal area is calculated based on the weak signal map, and the application data is downloaded and cached in advance based on this time period to ensure the smoothness of the application in the weak signal area.
  • GPS Global Positioning System
  • the above method of preloading application data is to obtain the real-time location through GPS and build a weak signal map based on GPS. Since the GPS signal can only be received outdoors, it is not suitable for indoor, subway and underground sections, which reduces the risk of weak signals. Suitability of signal maps.
  • FIG. 1 is a step flow chart of a communication prediction method provided by an embodiment of the present application.
  • the communication prediction method can be executed by a server.
  • the communication prediction method includes the following step:
  • the communication quality data of the terminal device in the scenario is collected.
  • the communication quality data includes multiple attribute dimensions corresponding to the scenario and the network quality of the terminal device in the scenario; multiple attributes corresponding to the scenario are
  • An attribute dimension can include spatial location, collection time period, currently running application, communication signal type of the terminal device, etc.
  • the attribute information is the detailed content of the attribute dimension; the network quality represents the quality of the connection signal of the terminal device in this scenario, which can express The smoothness and loading speed of the currently running application.
  • the attribute information is detailed content under attribute dimensions, for example, subway station A, application B, collection time period It is 17:00-18:00, operator C and 5G signal, and also covers the network quality score under the multiple attribute information (which can be expressed by communication rate, communication delay, throughput rate, etc.).
  • the communication quality data can be expressed in numerical form. If application B plays smoothly (the fluency can be determined by parameters such as communication rate, communication delay, and throughput rate), the corresponding numerical value of the communication quality data will be higher.
  • the combination of multiple attribute dimensions can constitute different scenarios.
  • the multiple attribute dimensions include at least two of the following: spatial location, time period, application type, operator, and communication standard.
  • the spatial location represents the area corresponding to the communication quality data, for example, a certain stop sign, a certain subway station, a certain office building, a certain shopping mall, etc.
  • the time period represents the time corresponding to the communication quality data, Including but not limited to date and a certain time period of a certain day
  • the application type indicates the application corresponding to the communication quality data
  • the communication format indicates the communication format of the terminal device when accessing the Internet.
  • S101 in Figure 1 above can be implemented in the following manner. Obtain multiple initial communication quality data in multiple attribute dimensions; based on the multiple initial communication quality data, determine multiple communication quality data corresponding to each initial communication quality data.
  • the data format of the initial communication quality data includes multiple dimensions such as time, space (location), application, user, operator, and communication standard.
  • Communication quality data includes different representations in the same dimension.
  • the spatial location dimension it can include different road sections, different subway stations, different indoors, etc.
  • the time dimension it can include different times of the day.
  • the application dimension it can include different application software for playing long videos, different application software for playing short videos, different application software for playing audio, different application software for information push, Different social software, etc.
  • the operator dimension it can include different operators
  • the communication standard dimension it represents the communication method, which can include the 4th generation mobile communication technology (4G), 5G, Wireless fidelity technology (Wi-Fi), etc.
  • the initial communication quality data also includes corresponding network quality parameters in different manifestations.
  • the initial communication quality data is scored according to the network quality parameters using a quality scoring method. , to obtain communication quality data.
  • Network quality parameters include but are not limited to communication rate, communication delay, throughput rate, other indicators that characterize communication quality, and the QoS metric formed by correlating these indicators.
  • the initial communication quality data can be processed based on the QoS metric. Score, and use the value calculated based on QoS metric as communication quality data.
  • the quality scoring method may be a weighted summation of communication rate, communication delay, throughput rate and other indicators, or other calculation methods, which are not limited by the embodiments of this application.
  • the initial communication quality data is QoS data and the communication quality map is a QoS map.
  • the QoS data The data collection method can be manual collection or crowdsourcing collection.
  • Manual collection refers to the collection of QoS data by testers at a designated time and space.
  • Crowdsourcing collection includes explicit and implicit. Among them, explicit refers to issuing crowdsourcing collection tasks to ordinary users, collecting tasks through users, and deploying the collection program to the user's terminal device, allowing users to complete the specified collection tasks by themselves.
  • the user is aware of the collection of QoS data; implicit refers to the collection without the user's active participation, the user is unaware of the collection of QoS data, after connecting to Wi-Fi, the terminal can actively report to the server, automatically Synchronize QoS data. Whether it is manual collection or crowdsourcing collection, upload the collected QoS data to the server so that the server can construct a QoS map based on the QoS data instead of detecting weak signal areas on the terminal device side to avoid excessive consumption of terminal device resources.
  • the plurality of communication quality data include a large amount of data reflecting network quality, including data with good network quality and data with poor network quality.
  • data with poor network quality is taken as the data of interest, and preset parameter thresholds are used to filter it out, thereby obtaining data with poor network quality (corresponding to the target communication quality data).
  • the target communication quality data includes a plurality of communication quality data that satisfy the conditions.
  • the area corresponding to the communication quality data can be determined.
  • the area corresponding to the filtered target communication quality data can also be determined. Since the target communication quality data is data of poor network quality, the area corresponding to the target communication quality data can represent a weak signal area or a low QoS area, and the weak signal areas that need attention are filtered out based on the preset parameter thresholds. Since the communication quality data carries time information (corresponding to attribute information), the target time information of the weak signal area can also be determined based on the filtered time information of the target communication quality data.
  • the target time information includes but is not limited to the duration of the weak signal. Duration and weak signal start time, etc.
  • S103 Determine a communication quality map based on the target attribute information in multiple attribute dimensions of the target communication quality data and the target time information; wherein, the communication quality map is used by the terminal device to predict communication quality for different applications in different scenarios.
  • the target communication quality data includes multiple communication quality data that meet the conditions. For one communication quality data that meets the conditions, it is marked on the map based on the attribute information corresponding to the communication quality data.
  • the map contains not only regional information with poor network quality, time period information, application information, operator information and communication standard information. Multiple such communication quality data are marked on the map to form a communication quality map.
  • the area corresponding to the target communication quality data is a weak signal area.
  • the multiple attribute dimensions of the target communication quality data include but are not limited to spatial location, time period, application type, operator and communication standard, combined with the target time. information to build a communication quality map.
  • the constructed communication quality map is a collection of multiple weak signal areas, as well as the attribute dimensions of each weak signal area.
  • the relevant characteristics carried by the weak signal area can be adapted to different applications and different scenarios.
  • the communication quality map includes weak signal areas with different attribute dimensions and is a collection of weak signal areas with multiple attribute dimensions, for example, a weak signal area for a certain application, a weak signal for a certain time period area, a weak signal area related to a certain spatial location, a weak signal area related to a certain communication standard, a weak signal area related to a certain operator, etc.
  • the terminal device can use the communication quality map to predict communication quality for different applications and different scenarios. There is no need to construct communication quality maps for different applications and different scenarios, which improves the versatility of the communication quality map.
  • the communication quality map in the embodiment of this application can be called a QoS map, a weak signal map, a communication semantic map, a communication technology availability map (for example, a 4G availability map, a 5G availability map), a quality of user experience (Quality of Experience, QoE) map, service availability map (for example, Virtual Reality (VR) service availability map, V2X (vehicle to everything) service availability map), peak/off-peak map, Global Positioning System (Global Positioning System, GPS) ) signal strength map, etc.
  • QoS map a weak signal map
  • a communication semantic map for example, a 4G availability map, a 5G availability map
  • QoE quality of user experience
  • service availability map for example, Virtual Reality (VR) service availability map, V2X (vehicle to everything) service availability map
  • peak/off-peak map for example, Global Positioning System (Global Positioning System, GPS)
  • each communication quality data corresponds to attribute information in multiple attribute dimensions; the multiple attribute dimensions represent different scenarios or Dimensions of application; determine target communication quality data and target time information corresponding to the target communication quality data based on multiple communication quality data and preset parameter thresholds.
  • the area corresponding to the target communication quality data represents the weak signal area.
  • the weak signal areas that need attention are filtered out according to the preset parameter threshold values.
  • the relevant characteristics carried by the weak signal area (for example, application, spatial location, weak signal time information, etc.) , can adapt to different applications and different scenarios.
  • the communication quality map is determined; among them, the communication quality map is used by the terminal device to predict communication quality for different applications in different scenarios.
  • the Internet access service can perform QoS prediction according to the constructed QoS map.
  • the initial communication quality data is QoS data
  • the communication quality map is a QoS map
  • the server is a cloud server.
  • Figure 2 is a communication prediction based on the QoS map provided by the embodiment of the present application.
  • the communication prediction method includes S11-S14.
  • the definition of QoS includes but is not limited to communication rate, communication delay, throughput rate, other indicators that characterize communication quality, and the QoS metric formed by correlating these indicators.
  • the QoS metric Score the QoS data and use the value calculated based on the QoS metric as the scoring data of the QoS data.
  • the QoS data collection frequency needs to meet certain requirements. For example, corresponding QoS data needs to be available every second, so that the constructed communication quality map can be adapted to different applications and different scenarios.
  • the QoS map download service can be deployed on a cloud server (which can also be called a cloud), and the terminal device downloads the QoS map by calling the open application programming interface (Application Programming Interface, API) of the cloud.
  • the download of the QoS map can be a full download for a city, or an incremental download for a certain spatial location, a certain time, a certain operator and other variables.
  • communication fluency can be improved by using QoS maps.
  • the QoS prediction method can use the QoS map to predict the QoS quality in the future time or future location based on the current time, location, applications used and other information, and notify the third-party applications of the prediction results to help them take countermeasures in advance.
  • S102 in Figure 1 above may also include S1021 and S1022.
  • Figure 3 is an optional step flow chart of another communication prediction method provided by an embodiment of the present application.
  • the communication quality data includes different representation forms (that is, different attribute information) in the same dimension. Different representation forms can be represented by different fields. When grouping, the corresponding data can be found according to the keywords. Data, thereby grouping communication quality data with consistent attribute information into the same group for statistical calculation.
  • the attribute information includes the collection time period. A piece of communication quality data is about the data at each time point in a certain time period, which means that the communication quality data is continuous in time. There is a plurality of attribute information, and the attribute information of the plurality of communication quality data in a set of communication quality data all correspond to the same one-to-one. The attribute information between each set of communication quality data is completely different or partially different.
  • a set of communication quality data includes multiple communication quality data.
  • the multiple communication quality data are collected at Station A, Application B, the collection time period is 8:00-9:00, Operator C, Fifth
  • the communication quality data under these conditions are the 5G New Radio (5G NR) signals of mobile communication, which means that the attribute information of the multiple communication quality data is consistent.
  • 5G NR 5G New Radio
  • the time period corresponds to 8:00-9:00. In actual situations, taking seconds as the interval as an example, there is corresponding data for every second, that is to say, This communication quality data is temporally continuous.
  • each group of communication quality data extraction is performed based on the preset parameter threshold value, and the data with poor network quality (corresponding to the target communication quality data) of each group of communication quality data is determined.
  • the spatial position in the attribute dimension of the quality data determines the weak signal area, thereby filtering out the weak signal area that needs attention.
  • the target time information corresponding to the target communication quality data the weak signal duration, weak signal start time, etc. of the weak signal area are determined.
  • the preset parameter threshold value includes: a preset threshold value; S1022 in the above-mentioned Figure 3 may include S201-S203.
  • S203 Determine target time information based on multiple time points corresponding to each first data.
  • the preset threshold value is a threshold value used to determine whether the communication quality data is low QoS data, that is, the threshold value below which the communication quality data is considered to be low QoS data.
  • Each first data corresponds to multiple consecutive time points. Multiple consecutive time points can constitute a certain time period. According to Multiple consecutive time periods can be used to determine the weak signal duration and weak signal start time.
  • the attribute information of the spatial location dimension corresponding to the multiple first data is consistent.
  • the weak signal area can be determined based on the attribute information of the spatial location dimension of the multiple first data, thereby obtaining the weak signal persistence of the weak signal area. time and weak signal start time.
  • the plurality of first data when the plurality of first data is greater than or equal to the preset number of samples, it indicates that the number of low QoS data is sufficient.
  • the plurality of first data are determined as data with poor network quality (corresponding to target communication quality data).
  • the plurality of first data is less than the preset number of samples, it means that the number of low QoS data is a little small and is not enough to be used as data of poor network quality.
  • the preset threshold value can be appropriately set by those skilled in the art according to the actual situation and the current scenario. It can be determined during the experiment of a large number of extraction processes. Different preset thresholds can be set for different scenarios. limit. The embodiments of this application are not limited to this.
  • the preset parameter threshold also includes: a preset ratio and a preset time period; after the above S201, the communication prediction method may also include S204-S208.
  • the low QoS data and other QoS data correspond to multiple consecutive time points. There will be a certain time point, although there is low QoS data. Data, but its proportion is small. In this case, it cannot be considered as data with poor network quality (corresponding to the target communication quality data).
  • For each time point calculate the data proportion between low QoS data and the total amount of data at that time point. When the data proportion is greater than or equal to the preset ratio, indicate the proportion of low QoS data at that time point. If it is larger, the next step of calculation can be carried out, which improves the accuracy of data statistics.
  • the target signal duration and target signal start time corresponding to the multiple third data Based on multiple time points corresponding to each third data, determine the target signal duration and target signal start time corresponding to the multiple third data; the target time information includes the target signal duration and the target signal start time.
  • the second data is part of the first data. Therefore, multiple continuous time points corresponding to each second data can be obtained based on the multiple continuous time points corresponding to each first data.
  • the third data is part of the second data, and multiple consecutive time points corresponding to each third data can be obtained.
  • a candidate duration corresponding to the second data is determined based on multiple consecutive time points corresponding to the second data.
  • the weak signal duration indicating the candidate duration may affect the normal use of the currently running application and is considered to be data of poor network quality (corresponding to the target communication quality data).
  • a plurality of third data are extracted from a plurality of second data according to the relationship between the candidate duration and the preset time period. Multiple third data are used as target communication quality data, and then based on multiple time points corresponding to each third data, the duration of the weak signal and the start time of the weak signal are determined.
  • the preset ratio and the preset time period can be appropriately set by those skilled in the art according to the actual situation and the current scenario. They can be determined during the experiment of a large number of extraction processes. Different settings can be set for different scenarios. Preset ratio and preset time period. The embodiments of this application are not limited to this.
  • the attribute information of the spatial location dimension corresponding to the multiple third data is consistent, and the weak signal area can be determined based on the attribute information of the spatial location dimension of the multiple third data, thereby obtaining the weak signal duration and Weak signal start time.
  • the effectiveness and accuracy of data statistics are improved.
  • the preset parameter threshold also includes: a preset deviation; after the above S206, the communication prediction method may also include S209-S212.
  • S211 Use multiple target data as target communication quality data, and obtain multiple time points corresponding to each target data.
  • the target signal duration and target signal start time corresponding to the multiple target data according to the multiple time points corresponding to each target data; the target time information includes the target signal duration and the target signal start time.
  • the plurality of third data are determined by calculating the target deviation between the plurality of third data. Volatility between tertiary data. The smaller the fluctuation between data, the more reference the data has.
  • multiple target data whose target deviation is less than or equal to the preset deviation are extracted. The fluctuation between the multiple target data is small, indicating that the performance of the multiple target data is stable and has reference value. Multiple target data are considered to be data with poor network quality (corresponding to target communication quality data).
  • the target data is part of the third data. Therefore, multiple continuous time points corresponding to each target data can be obtained based on the multiple continuous time points corresponding to each third data. Based on multiple time points corresponding to each target data, the duration of the weak signal and the start time of the weak signal are determined.
  • the attribute information of the spatial location dimension corresponding to multiple target data is consistent.
  • the weak signal area can be determined based on the attribute information of the spatial location dimension of the multiple target data, thereby obtaining the weak signal duration and weak signal of the weak signal area. Starting time. By setting preset deviations, the stability of data statistics is improved.
  • the target deviation includes: the first date deviation and the first data deviation; the preset deviation includes: the preset date fluctuation difference and the preset data fluctuation difference; the above S210 can also be implemented through the following two examples.
  • Example 1 Group multiple third data by date and determine the first date deviation between multiple third data on different dates; among multiple third data, determine that the first date deviation is less than or equal to the preset A plurality of fourth data of date fluctuation differences; determining a first data deviation of each fourth data; and determining a plurality of target data whose first data deviation is less than or equal to a preset data fluctuation difference among the plurality of fourth data.
  • the attribute information of the time dimension of the plurality of third data includes different time periods on different dates.
  • the third data is communication quality data for the time period of 8:00-9:00 every day.
  • the attribute information of multiple third data is consistent, their corresponding dates may be different. Therefore, multiple third data are grouped according to date, and the date deviation between different dates is calculated (corresponding to the first date deviation) . If the date deviations on different dates are too large, it means that these data are highly volatile and unstable on different dates. It is necessary to extract a plurality of fourth data whose date fluctuations are relatively stable (less than or equal to a preset date fluctuation difference) from a plurality of third data.
  • the data deviation (corresponding to the first data deviation) between the plurality of fourth data is calculated, and a plurality of target data whose data fluctuation is relatively stable (less than or equal to the preset data fluctuation difference) are extracted from the plurality of fourth data.
  • the target deviation includes a second date deviation and a second data deviation
  • the preset deviation includes: a preset date fluctuation difference and a preset data fluctuation difference
  • the above S210 can also be implemented through Example 2. Determine the second data deviation of the plurality of third data; determine the plurality of fifth data whose second data deviation is less than or equal to the preset data fluctuation difference among the plurality of third data; group the plurality of fifth data according to date , determine a second date deviation between a plurality of fifth data on different dates; determine a plurality of target data whose second date deviation is less than or equal to a preset date fluctuation difference among the plurality of fifth data.
  • the data deviation (corresponding to the second data deviation) between multiple third data is calculated.
  • the fluctuation of the data extracted from the multiple third data is relatively stable ( A plurality of fifth data less than or equal to the preset data fluctuation difference).
  • the attribute information of the time dimension of multiple fifth data includes different time periods on different dates.
  • the attribute information of multiple fifth data is consistent, but their corresponding dates may be different. Therefore, multiple fifth data are grouped according to date. , calculates the date deviation between different dates (corresponding to the second date deviation).
  • a plurality of target data whose date fluctuations are relatively stable (less than or equal to a preset date fluctuation difference) are extracted from the plurality of fifth data.
  • the communication quality map is a QoS map
  • the communication quality data is QoS data
  • the preset threshold is the low QoS threshold
  • the preset number of samples is N
  • the preset ratio is the low QoS scoring proportion.
  • Each low QoS area not only contains attribute information such as spatial location, time period, application, operator, communication standard, but also includes the start time and end time of low QoS information, as well as the duration of low QoS information.
  • Figure 4 is an exemplary flow chart of a low QoS information extraction algorithm provided by an embodiment of the present application.
  • Figure 4 is a low QoS information extraction algorithm for QoS packet data, which can include the following steps:
  • the low QoS scores in each group are grouped according to date, and the date deviation of the low QoS data on different dates is calculated.
  • the low QoS threshold indicates the threshold below which the QoS data is considered as low QoS data
  • the low QoS proportion threshold The road section is added to the QoS map
  • the duration T of the low QoS score indicates the time when the QoS score continues to be lower than the low QoS threshold.
  • the above scenes may include spatial locations.
  • the above parameter settings may be different for the subway environment and the home environment.
  • the above parameter settings for the study room and the kitchen in the home environment may be different. May be different.
  • the scenario can also include different applications.
  • the above parameter settings corresponding to games, long video APPs, and short video APPs may be different.
  • the extraction algorithm provided in Figure 4 above has high flexibility and improves the accuracy of QoS map construction, thereby improving the performance of QoS prediction using QoS maps.
  • Figure 5 is this The application embodiment provides an exemplary schematic diagram of a low QoS area.
  • Figure 5 shows the QoS scoring data collected in a subway scenario when the subway is running at a certain station.
  • the collection start time is 18:00, and the collection time period is approximately 260 seconds.
  • the ordinate in Figure 5 represents the QoS scoring data. The larger the score, the better the network quality. When building a QoS map, you need to pay attention to the low QoS scoring data.
  • Figure 5 takes the preset threshold value of 60 as an example.
  • QoS score data below 60 are regarded as low QoS data.
  • the abscissa in Figure 5 represents time.
  • the data collected starting from 18:00 are aligned according to time.
  • the multiple data collected are arranged in Figure 5 according to the time series distribution.
  • the curve represents the average of multiple data, and each second corresponds to multiple QoS scoring data.
  • the standard deviation of low QoS data fluctuations per second is the width of the upper and lower boundaries in Figure 5. If the width is larger, it means that the data is more volatile. If the width is smaller, it means that the data is more volatile. The accuracy is small, the data is stable and can be used as a reference. Since the QoS data shown in Figure 5 is about a certain subway station, the low QoS area can be determined. After the above data extraction process in Figure 4, the duration of the low QoS area and the start time of the low QoS area can be obtained. , represented by marked position 2 and marked position 3 in Figure 5.
  • the attribute dimensions of QoS maps are low. For example, dimensions such as scenarios and applications are not considered, and the prediction effect is limited.
  • the embodiment of this application uses a statistical algorithm of data extraction to construct a QoS map.
  • the preset parameter thresholds used in this algorithm can be dynamically adjusted. For different applications, different preset parameter thresholds can be set as needed, which improves QoS. Accuracy and versatility of maps.
  • this QoS map can be used to determine whether to start QoS prediction, without the need to build a separate QoS map for each application and each scenario.
  • QoS maps adapted to different applications are constructed, and a universal QoS map is constructed based on limited testing to reduce the construction cost of QoS maps.
  • the communication prediction method may also include S104 and S105.
  • S105 Update the communication quality map according to the feedback information and generate an updated communication quality map; the updated communication quality map is used by the terminal device to predict the next communication quality for different applications in different scenarios.
  • the prediction device in the terminal device communicates with the server.
  • the prediction device downloads the communication quality map from the server and uses the communication quality map to obtain the information about the impending entry into the weak signal area and the weak signal at a specific location or a specific time period.
  • the prediction device or application module in the terminal device determines whether to preload based on the status information of the running application.
  • the prediction device in the terminal device sends the feedback information of the third-party application to the terminal device.
  • the feedback information reflects the ability of the running application to withstand weak signals, as well as feedback on the current network quality.
  • the server can update the communication quality map based on feedback information.
  • the updated communication quality map is more suitable for a certain scenario or a certain application, thereby improving the accuracy of communication quality prediction.
  • Example 1 Adjust the preset parameter threshold value based on the feedback information to obtain the adjusted parameter threshold value; re-determine the updated communication quality map based on the adjusted parameter threshold value, and complete the communication quality map renew.
  • Example 2 Based on the feedback information, the information related to the running application in the communication quality map is updated to obtain an updated communication quality map.
  • the server updates the QoS map based on the feedback information fed back by the three-party applications (sent to the server by the prediction device), as shown in Figure 6.
  • Figure 6 shows the implementation of this application.
  • the example provides an exemplary schematic diagram of QoS map update based on feedback information from third-party applications.
  • the method for updating the QoS map in this embodiment of the present application may include the following two methods. One is to adjust the parameters in the low QoS information extraction algorithm (which can include preset thresholds, preset ratios, and preset time periods) based on feedback information from third-party applications, and then re-execute the low QoS information extraction algorithm to generate a new version of QoS. map.
  • Another method is to directly modify the attributes of existing low QoS areas based on feedback information from third-party applications to update the QoS map; for example, for a certain application, at a specific time or location, the third-party application provides QoS prediction services Low QoS alarms are always sent without any optimization strategy, indicating that the low QoS area has no impact on this application. Therefore, the low QoS area can be considered to be used for this application. Removed from the application's QoS map.
  • FIG. 6 takes the data flow as an example for explanation.
  • the feedback information of the third-party application is sent to the server by the prediction device of the terminal device (that is, the QoS prediction service device). Then the server performs the low QoS information extraction algorithm by modifying the low QoS area or adjusting the parameters in the low QoS information extraction algorithm, thereby releasing the QoS map version.
  • the QoS map is deployed on the terminal side of the server, and the QoS prediction service device downloads it from the terminal side to predict the next communication quality.
  • FIG. 7 is an optional step flow chart of yet another communication prediction method provided by an embodiment of the present application.
  • the communication prediction The method can be executed by a prediction device in the terminal device, and the communication prediction method includes the following steps:
  • the communication quality map is used to provide attribute information and time information corresponding to different applications in different scenarios when the communication quality is lower than the preset communication quality.
  • the terminal device can obtain its current location in real time based on GPS positioning during movement.
  • the terminal device can download the communication quality map from the server.
  • the communication quality map may be a universal QoS map constructed by the above-mentioned server using an extraction algorithm based on multiple communication quality data.
  • the universal QoS map is used to enable the terminal device to be in a specific location (and a specific time period). Obtain the low QoS data of the road ahead, and then determine whether to perform QoS prediction based on the application's lag characteristics and the general QoS map.
  • the communication quality map can also be an ordinary map constructed based on other methods. By further calculating the information included in the ordinary map, the low QoS data in the ordinary map is constructed, as long as the terminal device can be located at a specific location (and a specific time period). ), just obtain the low QoS data of the road section ahead.
  • Example 1 Obtain the current location through positioning; send a full map download request to the server, and obtain the communication quality map sent by the server in response to the full map download request.
  • Example 2 Send an incremental map download request to the server to obtain the incremental communication quality map sent by the server in response to the incremental map download request; based on the incremental communication quality map, update the obtained current communication quality map to obtain the communication quality map.
  • the QoS map is deployed on the server, that is, the download service is deployed on the server.
  • the terminal device downloads the map by calling the API interface opened by the server.
  • the QoS map download can be a full download for a region or city, or an incremental download for variables such as location, time, and operator.
  • map downloading is performed based on different requirements of terminal devices, which improves the diversity of downloading methods, and reduces resource consumption for incremental downloading.
  • the prediction device of the terminal equipment determines the communication prediction strategy, and the communication prediction
  • the policy can be a warning in a weak signal area or a policy indicating that a running application needs to be preloaded.
  • the prediction device sends the communication prediction strategy to the application module in the terminal device, so that the application module performs communication prediction on the running application.
  • the application module determines whether to preload based on the ability of the running application to withstand weak signals; if the prediction strategy is a strategy that indicates preloading, the application module determines whether to preload the running application according to the strategy. Perform preloading.
  • the current location and the general communication quality map sent by the server are obtained; the communication quality map is used to provide different applications in different scenarios when the communication quality is lower than the preset communication quality.
  • the corresponding attribute information and time information below; according to the communication quality map and the current location, the weak signal area ahead, the duration of the weak signal, the start time of the weak signal, etc. can be obtained at a specific location or a specific time period, and then based on the above-mentioned information related to the weak signal information determines communication prediction strategies.
  • the communication prediction result is to perform preloading on the running application or not to perform preloading on the running application.
  • the communication prediction The results match the situation of the running application, avoiding more video data preloading caused by triggering the prediction process, and reducing extra traffic consumption, extra power consumption and memory usage.
  • S302 in the above-mentioned Figure 7 may include S3021-S3023.
  • Figure 8 is an optional step flow chart of yet another communication prediction method provided by an embodiment of the present application.
  • the communication prediction strategy determines the communication prediction strategy based on the predicted duration; where , the arrival time represents the starting point from the current position Time to reach starting position.
  • the target area is a weak signal area
  • the prediction area is a weak signal area that terminal devices around the current location are likely to enter soon.
  • the prediction area represents the weak signal area that the terminal device is about to enter. .
  • the communication quality map is used to allow the terminal device to obtain the weak signal area ahead when it is at a specific location (corresponding to the boundary location).
  • the starting position of the prediction area is the location where the weak signal begins to appear. If the current position exceeds the boundary position, it means that the terminal device is about to enter a weak signal area. Of course, you can also judge whether the terminal device is about to enter a weak signal area through time sequence. Based on the current position, current time, starting position and obtained movement speed prediction, the arrival time of the terminal device to the weak signal area is predicted. If the arrival time is after the starting time, it means that the terminal device is about to enter the weak signal area. The terminal device determines the communication prediction strategy based on the predicted duration.
  • the communication prediction method before the above-mentioned S3023, the communication prediction method further includes S401 and S402.
  • S402. Determine the trigger reference value of the running application according to the status information of the running application.
  • the communication prediction strategy is determined according to the predicted duration in the above S3023, which can also be implemented in the following manner. If the predicted duration is greater than or equal to the triggering reference value, an execution strategy is generated; wherein the communication prediction strategy includes an execution strategy, and the execution strategy is used to cause the application module to pre-cache the running application.
  • video playback smoothness can be maintained for a period of time even without using the QoS prediction service. For example: a video APP continues to preload video data for the next 30 seconds under normal network signal conditions. Therefore, when the duration of the weak signal area it enters is less than 30 seconds, even if the QoS prediction service is not used, the smoothness of the application will decrease. It won't have much impact.
  • the prediction device does not directly perform the communication prediction strategy after detecting that the terminal device is about to enter the weak signal area, but also determines whether it is necessary to activate the communication prediction strategy based on the status information of the running application.
  • the status information of the running application may include the currently running service type, image quality (resolution), transmission rate, frame rate, etc. of the running application.
  • the trigger reference value represents the longest startup time for the running application to be triggered for pre-caching. If the prediction duration is greater than or equal to the trigger baseline value, it means that the prediction process needs to be triggered so that the application module can pre-cache the running application to maintain the smoothness of the currently running business type. If the prediction duration is less than the trigger baseline value, it means that the running application itself has continued to preload data for a period of time in the future under normal circumstances. Therefore, there is no need to trigger the prediction process, which reduces additional traffic consumption.
  • the prediction device obtains the status information of the running application; determines the trigger reference value based on the status information; if the prediction duration is greater than or equal to the trigger reference value, generates an execution strategy; causing the application module to adopt the execution strategy for the running application
  • the application is pre-cached.
  • the status information of the running application includes: the running configuration information of the running application; the above S401 can also be implemented in the following manner. Obtain the historical running configuration information of the running application and the current running configuration information of the running application; the historical running configuration information is the running configuration information in the historical time period adjacent to the current moment; based on the historical running configuration information and the current running configuration information, determine The running configuration information of the running application.
  • the status information includes running configuration information
  • the running configuration information includes but is not limited to: resolution, transmission rate, frame rate, etc.
  • Running configuration information has time-varying characteristics (for example, different videos in short videos will have different resolutions).
  • the historical running configuration information and the current running configuration information are averaged as the running configuration information.
  • the resolution, transmission rate, frame rate, etc. can be taken as the current value (corresponding to the current running configuration information) and the values within the past 5 seconds.
  • the values (corresponding to historical running configuration information) are averaged to obtain the running configuration information.
  • the historical running configuration information is the running configuration information in the historical time period adjacent to the current time, for example, the running configuration information corresponding to each second within 5 seconds adjacent to the current time.
  • the running configuration information of the running application is determined, which improves the accuracy of the running configuration information.
  • the status information of the running application includes: the current running service type of the running application and the running configuration information of the running application; the above S401 can also be implemented in the following manner.
  • the target configuration level is determined based on the running configuration information of the running application and the mapping relationship between the preset configuration interval and the preset level; the target configuration level represents the location of the running configuration information.
  • continuity indicators represented by numerical values.
  • the trigger reference value is determined based on the continuity index and the judgment conditions for determining whether pre-caching is required. At this time, these indicators can be simply divided into several grades as judgment conditions.
  • the mapping relationship between the preset configuration interval and the preset level represents the corresponding relationship between the configuration interval and the preset level.
  • the running configuration information is the resolution as an example for explanation. According to The size of the resolution is divided into intervals, and the number of resolution intervals can be multiple. For example, if the resolution interval is ⁇ 720P, its corresponding level is Level 1, and if the resolution interval is ⁇ 720P, its corresponding level is Level 2.
  • the mapping relationship between the preset level and the preset reference value represents the corresponding relationship between the level and the preset reference value. For example, the reference value corresponding to level 1 is 20 seconds, and the reference value corresponding to level 2 is 30 seconds.
  • interval division method of the running configuration information the mapping relationship between the preset configuration interval and the preset level, the mapping relationship between the preset level and the preset baseline value, and the size of the preset baseline value can all be used.
  • Those skilled in the art can make appropriate settings according to actual needs and are not limited to the mapping relationships listed above.
  • the embodiments of this application are only illustrative and not limiting.
  • the target configuration level is determined by the numerical value of the running configuration information.
  • the target level reflects the playback requirements of the running configuration information, which in turn affects the size of the triggering reference value.
  • the status information is a continuous value with a high degree of discreteness, which is classified hierarchically to reduce the degree of hashing. Setting the status information located in a certain interval to the same level reduces the resource consumption required to process a large amount of data and improves data processing efficiency.
  • the status information of the running application includes: the currently running service type of the running application; before the above S402, the communication prediction method further includes the following steps. Obtain the lag characteristics of multiple application samples; determine the initial trigger reference value of each application sample based on the lag characteristics of multiple application samples; where the initial trigger reference value represents the behavior of the application sample after entering the target area under preset operating conditions. Uptime.
  • the above S401 can also be implemented in the following ways. Determine the initial trigger baseline value of the running application based on the initial trigger baseline value of each application sample, or determine the initial trigger baseline value of the running application based on the current running service type of the running application and the initial trigger baseline value of each application sample. ; Use the initial trigger baseline value of the running application as the trigger baseline value of the running application.
  • the application's stuck characteristics are used to characterize the application's performance under specific conditions (for example: long video/short video, image quality, bit rate, transmission bandwidth, throughput, etc.) once it enters a weak signal area. , how long does it take for the lag to appear.
  • specific conditions for example: long video/short video, image quality, bit rate, transmission bandwidth, throughput, etc.
  • Figure 9 is an exemplary schematic diagram of the lagging feature of an application provided by the embodiment of the present application.
  • the long video playback of Application 1 enters the weak signal area from the normal network speed, for example, After the speed is limited to 1 kilobit per second (Kbps), lag begins to occur after an average of 78.3 seconds.
  • the lag characteristics of Application 1 include but are not limited to the long video lag delay (s) of Application 1 in Figure 9.
  • the short video playback of Application 2 enters the weak signal area from the normal network speed. For example, after the speed is limited to 1 kilobit per second (Kbps), lag begins to occur after an average of 38.8 seconds.
  • the lag characteristics of Application 2 include: Not limited to the short video freezing delay (s) of Application 2 in Figure 9.
  • kbps refers to the transmission rate of digital signals, and can also represent the transmission speed of the network, and Avg represents the time average.
  • the average freezing delay among the freezing characteristics of multiple corresponding application samples is used as the initial triggering reference value of the application type.
  • the average stall delay of multiple applications 1 is used as the initial triggering reference value of application 1.
  • the corresponding The initial trigger reference values can be different. Therefore, for each application sample, when obtaining its lag characteristics, the lag characteristics of the application sample when running different business types are obtained. Based on this, when determining the initial trigger baseline value, based on the current running business type of the running application and the initial trigger baseline value of each application sample to determine the initial trigger baseline value of the running application.
  • the average lagging delay of multiple applications 2 when playing short videos is used as the initial triggering reference value of application 2 when playing short video services
  • the average delay of multiple applications 2 when playing long videos is The average lag delay is used as the initial triggering baseline value for Application 2 when playing long video services.
  • the solution provided by the embodiment of the present application uses the statistical characteristics of stuttering of different applications to construct criteria for different applications to trigger QoS prediction under different conditions.
  • the terminal device combines the general QoS map and the trigger to perform QoS prediction. Criteria determine whether to perform QoS prediction services.
  • the QoS prediction service has the ability to adapt to different applications, avoid triggering unnecessary preloading and precaching processes, reduce traffic, power consumption and memory usage, and improve the overall performance of terminal equipment.
  • the embodiment of the present application proposes a method for determining whether to perform QoS prediction based on application lag characteristics and a universal QoS map, thereby increasing the flexible and elastic use of the universal QoS map.
  • Figure 10 is an exemplary flow chart of another communication method based on QoS map construction provided by an embodiment of the present application.
  • S14 in the above-mentioned Figure 2 may also include S141-S144.
  • the above QoS map can be used for QoS prediction. Therefore, the above QoS map can be called a universal QoS map.
  • the general QoS map can be the QoS map constructed in S12 in Figure 2, or can be obtained through other QoS map construction methods, as long as other QoS maps include information that can directly or indirectly infer that the mobile terminal enters a weak signal area. The location or time point, as well as the relevant information about the time required to cross the weak signal area.
  • QoS prediction startup benchmark values corresponding to different application APP configurations can also be combined with other conditions (for example, application service type, image quality/resolution, transmission rate and frame rate, etc.) to perform QoS prediction startup. Configuration of baseline values.
  • the QoS prediction startup benchmark value is used to determine whether to perform QoS prediction. That is, when the predicted weak signal duration is less than or equal to the startup benchmark value, QoS prediction will not be started; conversely, if the predicted weak signal duration is greater than the startup benchmark value, QoS prediction will be started. QoS prediction.
  • Table 1 is a QoS prediction execution criterion provided by the embodiment of the present application.
  • Table 1 shows the startup baseline values of some application configurations.
  • the criterion corresponding to number 1 indicates APP#A: When the service type being played is long video, the resolution is greater than or equal to 720P, the transmission rate is greater than or equal to 20Kbps, and the frame rate is greater than or equal to 30fps, if the terminal device is about to enter a weak signal area, If the predicted weak signal duration exceeds 30 seconds, the QoS prediction service will be triggered; if the predicted duration is less than 30 seconds, the QoS prediction service will not be triggered.
  • the criterion corresponding to number 2 indicates APP#A:
  • the resolution is less than 720P
  • the transmission rate is less than 20Kbps
  • the frame rate is less than 30fps
  • the criterion corresponding to number 4 indicates that under the same application (APP#A), without distinguishing other conditions, if the weak signal in front is maintained for more than 15 seconds, QoS will be triggered. Prediction services.
  • the criterion corresponding to number 3 means that under the same application (APP#A), as long as a short video is played, regardless of other conditions, if the weak signal in front is maintained for more than 18 seconds, the QoS prediction service will be triggered.
  • the meanings of numbers 5 to 7 in Table 1 can be understood with reference to the descriptions of numbers 1 to 4 above, and will not be described again here.
  • the prediction device in the terminal device determines whether to perform QoS prediction based on the current positioning position, the general QoS map, the currently running application, and the QoS prediction criteria.
  • the universal QoS map terminal device constructed based on S12 in Figure 2 when the universal QoS map terminal device constructed based on S12 in Figure 2 is at a specific location (and a specific time period), it obtains the low QoS data of the road section ahead, and then performs QoS prediction based on different application criteria and general The QoS map determines whether to perform QoS prediction. If the judgment result of the prediction device is that QoS prediction needs to be performed, the application module executes the preloading/cache release strategy of the currently running application (for example, video APP) to improve the smoothness of video playback (that is, the video playback lag rate is reduced) .
  • the preloading/cache release strategy of the currently running application for example, video APP
  • the embodiment of the present application is based on a universal QoS map and adapted according to the lag characteristics of different applications to reduce unnecessary preloading processes.
  • trigger criteria criteria for different applications to perform QoS prediction are constructed.
  • the terminal decides whether to perform QoS prediction based on the current positioning result, currently running application, QoS map and triggering reference value.
  • An execution method that determines whether to perform QoS prediction based on the application's stuck characteristics and a general QoS map.
  • the QoS prediction is executed more strategically. QoS prediction is only performed when it is pre-determined that stuck will occur when the user passes through a weak signal area. If If stuttering does not occur, QoS prediction will not be performed, which avoids triggering the prediction process and causing more video data preloading, reducing extra traffic consumption, extra power consumption, and memory usage.
  • the communication prediction method further includes the following steps. Receive execution feedback information for the execution policy sent by the application module; send execution feedback information to the server; wherein the execution feedback information is used to update the communication quality map.
  • the execution strategy is used to cause the application module to pre-cache the running application.
  • the application module pre-cache the running application, it also sends execution feedback information to the prediction device.
  • the prediction device can perform the pre-cache according to the execution feedback.
  • the information can obtain the updated status information of the running application, so that the next communication prediction strategy for the running application can be predicted based on the updated status information of the running application.
  • the execution feedback information is sent to the server through the prediction device, so that the server updates the communication quality map according to the execution feedback information.
  • the communication prediction strategy carries communication quality alarm information with a predicted duration; the communication quality alarm information is used to enable the application module to determine whether to adopt the corresponding execution strategy based on the status information of the running application.
  • the communication prediction strategy is communication quality alarm information, which is sent to the application module through the prediction device to inform the application module that it is about to enter the weak signal area.
  • the application module determines whether to perform preloading based on its own status information.
  • the judgment process Refer to the corresponding description on the prediction device side above.
  • the trigger reference value is obtained according to S401 and S402, and then a judgment is made based on the relationship between the prediction duration and the trigger reference value, which will not be described again here.
  • the communication prediction strategy carries communication quality alarm information with a predicted duration; after the above S303, the communication prediction method further includes the following steps. Receive application feedback information sent by the application module after taking or not taking the execution policy; send application feedback information to the server; wherein the application feedback information is used to update the communication quality map.
  • the prediction device can predict the next alarm of the same running application based on the execution results in the application feedback information. In other words, whether the application performs preloading can be used as a reference for predicting the next alarm of the device. For example, the application does not respond to this alarm information and does not preload.
  • the predicted duration carried by this alarm is 20 seconds. Then the next time the same application is run in the same scenario, if the predicted duration is less than or equal to 20 seconds, then Do not send alarm information to the application module.
  • the application feedback information is sent to the server through the prediction device, so that the server updates the communication quality map according to the application feedback information.
  • the application module always does not adopt any optimization strategy for the alarms sent by the prediction device, indicating that the weak signal area has no impact on the application. Therefore, the weak signal area can be considered The area is removed from the communication quality map used for this application.
  • the QoS prediction service (corresponding to the prediction device) in the terminal device and the third-party application (corresponding to the application module) in the terminal device can be connected through software interfaces, etc., thereby realizing the QoS prediction service and the three-party application interaction.
  • the interaction methods include the following two methods. One is when the QoS prediction service learns that a low-QoS road section is about to appear ahead based on the communication quality map. The QoS prediction service sends a warning to the third-party application that the low-QoS road section is about to enter. The third-party application determines whether the road section is about to enter.
  • Preloading is required for judgment; the other is when the QoS prediction service learns based on the communication quality map that a low QoS road section will appear ahead, the QoS prediction service determines whether the third-party application needs to be preloaded based on the status information of the running application. When it is determined that a third-party application needs to be preloaded, the QoS prediction service sends an execution policy to the third-party application. Description will be made below with reference to Figures 11 and 12.
  • Figure 11 is an exemplary schematic diagram of the interaction between a QoS prediction service and a third-party application provided by an embodiment of the present application.
  • the QoS prediction service sends a low QoS alarm to the third-party application that the low QoS road section is about to enter ahead (in Figure 11, sending a QoS prediction alarm indicates sending communication quality alarm information);
  • the third-party application takes corresponding countermeasures.
  • the corresponding countermeasure is to start preloading, or the third-party application makes a judgment based on the above QoS prediction execution criteria. If it is judged that the low QoS road section ahead will not affect the running application, No action is taken.
  • the third-party application then feeds back the corresponding countermeasures taken to the QoS prediction service (shown as a feedback alarm execution result in Figure 11); the third-party application also feeds back the final user quality of experience (Quality of Experience, QoE) to the QoS prediction service (shown as feedback application QoE in Figure 11), the QoS prediction service can obtain two kinds of feedback from three-party applications, one is the feedback of the optimization strategy (the response of the three-party application to the alarm), including the status information of the running application; the other The first is QoE feedback. Both types of feedback can be used in the update process of the QoS map.
  • Application feedback information includes alarm execution results and application QoE.
  • Figure 12 is an exemplary schematic diagram of the interaction between a Qos prediction service and a third-party application provided by an embodiment of the present application.
  • the status information of the running application fed back by the third-party application is obtained through the QoS prediction service ( (shown as updating application status in Figure 12), when the status information of the running application meets the criteria for executing QoS prediction, the QoS prediction service sends the QoS execution policy to the third-party application (shown as sending QoS prediction policy in Figure 12).
  • the third-party application then feeds back the policy execution results to the QoS prediction service (shown as feedback policy execution results in Figure 12); the third-party application also feeds back QoE to the QoS prediction service (shown as feedback application QoE in Figure 12).
  • the QoS prediction service Two types of feedback can be obtained from third-party applications. One is feedback on execution strategies, including status information of running applications; the other is QoE feedback. Execution feedback information includes policy execution results and application QoE. Among them, the status information of the running application fed back by the three-party applications in Figure 11 and Figure 12 includes but is not limited to the application's service type, image quality (that is, resolution), transmission rate, frame rate, etc.
  • Figure 13 is an optional structural schematic diagram of a server provided by the embodiment of the present application.
  • the server 130 includes: a first acquisition part 1301 configured to acquire multiple communication quality data in multiple attribute dimensions; each communication quality data corresponds to attribute information in multiple attribute dimensions; the multiple attribute dimensions represent different scenarios or applications. Dimensions;
  • the first determining part 1302 is configured to determine the target communication quality data and the target time information corresponding to the target communication quality data based on multiple communication quality data and preset parameter threshold values; according to the target communication quality data in multiple attribute dimensions
  • the target attribute information and target time information determine the communication quality map; wherein the communication quality map is configured for the terminal device to predict communication quality for different applications in different scenarios.
  • the multiple attribute dimensions include at least two of the following: spatial location, time period, application type, operator, and communication standard.
  • the first acquisition part 1301 is also configured to acquire multiple initial communication quality data in multiple attribute dimensions; based on the multiple initial communication quality data, determine multiple initial communication quality data corresponding to each initial communication quality data. communication quality data.
  • the first determining part 1302 is also configured to group multiple communication quality data according to the consistency of attribute information, and determine at least one group of communication quality data; the attribute information between each group of communication quality data is different, The attribute information of each communication quality data in each group of communication quality data is consistent and continuous in time; based on the preset parameter threshold value, each group of communication quality data is extracted separately to determine the target communication quality corresponding to each group of communication quality data.
  • the preset parameter threshold includes: a preset threshold
  • the first determining part 1302 is also configured to determine, from each set of communication quality data, a plurality of first data that are less than or equal to the preset threshold value, and obtain multiple time points corresponding to each first data; if multiple If the first data is greater than or equal to the preset number of samples, multiple first data are determined as target communication quality data; target time information is determined based on multiple time points corresponding to each first data; target time information includes target signal duration and target signal start time.
  • the preset parameter threshold also includes: a preset ratio and a preset time period;
  • the first determining part 1302 is also configured to determine the data proportion of the plurality of first data to the total number of data of each group of communication quality data if the plurality of first data is greater than or equal to the preset sample number; From a piece of data, extract multiple second data whose data proportion is greater than or equal to a preset ratio, and obtain multiple time points corresponding to each second data based on multiple time points corresponding to each first data; based on each second data Multiple time points corresponding to the data determine each candidate duration corresponding to each second data; among the multiple second data, multiple third data whose candidate duration is greater than or equal to the preset time period are determined as target communications quality data, and based on the multiple time points corresponding to each second data, obtain multiple time points corresponding to each third data; based on the multiple time points corresponding to each third data, determine the target signals corresponding to the multiple third data Duration and target signal start time; target time information includes target signal duration and target signal start time.
  • the preset parameter threshold also includes: a preset deviation
  • the first determining part 1302 is further configured to determine the target deviation between a plurality of third data whose candidate duration is greater than or equal to the preset time period; in the plurality of third data, determine that the target deviation is less than or equal to the preset deviation Multiple target data; use multiple target data as target communication quality data, and obtain multiple time points corresponding to each target data; determine the target signal duration corresponding to multiple target data based on the multiple time points corresponding to each target data duration and target signal start time; target time information includes target signal duration and target signal start time.
  • the target deviation includes: a first date deviation and a first data deviation
  • the preset deviation includes: a preset date fluctuation difference and a preset data fluctuation difference
  • the first determining part 1302 is also configured to group the plurality of third data according to date and determine the first date deviation between the plurality of third data on different dates; among the plurality of third data, determine the first A plurality of fourth data whose date deviation is less than or equal to the preset date fluctuation difference; determining the first data deviation of each fourth data; among the plurality of fourth data, determining that the first data deviation is less than or equal to the preset data fluctuation difference Multiple target data.
  • the target deviation includes a second date deviation and a second data deviation
  • the preset deviation includes: a preset date fluctuation difference and a preset data fluctuation difference
  • the first determining part 1302 is further configured to determine second data deviations of a plurality of third data; among the plurality of third data, determine a plurality of fifth data whose second data deviation is less than or equal to the preset data fluctuation difference; Group the plurality of fifth data according to date to determine the second date deviation between the plurality of fifth data on different dates; among the plurality of fifth data, determine that the second date deviation is less than or equal to the preset date fluctuation difference of multiple target data.
  • the server 130 also includes a first receiving part 1303 and an updating part 1304;
  • the first receiving part 1303 is configured to receive feedback information based on the communication quality map sent by the terminal device;
  • the update part 1304 is configured to update the communication quality map based on the feedback information and generate an updated communication quality map; the updated communication quality map is configured to enable the terminal device to predict the next communication quality for different applications in different scenarios. .
  • the update part 1304 is also configured to adjust the preset parameter threshold value according to the feedback information to obtain the adjusted parameter threshold value; based on the adjusted parameter threshold value, re-determine the updated parameter threshold value.
  • the communication quality map is updated to complete the communication quality map; or, based on the feedback information, the information related to the running application in the communication quality map is updated to obtain the updated communication quality map.
  • the server provided in the above embodiment performs communication prediction
  • only the division of the above program modules is used as an example.
  • the above processing can be allocated to different program modules according to needs, that is, the device
  • the internal structure is divided into different program modules to complete all or part of the processing described above.
  • the server and server-side method embodiments provided in the above embodiments belong to the same concept. The specific implementation process and beneficial effects can be found in the method embodiments, and will not be described again here.
  • the embodiment of the present application also provides a prediction device, as shown in Figure 14.
  • Figure 14 is an optional structural schematic diagram of a prediction device provided by the embodiment of the present application.
  • the prediction device 140 includes: a second acquisition part 1401 configured to acquire the current location and the communication quality map sent by the server; the communication quality map is configured to provide different applications with different performance when the communication quality is lower than the preset communication quality. Corresponding attribute information and time information in the scene;
  • the second determination part 1402 is configured to determine the communication prediction strategy according to the communication quality map and the current location;
  • the sending part 1403 is configured to send the communication prediction strategy to the application module in the terminal device, so that the application module can perform communication prediction on the running application according to the communication prediction strategy.
  • the second determination part 1402 is also configured to determine the prediction area according to the current location and the communication quality map; wherein the prediction area is the target area that is about to enter; and determine the boundary position of the prediction area according to the communication quality map. , the starting position of entering the prediction area, the starting time of entering the prediction area, and the prediction duration in the prediction area; if the distance between the current position and the starting position is less than or equal to the distance between the boundary position and the starting position distance, or if the arrival time is after the starting time, the communication prediction strategy is determined based on the predicted duration; where the arrival time represents the time from the current position to the starting position.
  • the second acquisition part 1401 is also configured to acquire status information of a running application
  • the second determination part 1402 is also configured to determine the trigger reference value of the running application based on the status information of the running application; if the predicted duration is greater than or equal to the trigger reference value, generate an execution strategy; wherein the communication prediction strategy includes execution Policy, the execution policy is configured to cause the application module to pre-cache the running application.
  • the status information of the running application includes: running configuration information of the running application;
  • the second acquisition part 1401 is also configured to acquire the historical running configuration information of the running application and the current running configuration information of the running application; the historical running configuration information is the running configuration information in the historical time period adjacent to the current moment;
  • the second determination part 1402 is also configured to determine the running configuration information of the running application based on the historical running configuration information and the current running configuration information.
  • the status information of the running application includes: the current running service type of the running application and the running configuration information of the running application;
  • the second determination part 1402 is also configured to determine the target configuration level based on the running configuration information of the running application and the mapping relationship between the preset configuration interval and the preset level; the target configuration level represents the configuration level to which the running configuration information belongs. ; Determine the trigger baseline value based on the current running service type of the running application, the target configuration level, and the mapping relationship between the preset level and the preset baseline value.
  • the status information of the running application includes: the current running service type of the running application;
  • the second acquisition part 1401 is also configured to acquire the lag characteristics of multiple application samples
  • the second determination part 1402 is also configured to determine the initial trigger reference value of each application sample based on the stuck characteristics of the multiple application samples; wherein the initial trigger reference value represents the initial trigger reference value after the application sample enters the target area under preset operating conditions. Normal running time; determine the initial trigger baseline value of the running application based on the initial trigger baseline value of each application sample; or determine the initial trigger baseline value of the running application based on the current running business type of the running application and the initial trigger baseline value of each application sample.
  • Initial trigger baseline value use the initial trigger baseline value of the running application as the trigger baseline value of the running application.
  • the prediction device 140 includes a second receiving portion 1404;
  • the second receiving part 1404 is configured to receive execution feedback information for the execution policy sent by the application module
  • the sending part 1403 is configured to send execution feedback information to the server; wherein the execution feedback information is configured to update the communication quality map.
  • the communication prediction strategy is communication quality alarm information carrying a predicted duration; the communication quality alarm information is configured to enable the application module to determine whether to adopt the corresponding execution strategy based on the status information of the running application.
  • the second receiving part 1404 is configured to receive application feedback information sent by the application module after taking or not taking the execution policy
  • the sending part 1403 is configured to send application feedback information to the server; where the application feedback information is configured to evaluate the communication quality.
  • the map is updated.
  • the second acquisition part 1401 is also configured to obtain the current location through positioning; send a full map download request to the server to obtain the communication quality map sent by the server in response to the full map download request; or send an incremental map download Request to the server to obtain the incremental communication quality map sent by the server in response to the incremental map download request; based on the incremental communication quality map, update the obtained current communication quality map to obtain the communication quality map.
  • FIG 15 is an optional structural schematic diagram of a terminal device provided by an embodiment of the present application.
  • the terminal The device 150 includes the prediction device 140 and the application module 1405 in the above-mentioned Figure 14.
  • the prediction device 140 is configured to obtain the current location and the communication quality map sent by the server; the communication quality map is configured to provide communication quality that is lower than the preset communication quality. Under the circumstances, the corresponding attribute information and time information of different applications in different scenarios; determine the communication prediction strategy based on the communication quality map and current location; send the communication prediction strategy to the application module in the terminal device;
  • the application module 1405 is configured to perform communication prediction on running applications according to the communication prediction policy.
  • Figure 16 is a schematic structural diagram of a server proposed by the embodiment of the present application.
  • the server 160 proposed by the embodiment of the present application includes a first processor 1601, a processor that stores executable computer programs.
  • the first memory 1602 and the first processor 1601 are configured to implement the communication prediction method provided by the server side in the embodiment of the present application when executing the executable computer program stored in the first memory 1602.
  • the server 160 may further include a first communication interface 1603, and a first bus 1604 for connecting the first processor 1601, the first memory 1602, and the first communication interface 1603.
  • the first bus 1604 is used to connect the first communication interface 1603, the first processor 1601, and the first memory 1602 to implement mutual communication between these devices.
  • Figure 17 is a schematic structural diagram of a terminal device proposed by the embodiment of the present application.
  • the terminal device 170 proposed by the embodiment of the present application includes a first processor 1701, a storage executable computer
  • the first memory 1702 of the program and the first processor 1701 are used to implement the communication prediction method provided by the prediction device side in the embodiment of the present application when executing the executable computer program stored in the first memory 1702 .
  • the terminal device 170 may further include a first communication interface 1703, and a first bus 1704 for connecting the first processor 1701, the first memory 1702, and the first communication interface 1703.
  • the first bus 1704 is used to connect the first communication interface 1703, the first processor 1701, and the first memory 1702 to realize mutual communication between these devices.
  • the above-mentioned processors can be an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a digital signal processor (Digital Signal Processor, DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (ProgRAMmable Logic Device, PLD), Field Programmable Gate Array (Field ProgRAMmable Gate Array, FPGA), Central Processing Unit (CPU), At least one of a controller, a microcontroller, and a microprocessor. It can be understood that for different devices, the electronic device used to implement the above processor function may also be other, which is not specifically limited in the embodiment of the present application.
  • ASIC Application Specific Integrated Circuit
  • DSP Digital Signal Processor
  • DSPD Digital Signal Processing Device
  • PLD Programmable Logic Device
  • Field Programmable Gate Array Field ProgRAMmable Gate Array
  • CPU Central Processing Unit
  • the memory (first memory 1602 and first memory 1702) is used to store executable computer programs and data.
  • the executable computer program includes computer operating instructions.
  • the memory may include high-speed RAM memory, Non-volatile memory, such as at least two disk memories, may also be included.
  • first memory 1602 and first memory 1702 can be a volatile memory (volatile memory), such as a random access memory (Random-Access Memory, RAM); or a non-volatile memory ( non-volatile memory), such as Read-Only Memory (ROM), flash memory (flash memory), Hard Disk Drive (HDD) or Solid-State Drive (SSD); or the above types
  • volatile memory such as a random access memory (Random-Access Memory, RAM)
  • non-volatile memory non-volatile memory
  • ROM Read-Only Memory
  • flash memory flash memory
  • HDD Hard Disk Drive
  • SSD Solid-State Drive
  • each functional module in this embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.
  • the above integrated units can be implemented in the form of hardware or software function modules.
  • the integrated unit is implemented in the form of a software function module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium.
  • the technical solution of this embodiment is essentially The contribution made to the prior art or all or 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 a number of instructions to enable a computer device (which can be a personal computer).
  • a computer, server, or network device, etc.) or processor executes all or part of the steps of the method in this embodiment.
  • the aforementioned storage media include: U disk, mobile hard disk, read only memory (Read Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other various media that can store program codes.
  • Embodiments of the present application provide a computer-readable storage medium that stores a computer program for implementing the communication prediction method described in any of the above embodiments when executed by a processor.
  • the program instructions corresponding to a communication prediction method in this embodiment can be stored on a storage medium such as an optical disk, a hard disk, a USB flash drive, etc., when the program instructions corresponding to a communication prediction method in the storage medium are When the electronic device reads or is executed, the communication prediction method as described in any of the above embodiments can be implemented.
  • a storage medium such as an optical disk, a hard disk, a USB flash drive, etc.
  • embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, magnetic disk storage and optical storage, etc.) embodying computer-usable program code therein.
  • a computer-usable storage media including, but not limited to, magnetic disk storage and optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory that causes a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction means, the instructions
  • the device implements the functions specified in implementing one process or multiple processes in the flow diagram and/or one block or multiple blocks in the block diagram.
  • These computer program instructions may also be loaded onto a computer or other programmable data processing device, causing a series of operating steps to be performed on the computer or other programmable device to produce computer-implemented processing, thereby executing on the computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in implementing a process or processes of the flowchart diagram and/or a block or blocks of the block diagram.
  • Embodiments of the present application provide a communication prediction method, server, prediction device, terminal equipment and storage medium.
  • multiple communication quality data in multiple attribute dimensions are obtained; each communication quality data corresponds to attribute information in multiple attribute dimensions; the multiple attribute dimensions represent the dimensions of different scenarios or applications; according to Multiple communication quality data and preset parameter threshold values are used to determine target communication quality data and target time information corresponding to the target communication quality data.
  • the area corresponding to the target communication quality data represents the weak signal area.
  • the weak signal areas that need attention are filtered out according to the preset parameter threshold values.
  • the relevant characteristics carried by the weak signal area (for example, application, spatial location, weak signal time information, etc.) , can adapt to different applications and different scenarios.
  • the communication quality map is determined; among them, the communication quality map is used by the terminal device to predict communication quality for different applications in different scenarios.

Abstract

本申请实施例公开了一种通信预测方法、服务器、预测装置、终端设备和存储介质。该方法包括：获取多个属性维度上的多个通信质量数据；每个通信质量数据对应多个属性维度下的属性信息；多个属性维度表征不同场景或应用的维度；根据多个通信质量数据和预设参数门限值，确定目标通信质量数据和目标通信质量数据对应的目标时间信息；根据目标通信质量数据在多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。

Description

通信预测方法、服务器、预测装置、终端设备和存储介质
相关申请的交叉引用
本申请基于申请号为202211060668.6、申请日为2022年08月30日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此引入本公开作为参考。
技术领域
本申请涉及通信技术领域，尤其涉及一种通信预测方法、服务器、预测装置、终端设备和存储介质。
背景技术
随着网络技术的快速发展和智能设备的广泛普及，终端设备上的功能(例如，应用软件、应用程序、应用)越来越丰富，终端设备可以通过网络从各种功能的客户端获取各种信息，例如，观看视频、收听音乐、看文件等。终端设备需要连接网络使用客户端，由于基站覆盖度较小，或者处于室内、搭乘地铁出行等原因，可能会出现网络信号较弱或者无网络信号的情况。因此，当终端设备进入网络信号较弱或者无网络信号的区域时，网络基本处于不可用状态，从而出现无法正常使用客户端的情况。
目前，通过预先采集并分析终端设备在固定出行路线移动时，网络信号质量(例如，空口传输时延、实时空口带宽、实时空口速率和丢包率等)的变化情况，学习出终端设备即将进入弱信号区域的时间，以及离开该弱信号区域的时间，也就是构建一个针对该终端设备的通信质量地图的过程。
然而，上述通信质量地图生成方法过于简单，降低了通信质量地图的通用性。
发明内容
本申请实施例提供一种通信预测方法、服务器、预测装置、终端设备和存储介质，提高了通信质量地图的通用性。
本申请实施例的技术方案是这样实现的：
第一方面，本申请实施例提供一种通信预测方法，所述方法包括：获取多个属性维度上的多个通信质量数据；每个所述通信质量数据对应所述多个属性维度下的属性信息；所述多个属性维度表征不同场景或应用的维度；根据多个所述通信质量数据和预设参数门限值，确定目标通信质量数据和所述目标通信质量数据对应的目标时间信息；根据所述目标通信质量数据在所述多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，所述通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。
第二方面，本申请实施例提供另一种通信预测方法，所述方法包括：获取当前位置，以及服务器发送的通信质量地图；所述通信质量地图用于提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息；根据所述通信质量地图和所述当前位置，确定通信预测策略；向终端设备中的应用模块发送所述通信预测策略，供所述应用模块根据所述通信预测策略对正在运行应用进行通信预测。
第三方面，本申请实施例提供一种服务器，所述服务器包括：第一获取部分，被配置为获取多个属性维度上的多个通信质量数据；每个所述通信质量数据对应所述多个属性维度下的属性信息；所述多个属性维度表征不同场景或应用的维度；第一确定部分，被配置为根据多个所述通信质量数据和预设参数门限值，确定目标通信质量数据和所述目标通信质量数据对应的目标时间信息；根据所述目标通信质量数据在所述多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，所述通信质量地图被配置为终端设备对不同应用在不同场景下进行通信质量预测。
第四方面，本申请实施例提供一种预测装置，所述预测装置包括：第二获取部分，被配置为获取当前位置，以及服务器发送的通信质量地图；所述通信质量地图被配置为提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息；第二确定部分，被配置为根据所述通信质量地 图和所述当前位置，确定通信预测策略；发送部分，被配置为向终端设备中的应用模块发送所述通信预测策略，供所述应用模块根据所述通信预测策略对正在运行应用进行通信预测。
第五方面，本申请实施例提供一种终端设备，所述终端设备包括：如第四方面所述的预测装置和应用模块；所述应用模块，被配置为根据所述通信预测策略对正在运行应用进行通信预测。
第六方面，本申请实施例提供一种服务器，所述服务器包括：第一存储器，用于存储可执行计算机程序；第一处理器，用于执行所述第一存储器中存储的可执行计算机程序时，实现上述第一方面的通信预测方法。
第七方面，本申请实施例提供一种终端设备，所述终端设备包括：第二存储器，用于存储可执行计算机程序；第二处理器，用于执行所述第二存储器中存储的可执行计算机程序时，实现上述第二方面的通信预测方法。
第八方面，本申请实施例提供一种计算机可读存储介质，存储有计算机程序，用于被处理器执行时，实现上述第一方面或第二方面的通信预测方法。
本申请实施例提供了一种通信预测方法、服务器、预测装置、终端设备和存储介质。根据本申请实施例提供的方案，获取多个属性维度上的多个通信质量数据；每个通信质量数据对应多个属性维度下的属性信息；多个属性维度表征不同场景或应用的维度；根据多个通信质量数据和预设参数门限值，确定目标通信质量数据和目标通信质量数据对应的目标时间信息。目标通信质量数据对应的区域表示弱信号区域，根据预设参数门限值将需要关注的弱信号区域筛选出来，弱信号区域携带的相关特性(例如，应用、空间位置、弱信号时间信息等)，可以适配不同应用和不同场景。根据目标通信质量数据在多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。根据有限的测试生成通信质量地图，即可针对不同的应用、不同的场景(例如，地铁出行场景、室内场景、偏远场景、上次班高峰场景等)，采用该通信质量地图判断是否启动通信质量预测，提高了通信质量地图的通用性。
附图说明
图1为本申请实施例提供的一种通信预测方法的可选的步骤流程图；
图2为本申请实施例提供的一种基于QoS地图构建的通信预测方法的示例性的流程示意图；
图3为本申请实施例提供的另一种通信预测方法的可选的步骤流程图；
图4为本申请实施例提供的一种低QoS信息抽取算法的示例性的流程示意图；
图5为本申请实施例提供的一种低QoS区域的示例性的示意图；
图6为本申请实施例提供的一种基于三方应用的反馈信息的QoS地图更新的示例性的示例图；
图7为本申请实施例提供的再一种通信预测方法的可选的步骤流程图；
图8为本申请实施例提供的又一种通信预测方法的可选的步骤流程图；
图9为本申请实施例提供的一种应用的卡顿特征的示例性的示意图；
图10为本申请实施例提供的另一种基于QoS地图构建的通信预测方法的示例性的流程示意图；
图11为本申请实施例提供的一种Qos预测服务与三方应用交互的示例性的示意图；
图12为本申请实施例提供的另一种Qos预测服务与三方应用交互的示例性的示意图；
图13为本申请实施例提供的一种服务器的可选的结构示意图；
图14为本申请实施例提供的一种预测装置的可选的结构示意图；
图15为本申请实施例提供的一种终端设备的可选的结构示意图
图16为本申请实施例提供的一种服务器组成结构示意图；
图17为本申请实施例提供的一种终端设备组成结构示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述。应当理解的是，此处所描述的一些实施例仅仅用以解释本申请的技术方案，并不用于限定本申请的技术范围。
为了更好地理解本申请实施例中提供的通信预测方法，在对本申请实施例的技术方案进行介绍之前，先对相关技术进行说明。
相关技术中，在学习出终端设备即将进入弱信号区域的时间，以及离开该弱信号区域的时间之后，当终端设备在该出行路线移动时，预测终端设备即将进入弱信号区域，以及该弱信号区域覆盖的时间长 度。根据上述预测信息，当某个上网业务(例如，视频APP)提前知道通信服务质量即将变差，可及早采取应对措施，例如，触发视频预加载，以提升上网业务的流畅度。
然而，通过实时监控网络信号质量，所考虑的网络信号质量是以空口为主的信号质量进行度量，未能反应人流因素所造成的通信服务质量影响。示例性的，针对地铁出行的出行场景，通信服务质量较差的区域通常发生在人流高峰时段(例如，上下班通勤时段)，在该场景下容易漏检弱信号区域。该方案中构建的通信质量(Quality of Service，QoS)地图的属性维度低，例如，没有考虑到场景、应用等维度，预测效果有限，没有考虑到QoS地图在针对不同场景与不同应用时的适配性。而且，未考虑QoS地图的更新策略，也未考虑与三方应用之间交互，以及利用三方应用的反馈信息更新QoS地图。QoS是指客户端对通信服务整体性能的描述或度量，为了量化衡量服务质量，通常会考虑几个相关指标，例如，丢包、码率、吞吐量、传输延迟、可用性、抖动等。
相关技术中提供的应用数据预加载方法，示例性的，移动终端通过全球定位系统(Global Positioning System，GPS)实时获取定位位置，根据当前所在位置以及基于GPS所构建的弱信号地图，判断是否即将进入弱信号区域。若判定即将进入弱信号区域，则基于弱信号地图计算穿过弱信号区域所需时长，并根据该时长提前下载并缓存应用数据，以保证弱信号区域内该应用的流畅度。
然而，上述应用数据预加载的方法是通过GPS获取实时位置，并基于GPS构建弱信号地图，由于GPS信号只能在户外接收，因此，不适用于室内、地铁和地下路段的场景，降低了弱信号地图的适用性。
本申请实施例提供一种通信预测方法，如图1所示，图1为本申请实施例提供的一种通信预测方法的步骤流程图，该通信预测方法可以由服务器执行，通信预测方法包括以下步骤：
S101、获取多个属性维度上的多个通信质量数据；每个通信质量数据对应多个属性维度下的属性信息；多个属性维度表征不同场景或应用的维度。
在本申请实施例中，针对某个场景，采集该场景下终端设备的通信质量数据，通信质量数据包括该场景对应的多个属性维度以及该场景下终端设备的网络质量；该场景对应的多个属性维度可以包括空间位置、采集时间段、当前运行应用、终端设备的通信信号类型等，属性信息是属性维度的详细内容；网络质量大小表示该场景下终端设备连接信号的好坏，可以表现在当前运行应用的流畅度、加载速度等。
在本申请实施例中，对于某个通信质量数据来说，其不仅涵盖多个属性维度下的属性信息，属性信息是属性维度下的详细内容，例如，地铁站A、应用B、采集时间段是17:00-18:00、运营商C和5G信号，还涵盖该多个属性信息下的网络质量得分(可以通过通信速率、通信时延、吞吐率等表示)。通信质量数据可以表示成数值形式，若应用B播放很流畅(流畅度可以通过通信速率、通信时延、吞吐率等参数进行确定)，则通信质量数据对应的数值较高。多个属性维度的组合可以构成不同的场景。
在本申请一些实施例中，多个属性维度包括以下至少两个：空间位置、时间段、应用类型、运营商和通信制式。
在本申请实施例中，空间位置表示该通信质量数据对应的区域，例如，某一站牌、某一地铁站，某一写字楼、某一商场等；时间段表示该通信质量数据对应的时间，包括但不限于日期、某天中的某一时间段；应用类型表示通信质量数据对应的应用；通信制式表示终端设备上网时的通信格式。
在本申请一些实施例中，上述图1中S101可以通过以下方式实现。获取多个属性维度上的多个初始的通信质量数据；基于多个初始的通信质量数据，确定各个初始的通信质量数据对应的多个通信质量数据。
在本申请实施例中，初始的通信质量数据的数据格式包括时间、空间(位置)、应用、用户、运营商、通信制式等多个维度。通信质量数据在同一维度下包括不同的表现形式，示例性的，对于空间位置维度，其可以包括不同的路段、不同的地铁站、不同的室内等；对于时间维度，其可以包括一天当中不同时间；对于应用维度，其可以包括用于播放长视频的不同的应用软件、用于播放短视频的不同的应用软件、用于播放音频的不同的应用软件、用于信息推送的不同的应用软件、不同的社交软件等；对于运营商维度，其可以包括不同的运营商；对于通信制式维度，其表征通信方式，可以包括第四代移动通信技术(the 4th generation mobile communication technology，4G)、5G、无线保真技术(Wi-Fi)等。
在本申请实施例中，初始的通信质量数据还包括不同的表现形式下对应的网络质量参数，在本申请实施例中，根据网络质量参数，利用质量打分方法，对初始的通信质量数据进行打分，得到通信质量数据。网络质量参数包括但不限于通信速率、通信时延、吞吐率、其他表征通信质量的指标以及通过关联这些指标所形成的QoS衡量指标(QoS metric)，根据QoS metric可以对初始的通信质量数据进行打分，将根据QoS metric计算得出的数值作为通信质量数据。质量打分方法可以是将通信速率、通信时延、吞吐率等指标进行加权求和，也可以是其他计算方法，对此本申请实施例不做限制。
示例性的，以初始的通信质量数据是QoS数据、通信质量地图是QoS地图为例进行说明，QoS数 据的采集方式可以是人工采集或众包采集等方式。人工采集是指由测试人员在指定时间、指定空间进行QoS数据的采集。众包采集包括显式与隐式，其中，显式是指向普通用户发布众包采集任务，通过用户领取任务，并将采集的程序部署到用户的终端设备中，让用户自行完成规定的采集任务，用户有感知的采集QoS数据；隐式是指在用户不需要主动参与的情况下进行采集，用户无感知的采集QoS数据，在连上Wi-Fi之后，终端可以主动向服务器进行上报，自动同步QoS数据。无论是人工采集还是众包采集，将采集的QoS数据上传服务器，使得服务器根据QoS数据构建QoS地图，而不是在终端设备侧进行弱信号区域的检测，避免终端设备资源消耗过大。
S102、根据多个通信质量数据和预设参数门限值，确定目标通信质量数据和目标通信质量数据对应的目标时间信息。
在本申请实施例中，多个通信质量数据包括大量的反映网络质量的数据，既有网络质量较好的数据，又有网络质量较差的数据。在构建通信质量地图时，将网络质量较差的数据作为关注数据，利用预设参数门限值将其筛选出来，从而得到网络质量较差的数据(对应于目标通信质量数据)。目标通信质量数据包括多个满足条件的通信质量数据。
在本申请实施例中，由于通信质量数据包括空间位置维度的属性信息，因此，可以确定通信质量数据对应的区域，同理，也可以确定筛选出来的目标通信质量数据对应的区域。由于目标通信质量数据是网络质量较差的数据，目标通信质量数据对应的区域可以表示弱信号区域或低QoS区域，根据预设参数门限值将需要关注的弱信号区域筛选出来。由于通信质量数据携带有时间信息(对应于属性信息)，因此还可以根据筛选出来的目标通信质量数据的时间信息，确定该弱信号区域的目标时间信息，目标时间信息包括但不限于弱信号持续时长和弱信号开始时间等。
S103、根据目标通信质量数据在多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。
在本申请实施例中，目标通信质量数据包括多个满足条件的通信质量数据，针对一个满足条件的通信质量数据来说，基于该通信质量数据对应的属性信息，将其标注在地图中，标注在地图中的不仅有网络质量差的区域信息、时间段信息、应用信息、运营商信息和通信制式信息。将多个这样的通信质量数据均标注在地图中，从而构成通信质量地图。
在本申请实施例中，目标通信质量数据对应的区域是弱信号区域，目标通信质量数据的多个属性维度包括但不限于空间位置、时间段、应用类型、运营商和通信制式，结合目标时间信息，构建通信质量地图。构建出来的通信质量地图是多个弱信号区域的集合，以及各个弱信号区域的属性维度，弱信号区域携带的相关特性(例如，应用、空间位置、弱信号时间信息等)，可以适配不同应用和不同场景。
在本申请实施例中，通信质量地图包括不同属性维度的弱信号区域，是多个属性维度的弱信号区域的集合，例如，关于某一应用的弱信号区域、关于某一时间段的弱信号区域、关于某一空间位置的弱信号区域、关于某一通信制式的弱信号区域、关于某一运营商的弱信号区域等。在构建通信质量地图之后，终端设备对不同应用在不同场景，均可以利用该通信质量地图进行通信质量预测，无需针对不同应用和不同场景分别构建通信质量地图，提高了通信质量地图的通用性。
需要说明的是，本申请实施例中的通信质量地图可以称为QoS地图、弱信号地图、通信语义地图、通信技术可用性地图(例如，4G可用性地图、5G可用性地图)，用户体验质量(Quality of Experience，QoE)地图、服务可用性地图(例如，虚拟现实(Virtual Reality，VR)服务可用性地图、V2X(vehicle to everything)服务可用性地图)、高峰/离峰地图、全球定位系统(Global Positioning System，GPS)信号强弱地图等。
可以理解的是，根据本申请实施例提供的方案，获取多个属性维度上的多个通信质量数据；每个通信质量数据对应多个属性维度下的属性信息；多个属性维度表征不同场景或应用的维度；根据多个通信质量数据和预设参数门限值，确定目标通信质量数据和目标通信质量数据对应的目标时间信息。目标通信质量数据对应的区域表示弱信号区域，根据预设参数门限值将需要关注的弱信号区域筛选出来，弱信号区域携带的相关特性(例如，应用、空间位置、弱信号时间信息等)，可以适配不同应用和不同场景。根据目标通信质量数据在多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。根据有限的测试生成通信质量地图，即可针对不同的应用、不同的场景(例如，地铁出行场景、室内场景、偏远场景、上次班高峰场景等)，采用该通信质量地图判断是否启动通信质量预测，提高了通信质量地图的通用性。
在本申请实施例中，上网业务可以依照所构建的QoS地图执行QoS预测。以初始的通信质量数据是QoS数据、通信质量地图是QoS地图、服务器是云服务器为例进行说明，如图2所示，图2为本申请实施例提供的一种基于QoS地图构建的通信预测方法的示例性的流程示意图，该通信预测方法包括S11-S14。
S11、采集QoS数据。
在本申请实施例中，QoS的定义包括但不限于通信速率、通信时延、吞吐率、其他表征通信质量的指标以及通过关联这些指标所形成的QoS衡量指标(QoS metric)，根据QoS metric可以对QoS数据进行打分，将根据QoS metric计算得出的数值作为QoS数据的打分数据。
需要说明的是，QoS数据的采集频率需要满足一定要求，例如，每一秒都需要有对应的QoS数据，这样可以使得构建的通信质量地图能够适配于不同的应用、不同的场景。
S12、构建QoS地图。
S13、将QoS地图部署到云服务器。
在本申请实施例中，可以将QoS地图下载服务部署在云服务器(也可以称为云端)，终端设备通过调用云端开放的应用程序接口(Application Programming Interface，API)进行QoS地图的下载。QoS地图的下载可以是针对城市的全量下载，也可以是针对某一空间位置、某一时间、某一运营商等变量的增量下载。
S14、QoS预测。
在本申请实施例中，可以通过使用QoS地图提升通信流畅度。QoS预测方式可以根据当前时间、位置、所使用应用等信息，利用QoS地图预测未来时间或未来位置的QoS质量，并将预测的结果通知给三方应用，以帮助其提前采取应对措施。
在本申请一些实施例中，上述图1中S102还可以包括S1021和S1022。如图3所示，图3为本申请实施例提供的另一种通信预测方法的可选的步骤流程图。
S1021、对多个通信质量数据按照属性信息一致性进行分组，确定至少一组通信质量数据；各组通信质量数据之间的属性信息不同，每组通信质量数据内各个通信质量数据的属性信息一致，且时间上连续。
在本申请实施例中，通信质量数据在同一维度下包括不同的表现形式(即属性信息不同)，不同的表现形式可以采用不同的字段表示，在进行分组时，可以根据关键字找出对应的数据，从而将属性信息一致的通信质量数据分为同一组，以便进行统计计算。属性信息中包括采集时间段，一个通信质量数据是关于某一时间段中各个时间点上的数据，也就是说通信质量数据在时间上是连续的。属性信息有多个，一组通信质量数据中的多个通信质量数据的属性信息均一一对应相同。各组通信质量数据之间的属性信息完全不同或部分不同。
示例性的，一组通信质量数据包括多个通信质量数据，该多个通信质量数据均是在站牌A、应用B、采集时间段是8:00-9:00、运营商C、第五代移动通信新空口(5G New Radio，5G NR)信号这些条件下的通信质量数据，也就是说该多个通信质量数据的属性信息一致。对于每个通信质量数据来说，其时间段对应8:00-9:00，在实际情况中，以秒为间隔时间为例，对于每秒，均有与之对应的数据，也就是说，该通信质量数据在时间上连续。
S1022、基于预设参数门限值，对各组通信质量数据分别进行抽取，确定各组通信质量数据分别对应的目标通信质量数据和目标通信质量数据对应的目标时间信息。
在本申请实施例中，对于每组通信质量数据，均基于预设参数门限值进行抽取，确定每组通信质量数据的网络质量较差的数据(对应于目标通信质量数据)，根据目标通信质量数据的属性维度中的空间位置确定弱信号区域，从而将需要关注的弱信号区域筛选出来。再结合目标通信质量数据对应的目标时间信息，确定该弱信号区域的弱信号持续时长、弱信号开始时间等。
可以理解的是，按照属性信息的一致性对多个通信质量数据进行分组，降低了统计数据的资源消耗量，然后针对每组通信质量数据，进行弱信号区域的筛选，提高了数据统计结果的准确性。
在本申请一些实施例中，预设参数门限值包括：预设门限值；上述图3中S1022可以包括S201-S203。
S201、从每组通信质量数据中，确定小于或等于预设门限值的多个第一数据，并获取各个第一数据对应的多个时间点。
S202、若多个第一数据大于或等于预设样本数量，则将多个第一数据确定为目标通信质量数据。
S203、基于各个第一数据对应的多个时间点，确定目标时间信息。
在本申请实施例中，预设门限值为用于判断通信质量数据是否为低QoS数据的门限值，即，通信质量数据低于什么门限值认定为低QoS数据。在每组通信质量数据中选取小于或等于预设门限值的多个第一数据，每个第一数据对应多个连续的时间点，多个连续的时间点可以构成某个时间段，根据多个连续的时间段可以确定弱信号持续时间和弱信号开始时间。
在本申请实施例中，多个第一数据对应的空间位置维度的属性信息一致，可以根据多个第一数据的空间位置维度的属性信息确定弱信号区域，从而得到弱信号区域的弱信号持续时间和弱信号开始时间。
在本申请实施例中，当多个第一数据大于或等于预设样本数量时，说明低QoS数据的数量足够多。 将多个第一数据确定为网络质量较差的数据(对应于目标通信质量数据)。当多个第一数据小于预设样本数量时，说明低QoS数据的数量有点少，不足以作为网络质量较差的数据。
需要说明的是，预设门限值可以由本领域技术人员根据实际情况和当前场景进行适当设置，可以是由大量的抽取过程的实验过程中进行确定的，对于不同场景可以设置不同的预设门限值。对此本申请实施例不做限制。
在本申请实施例中，通过设置预设门限值，筛选出低QoS数据(对应于多个第一数据)；通过设置预设样本数量，提高了数据统计的有效性。
在本申请一些实施例中，预设参数门限值还包括：预设比值和预设时间段；上述S201之后，该通信预测方法还可以包括S204-S208。
S204、若多个第一数据大于或等于预设样本数量，则确定多个第一数据占每组通信质量数据的数据总数量的数据占比。
S205、从多个第一数据中，抽取数据占比大于或等于预设比值的多个第二数据，并基于各个第一数据对应的多个时间点，获取各个第二数据对应的多个时间点。
在本申请实施例中，一组通信质量数据中存在低QoS数据以及其他QoS数据，低QoS数据和其他QoS数据都对应于多个连续时间点，会存在某一时间点上，虽然有低QoS数据，但是其所占的比重较小的情况，此种情况下不能将其认为是网络质量较差的数据(对应于目标通信质量数据)。针对每个时间点，计算低QoS数据与该时间点上数据总数量之间的数据占比，当数据占比大于或等于预设比值时，说明在该时间点上低QoS数据所占的比重较大，可以进行下一步的计算，提高了数据统计的准确性。
S206、根据各个第二数据对应的多个时间点，确定各个第二数据对应的各个候选持续时长。
S207、在多个第二数据中，将候选持续时长大于或等于预设时间段的多个第三数据，确定为目标通信质量数据，并基于各个第二数据对应的多个时间点，获取各个第三数据对应的多个时间点。
S208、基于各个第三数据对应的多个时间点，确定多个第三数据对应的目标信号持续时长和目标信号开始时间；目标时间信息包括目标信号持续时长和目标信号开始时间。
在本申请实施例中，第二数据是第一数据中的一部分，因此，可以基于各个第一数据对应的多个连续时间点，获取各个第二数据对应的多个连续时间点。同理，第三数据是第二数据中的一部分，可以获取各个第三数据对应的多个连续时间点。针对每个第二数据，根据第二数据对应的多个连续时间点，确定第二数据对应的候选持续时长。当候选持续时长大于或等于预设时间段时，说明该候选持续时长的弱信号持续时长可能会影响到当前运行应用的正常使用，将其认为是网络质量较差的数据(对应于目标通信质量数据)。根据候选持续时长和预设时间段之间的关系，在多个第二数据中抽取多个第三数据。将多个第三数据作为目标通信质量数据，然后基于各个第三数据对应的多个时间点，确定弱信号持续时长和弱信号开始时间。
需要说明的是，预设比值和预设时间段可以由本领域技术人员根据实际情况和当前场景进行适当设置，可以是由大量的抽取过程的实验过程中进行确定的，对于不同场景可以设置不同的预设比值和预设时间段。对此本申请实施例不做限制。
可以理解的是，多个第三数据对应的空间位置维度的属性信息一致，可以根据多个第三数据的空间位置维度的属性信息确定弱信号区域，从而得到弱信号区域的弱信号持续时间和弱信号开始时间。通过设置预设比值和预设时间段，提高了数据统计的有效性和准确性。
在本申请一些实施例中，预设参数门限值还包括：预设偏差；上述S206之后，该通信预测方法还可以包括S209-S212。
S209、确定候选持续时长大于或等于预设时间段的多个第三数据之间的目标偏差。
S210、在多个第三数据中，确定目标偏差小于或等于预设偏差的多个目标数据。
S211、将多个目标数据作为目标通信质量数据，并获取各个目标数据对应的多个时间点。
S212、根据各个目标数据对应的多个时间点，确定多个目标数据对应的目标信号持续时长和目标信号开始时间；目标时间信息包括目标信号持续时长和目标信号开始时间。
在本申请实施例中，在多个第二数据中，抽取候选持续时长大于或等于预设时间段的多个第三数据之后，通过计算多个第三数据之间的目标偏差，确定多个第三数据之间的波动性。数据之间的波动越小，说明数据具有参考性。在多个第三数据中，抽取目标偏差小于或等于预设偏差的多个目标数据，多个目标数据之间的波动性小，说明多个目标数据的性能稳定，具有参考性。将多个目标数据认为是网络质量较差的数据(对应于目标通信质量数据)。
在本申请实施例中，目标数据是第三数据中的一部分，因此，可以基于各个第三数据对应的多个连续时间点，获取各个目标数据对应的多个连续时间点。基于各个目标数据对应的多个时间点，确定弱信号持续时长和弱信号开始时间。
可以理解的是，多个目标数据对应的空间位置维度的属性信息一致，可以根据多个目标数据的空间位置维度的属性信息确定弱信号区域，从而得到弱信号区域的弱信号持续时间和弱信号开始时间。通过设置预设偏差，提高了数据统计的稳定性。
在本申请一些实施例中，目标偏差包括：第一日期偏差和第一数据偏差；预设偏差包括：预设日期波动差和预设数据波动差；上述S210还可以通过以下两个示例实现。示例一、将多个第三数据按照日期进行分组，确定在不同日期下多个第三数据之间的第一日期偏差；在多个第三数据中，确定第一日期偏差小于或等于预设日期波动差的多个第四数据；确定各个第四数据的第一数据偏差；在多个第四数据中，确定第一数据偏差小于或等于预设数据波动差的多个目标数据。
在本申请实施例中，多个第三数据的时间维度的属性信息包括不同日期下不同时间段，例如，第三数据是针对每天当中8:00-9:00这一时间段的通信质量数据，虽然多个第三数据的属性信息一致，但是其对应的日期可能不一样，因此，按照日期对多个第三数据进行分组，计算不同日期之间的日期偏差(对应于第一日期偏差)。如果不同日期下的日期偏差太大，说明这些数据在不同日期下的波动性较大，不稳定。需要在多个第三数据中，抽取日期波动较稳定(小于或等于预设日期波动差)的多个第四数据。然后计算多个第四数据之间的数据偏差(对应于第一数据偏差)，在多个第四数据中抽取数据波动较稳定(小于或等于预设数据波动差)的多个目标数据。
在本申请一些实施例中，目标偏差包括第二日期偏差和第二数据偏差，预设偏差包括：预设日期波动差和预设数据波动差；上述S210还可以通过示例二实现。确定多个第三数据的第二数据偏差；在多个第三数据中，确定第二数据偏差小于或等于预设数据波动差的多个第五数据；将多个第五数据按照日期进行分组，确定在不同日期下多个第五数据之间的第二日期偏差；在多个第五数据中，确定第二日期偏差小于或等于预设日期波动差的多个目标数据。
在本申请实施例中，计算多个第三数据之间的数据偏差(对应于第二数据偏差)，数据偏差越小，说明数据越稳定，在多个第三数据中抽取数据波动较稳定(小于或等于预设数据波动差)的多个第五数据。多个第五数据的时间维度的属性信息包括不同日期下不同时间段，多个第五数据的属性信息一致，但是其对应的日期可能不一样，因此，按照日期对多个第五数据进行分组，计算不同日期之间的日期偏差(对应于第二日期偏差)。在多个第五数据中抽取日期波动较稳定(小于或等于预设日期波动差)的多个目标数据。
下面，将说明本申请实施例在一个实际的应用场景中的示例性应用。
在本申请实施例中，以通信质量地图是QoS地图、通信质量数据是QoS数据、预设门限值是低QoS门限值、预设样本数量是N、预设比值是低QoS打分占比门限X％、预设时间段是低QoS打分的持续时间T、预设日期波动差是θ、预设数据波动差是σ为例进行说明，本申请实施例基于统计分析的方法构建QoS地图。在大量采集的QoS数据中提取出低QoS区域，以下进行说明。将QoS数据按照空间位置、时间段、应用、运营商、通信制式的属性维度进行分组。对每个分组中的QoS数据运行低QoS信息抽取算法，得到低QoS区域。每个低QoS区域不仅包含空间位置、时间段、应用、运营商、通信制式等属性信息外，还包括低QoS信息开始时间与结束时间，以及低QoS信息持续时间。如图4所示，图4为本申请实施例提供的一种低QoS信息抽取算法的示例性的流程示意图。图4是对QoS分组数据进行低QoS信息抽取算法，可以包括以下步骤：
S21、设置低QoS门限值，统计低于低QoS门限值的QoS样本数量。
S22、判断QoS样本数量是否大于或等于N；若是，则执行S23，若否，则结束抽取过程。
S23、统计每个分组内QoS打分低于低QoS门限值的低分次数占比。
S24、判断低分次数占比是否大于或等于X％；若是，则执行S25，若否，则结束抽取过程。
S25、统计每个分组内低QoS打分的持续时间。即，统计QoS打分持续低于低QoS门限值的时间。
S26、判断持续时间是否大于或等于T；若是，则执行S27，若否，则结束抽取过程。
S27、统计低QoS不同日期的不一致性。
在本申请实施例中，将每个分组内低QoS打分再按照日期分组，统计低QoS数据不同日期下的日期偏差。
S28、判断低QoS数据不同日期下的日期偏差是否小于或等于θ；若是，则执行S29，若否，则结束抽取过程。
S29、统计低QoS区域内，低QoS数据波动的标准差。
S30、判断低QoS数据波动的标准差是否小于σ；若是，则执行S31，若否，则结束抽取过程。
S31、估计低QoS区域的开始时间。
S32、记录低QoS信息。
在本申请实施例中，上述图4的预设参数门限值中低QoS门限值、低QoS打分占比门限X％和低 QoS打分的持续时间T(分别对应于预设门限值、预设比值和预设时间段)，可以根据场景进行动态调整。其中，低QoS门限值表示QoS数据低于什么门限认定为低QoS数据；低QoS占比门限X％表示在该路段采集的QoS数据中，低QoS打分的占比达到某一个门限才考虑将该路段加入QoS地图中；低QoS打分的持续时间T表示QoS打分持续低于低QoS门限值的时间。上述场景可以包括空间位置，例如，地铁环境和家庭环境对应的上述参数设置可能不同，也可以是同一环境下的不同区域设置不同参数数值，例如，对于家庭环境中书房和厨房对应的上述参数设置可能不同。同时，场景还可以包括不同的应用，例如，游戏、长视频APP、短视频APP对应的上述参数设置可能不同。上述图4中提供的抽取算法具有较高的灵活性，提高了QoS地图构建的准确性，从而提高了利用QoS地图进行QoS预测的性能。
在本申请实施例中，基于上述图4所示的抽取算法，得到低QoS区域、低QoS区域的低QoS持续时长和低QoS区域的低QoS开始时间，如图5所示，图5为本申请实施例提供的一种低QoS区域的示例性的示意图。图5中示出了在地铁场景下，地铁运行在某一站时，所采集的QoS打分数据，采集开始时间在18：00，采集时间段大约在260秒。图5中纵坐标表示QoS打分数据，分值越大说明网络质量越好，在构建QoS地图时，需要关注低QoS打分数据，图5中是以预设门限值是60为例进行说明，将低于60的QoS打分数据作为低QoS数据。图5中横坐标表示时间，将18:00开始采集的数据按照时间对齐，将采集的多个数据按照时间序列分布排列在图5中，图5中上下边界中间是多个数据，图5中曲线表示多个数据的均值，对于每秒，均对应有多个QoS打分数据。在进行数据抽取算法时，先将低QoS数据抽取出来，然后统计每秒上低QoS数据的占比，对于图5中的标注位置1，虽然也有低QoS数据，但是由于其占比较小，因此不予考虑将其加入低QoS区域。图5中还可以看出对于每秒，低QoS数据波动的标准差，也就是图5中上下边界的宽度，若宽度较大，则说明数据波动性大，若宽度较小，则说明数据波动性小，数据稳定，具有可参考性。由于图5中示出的QoS数据是关于某一地铁站的，因此，可以确定低QoS区域，经过上述图4中的数据抽取过程，可以得到低QoS区域的持续时长、低QoS区域的开始时间，图5中以标注位置2和标注位置3表示。
相关技术中，QoS地图的属性维度低，例如，场景、应用等维度没有考虑，预测效果有限。本申请实施例采用数据抽取的统计算法构建QoS地图，该算法中所采用的预设参数门限值可以动态调节，对于不同应用，可以根据需要设置不同的预设参数门限值，提高了QoS地图的准确性和通用性。并且，针对不同应用和不同场景，均可以采用该QoS地图判断是否启动QoS预测，无需针对每种应用和每种场景分别构建QoS地图。通过单次QoS数据采集，构建出适配于不同应用的QoS地图，根据有限的测试构建一个通用的QoS地图，降低QoS地图的构建成本。
在本申请一些实施例中，在上述图1中S103之后，该通信预测方法还可以包括S104和S105。
S104、接收终端设备发送的基于通信质量地图的反馈信息。
S105、根据反馈信息，对通信质量地图进行更新，生成更新后的通信质量地图；更新后的通信质量地图用于终端设备对不同应用在不同场景下进行下一次通信质量预测。
在本申请实施例中，终端设备中预测装置与服务器之间互相通信，预测装置从服务器下载通信质量地图，利用通信质量地图，在特定位置或特定时间段得到即将进入弱信号区域，以及弱信号持续时长的信息，然后终端设备中的预测装置或应用模块根据正在运行应用的状态信息判断是否要进行预加载。终端设备中的预测装置将三方应用的反馈信息发送给终端设备。反馈信息反映了正在运行应用的抗弱信号能力，以及对当前网络质量的反馈。服务器可以根据反馈信息对通信质量地图进行更新，更新后的通信质量地图更适用于某一场景或某一应用，从而提高通信质量预测的准确度。
在本申请一些实施例中，上述S105可以通过以下两种示例实现：
示例一、根据反馈信息，对预设参数门限值进行调整，得到调整后的参数门限值；基于调整后的参数门限值，重新确定更新后的通信质量地图，完成对通信质量地图的更新。
示例二、根据反馈信息，对通信质量地图中与正在运行应用相关的信息进行更新，得到更新后的通信质量地图。
在本申请实施例中，以通信质量地图是QoS地图为例，服务器基于三方应用反馈的反馈信息(由预测装置发送给服务器)进行QoS地图更新，如图6所示，图6为本申请实施例提供的一种基于三方应用的反馈信息的QoS地图更新的示例性的示意图。本申请实施例中更新QoS地图的方法可以包括以下两种方式。一种是根据三方应用的反馈信息调整低QoS信息抽取算法中的参数(可以包括预设门限值、预设比值和预设时间段)，然后重新执行低QoS信息抽取算法生成新版本的QoS地图。另一种方法是根据三方应用的反馈信息直接修改已有低QoS区域的属性，从而对QoS地图进行更新；示例性的，对于某一应用，在特定时间或特定位置，三方应用对于QoS预测服务发送的低QoS告警，总是不采取任何优化策略，说明该低QoS区域对于这个应用没有影响，因此，可以考虑将该低QoS区域从用于该 应用的QoS地图中删除。
需要说明的是，图6中是以数据流向为例进行说明，三方应用的反馈信息是由终端设备的预测装置(也就是QoS预测服务装置)向服务器发送的。然后服务器通过低QoS区域的修改，或者调整低QoS信息抽取算法中的参数，执行低QoS信息抽取算法，从而进行QoS地图版本发布。将QoS地图部署在服务器的端侧，由QoS预测服务装置从端侧进行下载，进行下一次通信质量预测。
本申请实施例还提供一种通信预测方法，如图7所示，如图7所示，图7为本申请实施例提供的再一种通信预测方法的可选的步骤流程图，该通信预测方法可以由终端设备中的预测装置执行，通信预测方法包括以下步骤：
S301、获取当前位置，以及服务器发送的通信质量地图；通信质量地图用于提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息。
在本申请实施例中，终端设备在移动过程中，可以根据GPS定位实时获取当前位置。终端设备可以从服务器上下载通信质量地图。
在本申请实施例中，通信质量地图可以是上述服务器根据多个通信质量数据采取抽取算法，构建出的通用QoS地图，通用QoS地图用于使得终端设备在特定位置(及特定时间段)时，获取前方路段的低QoS数据，进而基于应用的卡顿特征和通用QoS地图判断是否执行QoS预测。通信质量地图也可以是基于其他的方式构建的普通地图，通过对普通地图所包括的信息进行进一步的运算，构普通地图中的低QoS数据，只要能够使得终端设备在特定位置(及特定时间段)时，获取前方路段的低QoS数据即可。
在本申请一些实施例中，上述S301中可以通过以下两个示例实现：
示例一、通过定位获取当前位置；发送全量地图下载请求至服务器，得到服务器针对全量地图下载请求发送的通信质量地图。
示例二、发送增量地图下载请求至服务器，得到服务器针对增量地图下载请求发送的增量通信质量地图；基于增量通信质量地图，对获取的当前通信质量地图进行更新，得到通信质量地图。
在本申请实施例中，以通信质量地图是QoS地图为例，将QoS地图部署在服务器，也就是将下载服务部署在服务器。终端设备通过调用服务器开放的API接口进行地图的下载。其中，QoS地图的下载可以是针对区域或城市的全量下载，也可以是针对位置、时间、运营商等变量的增量下载。
在本申请实施例中，通过终端设备不同的需求进行地图下载，提高了下载方式的多样性，针对增量下载的方式，减少了资源消耗。
S302、根据通信质量地图和当前位置，确定通信预测策略。
S303、向终端设备中的应用模块发送通信预测策略，供应用模块根据通信预测策略对正在运行应用进行通信预测。
在本申请实施例中，根据通信质量地图和当前位置，可以判断出终端设备是否进入弱信号区域，当判断出终端设备进入弱信号区域时，由终端设备的预测装置确定通信预测策略，通信预测策略可以是弱信号区域告警，也可以是指示正在运行应用需要进行预加载的策略。预测装置向终端设备中的应用模块发送通信预测策略，使得应用模块对正在运行应用进行通信预测。若通信预测策略是弱信号区域告警，则应用模块根据正在运行应用的自身抗弱信号能力，判断是否进行预加载；若预测策略是指示进行预加载的策略，则应用模块根据策略对正在运行应用进行预加载。
可以理解的是，根据本申请实施例提供的方案，获取当前位置，以及服务器发送的通用的通信质量地图；通信质量地图用于提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息；根据通信质量地图和当前位置，可以在特定位置或特定时间段，获取前方弱信号区域、弱信号持续时长、弱信号开始时间等，进而根据上述与弱信号相关的信息确定通信预测策略。向终端设备中的应用模块发送通信预测策略，供应用模块根据通信预测策略对正在运行应用进行通信预测，通信预测结果为对正在运行应用执行预加载或对正在运行应用不执行预加载，通信预测结果与正在运行应用的情况相匹配，避免因触发预测流程带来更多的视频数据预加载，减少了额外的流量消耗、额外的功率消耗和内存占用。
在本申请一些实施例中，上述图7中S302可以包括S3021-S3023。如图8所示，图8为本申请实施例提供的又一种通信预测方法的可选的步骤流程图，
S3021、根据当前位置和通信质量地图，确定预测区域；其中，预测区域为即将进入的目标区域。
S3022、根据通信质量地图，确定预测区域的边界位置、进入预测区域的起始位置、进入预测区域的起始时刻，以及处于预测区域的预测持续时长。
S3023、若当前位置与起始位置之间的距离，小于或等于边界位置与起始位置之间的距离，或者，到达时刻在起始时刻之后，则根据预测持续时长，确定通信预测策略；其中，到达时刻表征从当前位置 到达起始位置的时间。
在本申请实施例中，目标区域是弱信号区域，预测区域为当前位置周围的终端设备有可能即将进入的弱信号区域。通信质量地图上的目标区域有很多，只有终端设备即将要进入的目标区域才是此次通信预测方法中需要被关注的弱信号区域，在此以预测区域表示终端设备即将要进入的弱信号区域。
在本申请实施例中，通信质量地图用于让终端设备在特定位置(对应于边界位置)时，获取前方弱信号区域，预测区域的起始位置是弱信号开始出现的位置。若当前位置超过边界位置，则说明终端设备即将进入弱信号区域。当然，也可以通过时间上的先后判断终端设备是否即将进入弱信号区域。根据当前位置、当前时刻、起始位置和获取的移动速度预测，预测终端设备到达弱信号区域的到达时刻，若到达时刻在起始时刻之后，说明终端设备即将进入弱信号区域。终端设备根据预测持续时长，确定通信预测策略。
在本申请实施例中，通过计算位置或时间之间的关系，确定终端设备是否要进入弱信号区域，提高了判断是否进入弱信号区域的多样性。
在本申请一些实施例中，在上述S3023之前，该通信预测方法还包括S401和S402。
S401、获取正在运行应用的状态信息。
S402、根据正在运行应用的状态信息，确定正在运行应用的触发基准值。
基于上述S401和S402中得到的触发基准值，上述S3023中根据预测持续时长，确定通信预测策略，还可以通过以下方式实现。若预测持续时长大于或等于触发基准值，则生成执行策略；其中，通信预测策略包括执行策略，执行策略用于使得应用模块对正在运行应用进行预缓存。
示例性的，以当前运行应用是视频APP为例，不同的第三方视频APP有其独特的视频预加载/缓存释放算法，其应用本身就决定了该视频APP的基础抗弱信号能力，使得在某些弱信号区域，即使不使用QoS预测服务，也能维持一段时间的视频播放流畅度。例如：某视频APP在网络信号条件正常情况下，持续预加载未来30秒视频数据，因此，当所进入的弱信号区域的时长低于30秒时，即使不使用QoS预测服务，该应用的流畅度不会有太大影响。相反地，若使用QoS预测服务，则可能因触发预测流程带来更多的视频数据预加载，增加额外的流量消耗、额外的功率消耗和内存占用。因此，预测装置不是在检测到终端设备即将进入弱信号区域之后，就直接进行通信预测策略，而是还结合正在运行应用的状态信息，确定是否需要启动通信预测策略。
在本申请实施例中，正在运行应用的状态信息可以包括正在运行应用的当前运行业务类型、画质(分辨率)、传输率和帧率等。根据正在运行应用的状态信息，可以确定正在运行应用的抗弱信号能力，触发基准值表征正在运行应用被触发进行预缓存的最长启动时间。若预测持续时长大于或等于触发基准值，则说明需要触发预测流程，使得应用模块对正在运行应用进行预缓存，以维持当前运行业务类型的流畅度。若预测持续时长小于触发基准值，则说明正在运行应用自身在正常情况下，已经持续预加载未来一段时间内的数据，因此，无需触发预测流程，减少了额外的流量消耗。
可以理解的是，由预测装置获取正在运行应用的状态信息；根据状态信息，确定触发基准值；若预测持续时长大于或等于触发基准值，则生成执行策略；使得应用模块采用执行策略对正在运行应用进行预缓存。通过结合状态信息，确定触发基准值，提高了通信预测策略的实时性。
在本申请一些实施例中，正在运行应用的状态信息包括：正在运行应用的运行配置信息；上述S401还可以通过以下方式实现。获取正在运行应用的历史运行配置信息和正在运行应用的当前运行配置信息；历史运行配置信息为与当前时刻相邻历史时间段内的运行配置信息；根据历史运行配置信息和当前运行配置信息，确定正在运行应用的运行配置信息。
在本申请实施例中，状态信息包括运行配置信息，运行配置信息包括但不限于：分辨率、传输率和帧率等。运行配置信息具有时变特性(例如，短视频中不同影片会有不同的分辨率)。在将这些实时变化的指标作为确定触发基准值、判断是否需要进行预缓存的判断条件时，可根据当前值和过去一段历史值来形成判断条件。
示例性的，将历史运行配置信息和当前运行配置信息取均值，作为运行配置信息，例如，分辨率、传输率和帧率等可以取当前值(对应于当前运行配置信息)和过去5秒内的值(对应于历史运行配置信息)做平均，从而得到运行配置信息。为保证运行配置信息的时效性，历史运行配置信息为与当前时刻相邻历史时间段内的运行配置信息，例如，与当前时刻过去相邻5秒内每秒对应的运行配置信息。
可以理解的是，由于运行配置信息的时变特性，因此。根据历史运行配置信息和当前运行配置信息，确定正在运行应用的运行配置信息，提高了运行配置信息的准确性。
在本申请一些实施例中，正在运行应用的状态信息包括：正在运行应用的当前运行业务类型和正在运行应用的运行配置信息；上述S401还可以通过以下方式实现。根据正在运行应用的运行配置信息，以及预设配置区间与预设等级之间的映射关系，确定目标配置等级；目标配置等级表征运行配置信息所 属的配置等级；根据正在运行应用的当前运行业务类型、目标配置等级，以及预设等级与预设基准值之间的映射关系，确定触发基准值。
在本申请实施例中，上述判断条件中，分辨率、传输率和帧率等是以数值型表示的连续性指标，在根据连续性指标确定触发基准值、判断是否需要进行预缓存的判断条件时，可将这些指标简单划分为若干档次作为判断条件。
在本申请实施例中，预设配置区间与预设等级之间的映射关系表示配置区间与预设等级之间的对应关系，示例性的，以运行配置信息是分辨率为例进行说明，按照分辨率的大小进行区间划分，分辨率区间的数量可以是多个。例如，分辨率区间<720P，其对应的等级为1级，分辨率区间≥720P，其对应的等级为2级。预设等级与预设基准值之间的映射关系表示等级与预设基准值之间的对应关系，示例性的，1级对应的基准值是20秒，2级对应的基准值是30秒。
需要说明的是，运行配置信息的区间划分方式、预设配置区间与预设等级之间的映射关系、预设等级与预设基准值之间的映射关系、预设基准值的大小，均可以由本领域技术人员根据实际需求进行适当设置，并不局限于上述列举的映射关系，对此本申请实施例仅是示例性说明，不做限制。
在本申请实施例中，通过运行配置信息的数值确定目标配置等级，目标等级反映了运行配置信息的播放需求，进而影响触发基准值的大小。状态信息是连续性数值，离散程度高，对其进行等级分类，以降低散列程度。将位于某一区间的状态信息设置为同一级别，降低了需要处理大量数据的资源消耗，提高了数据处理效率。
在本申请一些实施例中，正在运行应用的状态信息包括：正在运行应用的当前运行业务类型；在上述S402之前，该通信预测方法还包括以下步骤。获取多个应用样本的卡顿特征；根据多个应用样本的卡顿特征，确定各个应用样本的初始触发基准值；其中，初始触发基准值表征应用样本在预设运行条件下进入目标区域之后的正常运行时间。
上述S401还可以通过以下方式实现。根据各个应用样本的初始触发基准值，确定正在运行应用的初始触发基准值，或者，根据正在运行应用的当前运行业务类型和各个应用样本的初始触发基准值，确定正在运行应用的初始触发基准值；将正在运行应用的初始触发基准值作为正在运行应用的触发基准值。
在本申请实施例中，应用的卡顿特征用于表征该应用在特定条件下(例如：长视频/短视频、画质、码率、传输带宽、吞吐量等)，一旦进入弱信号区域后，经过多长时间开始显现卡顿状态。在对应用的卡顿特征进行测试时，可以通过强制限制终端设备的上行/下行传输率、关闭移动数据或进入飞行模式等方法模拟弱信号场景。
示例性的，如图9所示，图9为本申请实施例提供的一种应用的卡顿特征的示例性的示意图，应用1的长视频播放从正常网速进入到弱信号区域，例如，限速至1千比特每秒(Kbps)后，经过平均78.3秒开始出现卡顿延迟，应用1的卡顿特征包括但不限于图9中应用1的长视频卡顿延迟(s)。应用2的短视频播放从正常网速进入到弱信号区域，例如，限速至1千比特每秒(Kbps)后，经过平均38.8秒后开始出现卡顿延迟，应用2的卡顿特征包括但不限于图9中应用2的短视频卡顿延迟(s)。其中，kbps指的是数字信号的传输速率，也可以表示网络的传输速度，Avg表示时间平均值。
在本申请实施例中，在获取各个应用样本的卡顿特征之后，对于同一应用类型，将其对应的多个应用样本的卡顿特征中卡顿延迟均值作为该应用类型的初始触发基准值。例如，针对图9中应用1，将多个应用1的卡顿延迟均值作为应用1的初始触发基准值。
在本申请实施例中，对于某些视频APP，其既支持长视频播放，也支持短视频播放，也就是说应用对应的业务类型有多个，对于同一应用的不同的业务类型，其对应的初始触发基准值可以不相同。因此，对于每个应用样本，在获取其卡顿特征时，是获取应用样本运行不同业务类型时的卡顿特征，基于此，在确定初始触发基准值时，根据正在运行应用的当前运行业务类型和各个应用样本的初始触发基准值，确定正在运行应用的初始触发基准值。例如，针对图9中应用2，将多个应用2在播放短视频时的卡顿延迟均值作为应用2在播放短视频业务时的初始触发基准值，将多个应用2在播放长视频时的卡顿延迟均值作为应用2在播放长视频业务时的初始触发基准值。
可以理解的是，本申请实施例提供的方案，利用不同应用的卡顿统计特征，构建出不同应用在不同条件下触发执行QoS预测的准则，终端设备结合通用QoS地图以及该触发执行QoS预测的准则决定是否执行QoS预测服务。QoS预测服务具备适配不同应用的能力，避免触发非必要的预加载和预缓存流程，减少流量、功率消耗以及内存占用，提高了终端设备的整体性能。
下面，将说明本申请实施例在一个实际的应用场景中的示例性应用。
本申请实施例提出一种基于应用的卡顿特征和通用QoS地图判断是否执行QoS预测的方法，增加通用QoS地图的灵活运用和弹性运用。基于上述图2，如图10所示，图10为本申请实施例提供的另一种基于QoS地图构建的通信方法的示例性的流程示意图。上述图2中S14还可以包括S141-S144。
S141、获取通用QoS地图。
示例性的，以当前运行应用是视频APP为例，对于不同应用、不同场景、不同用户来说，均可以采用上述QoS地图进行QoS预测，因此，可以将上述QoS地图称为通用QoS地图。
在本申请实施例中，通用QoS地图可以是图2中S12所构建的QoS地图，也可以是通过其他QoS地图构建方法取得，只要其他QoS地图包括能直接或间接推算出移动终端进入弱信号区域的位置或时间点，以及穿越该弱信号区域所需时间的相关信息即可。
S142、获取不同应用的卡顿特征。
S143、构建不同应用执行QoS预测的准则。
在本申请实施例中，针对不同应用APP配置对应的QoS预测启动基准值，也可结合其他条件(例如，应用的业务类型、画质/分辨率、传输率和帧率等)进行QoS预测启动基准值的配置。QoS预测启动基准值用于判断是否执行QoS预测，即当预测的弱信号时长小于或等于该启动基准值时，不启动QoS预测；反之，若预测的弱信号时长大于该启动基准值，则启动QoS预测。
示例性的，如表1所示，表1为本申请实施例提供的一种QoS预测执行准则，表1中示出了部分应用配置的启动基准值。编号1对应的准则表示APP#A：当所播放业务类型为长视频、分辨率大于或等于720P、传输率大于或等于20Kbps，以及帧率大于或等于30fps时，若终端设备即将进入弱信号区域，且预测弱信号时长超过30秒，则触发QoS预测服务；若该预测时长在30秒以下，则不触发QoS预测服务。编号2对应的准则表示APP#A：当所播放业务类型为短视频、分辨率小于720P、传输率小于20Kbps，以及帧率小于30fps时，若终端设备即将进入弱信号区域，且预测弱信号时长超过20秒，则触发QoS预测服务；若该预测时长在20秒以下，则不触发QoS预测服务。
在本申请实施例中，做为一种可能的配置方式，编号4对应的准则表示在同一应用(APP#A)下，不区分其他条件，若前方弱信号维持时间超过15秒，则触发QoS预测服务。同理，编号3对应的准则代表在同一应用(APP#A)下，只要是播放短视频，不区分其他条件，若前方弱信号维持时间超过18秒，则触发QoS预测服务。对于表1中编号5-编号7的表示含义可以参照上述编号1-4中的描述进行理解，在此不再赘述。
表1
S144、判定是否执行QoS预测。
终端设备中预测装置根据当前定位位置、通用QoS地图、当前运行应用以及QoS预测准则，判定是否执行QoS预测。
在本申请实施例中，根据上述图2中S12构建的通用QoS地图终端设备在特定位置(及特定时间段)时，获取前方路段的低QoS数据，进而基于不同应用执行QoS预测的准则和通用QoS地图判断是否执行QoS预测。若预测装置的判断结果是需要执行QoS预测，则应用模块执行当前运行应用(例如，视频APP)的预加载/缓存释放的策略，提高视频播放的流畅度(即，视频播放卡顿率下降)。
相关技术中，不同的应用有其自身的的视频预加载/缓存释放算法，使得该应用在经过弱信号区域时，在一定程度上能缓解卡顿现象发生，若完全按QoS地图中的预测数据，将触发额外的预加载流程，导致非必要的视频内容预加载。
可以理解的是，本申请实施例在通用的QoS地图基础上，根据不同应用的卡顿特征进行适配，以减少非必要的预加载流程。通过分析不同应用的卡顿统计特征，构建出不同应用执行QoS预测的触发基准值(criteria)。终端根据当前定位结果、当前运行应用、QoS地图以及触发基准值决定是否执行QoS预测。基于应用的卡顿特征和通用QoS地图判断是否执行QoS预测的执行方法，在策略上更有针对性地执行QoS预测，当预先判断用户经过弱信号区域时会发生卡顿才执行QoS预测，若不不发生卡顿则不执行QoS预测，避免因触发预测流程带来更多的视频数据预加载，减少了额外的流量消耗、额外的功率消耗和内存占用。
在本申请一些实施例中，在上述S303之后，该通信预测方法还包括以下步骤。接收应用模块发送的针对执行策略的执行反馈信息；向服务器发送执行反馈信息；其中，执行反馈信息用于对通信质量地图进行更新。
在本申请实施例中，执行策略用于使得应用模块对正在运行应用进行预缓存，应用模块在对正在运行应用运行预缓存之后，还将执行反馈信息发送给预测装置，预测装置可以根据执行反馈信息可以获取到正在运行应用更新后的状态信息，以便根据正在运行应用更新后的状态信息，对正在运行应用进行下一次通信预测策略的预测。通过预测装置将执行反馈信息发送给服务器，使得服务器根据执行反馈信息对通信质量地图进行更新。
在本申请一些实施例中，通信预测策略是携带预测持续时长的通信质量告警信息；通信质量告警信息用于使得应用模块根据正在运行应用的状态信息，判断是否采取对应的执行策略。
在本申请实施例中，通信预测策略是通信质量告警信息，通过预测装置发送给应用模块，告知应用模块即将进入弱信号区域，由应用模块根据自身的状态信息判断是否要进行预加载，判断过程可以参见上述预测装置侧相对应的描述，例如，根据S401和S402得到触发基准值，然后基于预测持续时长与触发基准值之间的关系进行判断，在此不再赘述。
在本申请一些实施例中，通信预测策略是携带预测持续时长的通信质量告警信息；在上述S303之后，该通信预测方法还包括以下步骤。接收应用模块发送的在采取或不采取执行策略之后的应用反馈信息；向服务器发送应用反馈信息；其中，应用反馈信息用于对通信质量地图进行更新。
在本申请实施例中，无论应用模块是否进行预加载，均将其对应的应用反馈信息发送给预测装置，预测装置可以根据应用反馈信息中的执行结果对下一次相同运行应用的告警进行预测。也就是，应用是否执行了预加载，可以作为预测装置在下一次告警时的参考。例如，应用对于此次告警信息没有反应，没有进行预加载，此次告警携带的预测持续时长为20秒，那么在下一次相同运行应用、相同场景下，若预测持续时长小于或等于20秒，则不向应用模块发送告警信息。通过预测装置将应用反馈信息发送给服务器，使得服务器根据应用反馈信息对通信质量地图进行更新。例如，对于某一应用，在特定时间或特定位置，应用模块对于预测装置发送的告警，总是不采取任何优化策略，说明该弱信号区域对于这个应用没有影响，因此，可以考虑将该弱信号区域从用于该应用的通信质量地图中删除。
在本申请实施例中，终端设备中的QoS预测服务(对应于预测装置)与终端设备中的三方应用(对应于应用模块)之间可以通过软件接口等进行连接，从而实现QoS预测服务与三方应用的交互。交互方式包括以下两种，一种是当QoS预测服务根据通信质量地图得知前方将要出现低QoS路段时，QoS预测服务向三方应用发送前方即将进入低QoS路段的告警，由三方应用对其是否要进行预加载进行判断；另一种是当QoS预测服务根据通信质量地图得知前方将要出现低QoS路段时，由QoS预测服务根据正在运行应用的状态信息判断出三方应用是否需要进行预加载，当判断出三方应用需要进行预加载时，QoS预测服务向三方应用发送执行策略。以下结合图11和图12进行说明。
示例性的，如图11所示，图11为本申请实施例提供的一种Qos预测服务与三方应用交互的示例性的示意图，在使用QoS预测服务过程中，当预测装置根据通信质量地图得知前方将要出现低QoS路段(对应于低QoS区域)时，QoS预测服务向三方应用发送前方即将进入低QoS路段的低QoS告警(图11中以发送QoS预测告警表示发送通信质量告警信息)；三方应用收到低QoS告警后，采取对应的应对措施，对应的应对措施为启动预加载，或者，三方应用根据上述QoS预测执行准则进行判断，若判断前方低QoS路段不会影响正在运行应用，则不采取任何操作。三方应用再将采取的对应的应对措施反馈给QoS预测服务(图11中以反馈告警执行结果示出)；三方应用还会将最终的用户体验质量(Quality of Experience，QoE)反馈给QoS预测服务(图11中以反馈应用QoE示出)，QoS预测服务可以从三方应用获得两种反馈，一种是优化策略的反馈(三方应用针对告警的反应)，包括正在运行应用的状态信息；另一种是QoE反馈。这两种反馈可用于对QoS地图的更新过程。应用反馈信息包括告警执行结果和应用QoE。
示例性的，如图12所示，图12为本申请实施例提供的一种Qos预测服务与三方应用交互的示例性的示意图，通过QoS预测服务获取三方应用反馈的正在运行应用的状态信息(图12中以更新应用状态示出)，当正在运行应用的状态信息满足执行QoS预测的准则时，QoS预测服务向三方应用发送QoS执行策略(图12中以发送QoS预测策略表示发送执行策略)。三方应用再将策略执行结果反馈给QoS预测服务(图12中以反馈策略执行结果示出)；三方应用还将QoE反馈给QoS预测服务(图12中以反馈应用QoE示出)，QoS预测服务可以从三方应用获得两种反馈，一种是执行策略的反馈，包括正在运行应用的状态信息；另一种是QoE反馈。执行反馈信息包括策略执行结果和应用QoE。其中，上述图11和图12中三方应用所反馈的正在运行应用的状态信息包括但不限于应用的业务类型、画质(即分辨率)、传输率及帧率等。
为实现本申请实施例服务器侧的通信预测方法，本申请实施例还提供一种服务器，如图13所示，图13为本申请实施例提供的一种服务器的可选的结构示意图，该服务器130包括：第一获取部分1301，被配置为获取多个属性维度上的多个通信质量数据；每个通信质量数据对应多个属性维度下的属性信息；多个属性维度表征不同场景或应用的维度；
第一确定部分1302，被配置为根据多个通信质量数据和预设参数门限值，确定目标通信质量数据和目标通信质量数据对应的目标时间信息；根据目标通信质量数据在多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，通信质量地图被配置为终端设备对不同应用在不同场景下进行通信质量预测。
在一些实施例中，多个属性维度包括以下至少两个：空间位置、时间段、应用类型、运营商和通信制式。
在一些实施例中，第一获取部分1301，还被配置为获取多个属性维度上的多个初始的通信质量数据；基于多个初始的通信质量数据，确定各个初始的通信质量数据对应的多个通信质量数据。
在一些实施例中，第一确定部分1302，还被配置为对多个通信质量数据按照属性信息一致性进行分组，确定至少一组通信质量数据；各组通信质量数据之间的属性信息不同，每组通信质量数据内各个通信质量数据的属性信息一致，且时间上连续；基于预设参数门限值，对各组通信质量数据分别进行抽取，确定各组通信质量数据分别对应的目标通信质量数据和目标通信质量数据对应的目标时间信息。
在一些实施例中，预设参数门限值包括：预设门限值；
第一确定部分1302，还被配置为从每组通信质量数据中，确定小于或等于预设门限值的多个第一数据，并获取各个第一数据对应的多个时间点；若多个第一数据大于或等于预设样本数量，则将多个第一数据确定为目标通信质量数据；基于各个第一数据对应的多个时间点，确定目标时间信息；目标时间信息包括目标信号持续时长和目标信号开始时间。
在一些实施例中，预设参数门限值还包括：预设比值和预设时间段；
第一确定部分1302，还被配置为若多个第一数据大于或等于预设样本数量，则确定多个第一数据占每组通信质量数据的数据总数量的数据占比；从多个第一数据中，抽取数据占比大于或等于预设比值的多个第二数据，并基于各个第一数据对应的多个时间点，获取各个第二数据对应的多个时间点；根据各个第二数据对应的多个时间点，确定各个第二数据对应的各个候选持续时长；在多个第二数据中，将候选持续时长大于或等于预设时间段的多个第三数据，确定为目标通信质量数据，并基于各个第二数据对应的多个时间点，获取各个第三数据对应的多个时间点；基于各个第三数据对应的多个时间点，确定多个第三数据对应的目标信号持续时长和目标信号开始时间；目标时间信息包括目标信号持续时长和目标信号开始时间。
在一些实施例中，预设参数门限值还包括：预设偏差；
第一确定部分1302，还被配置为确定候选持续时长大于或等于预设时间段的多个第三数据之间的目标偏差；在多个第三数据中，确定目标偏差小于或等于预设偏差的多个目标数据；将多个目标数据作为目标通信质量数据，并获取各个目标数据对应的多个时间点；根据各个目标数据对应的多个时间点，确定多个目标数据对应的目标信号持续时长和目标信号开始时间；目标时间信息包括目标信号持续时长和目标信号开始时间。
在一些实施例中，目标偏差包括：第一日期偏差和第一数据偏差；预设偏差包括：预设日期波动差和预设数据波动差；
第一确定部分1302，还被配置为将多个第三数据按照日期进行分组，确定在不同日期下多个第三数据之间的第一日期偏差；在多个第三数据中，确定第一日期偏差小于或等于预设日期波动差的多个第四数据；确定各个第四数据的第一数据偏差；在多个第四数据中，确定第一数据偏差小于或等于预设数据波动差的多个目标数据。
在一些实施例中，目标偏差包括第二日期偏差和第二数据偏差，预设偏差包括：预设日期波动差和预设数据波动差；
第一确定部分1302，还被配置为确定多个第三数据的第二数据偏差；在多个第三数据中，确定第二数据偏差小于或等于预设数据波动差的多个第五数据；将多个第五数据按照日期进行分组，确定在不同日期下多个第五数据之间的第二日期偏差；在多个第五数据中，确定第二日期偏差小于或等于预设日期波动差的多个目标数据。
在一些实施例中，服务器130还包括第一接收部分1303和更新部分1304；
第一接收部分1303，被配置为接收终端设备发送的基于通信质量地图的反馈信息；
更新部分1304，被配置为根据反馈信息，对通信质量地图进行更新，生成更新后的通信质量地图；更新后的通信质量地图被配置为终端设备对不同应用在不同场景下进行下一次通信质量预测。
在一些实施例中，更新部分1304，还被配置为根据反馈信息，对预设参数门限值进行调整，得到调整后的参数门限值；基于调整后的参数门限值，重新确定更新后的通信质量地图，完成对通信质量地图的更新；或者，根据反馈信息，对通信质量地图中与正在运行应用相关的信息进行更新，得到更新后的通信质量地图。
需要说明的是，上述实施例提供的服务器在进行通信预测时，仅以上述各程序模块的划分进行举例说明，实际应用中，可以根据需要而将上述处理分配由不同的程序模块完成，即将装置的内部结构划分成不同的程序模块，以完成以上描述的全部或者部分处理。另外，上述实施例提供的服务器与服务器侧的方法实施例属于同一构思，其具体实现过程及有益效果详见方法实施例，这里不再赘述。对于本装置实施例中未披露的技术细节，请参照本申请方法实施例的描述而理解。
为实现本申请实施例预测装置侧的通信预测方法，本申请实施例还提供一种预测装置，如图14所示，图14为本申请实施例提供的一种预测装置的可选的结构示意图，该预测装置140包括：第二获取部分1401，被配置为获取当前位置，以及服务器发送的通信质量地图；通信质量地图被配置为提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息；
第二确定部分1402，被配置为根据通信质量地图和当前位置，确定通信预测策略；
发送部分1403，被配置为向终端设备中的应用模块发送通信预测策略，供应用模块根据通信预测策略对正在运行应用进行通信预测。
在一些实施例中，第二确定部分1402，还被配置为根据当前位置和通信质量地图，确定预测区域；其中，预测区域为即将进入的目标区域；根据通信质量地图，确定预测区域的边界位置、进入预测区域的起始位置、进入预测区域的起始时刻，以及处于预测区域的预测持续时长；若当前位置与起始位置之间的距离，小于或等于边界位置与起始位置之间的距离，或者，到达时刻在起始时刻之后，则根据预测持续时长，确定通信预测策略；其中，到达时刻表征从当前位置到达起始位置的时间。
在一些实施例中，第二获取部分1401，还被配置为获取正在运行应用的状态信息；
第二确定部分1402，还被配置为根据正在运行应用的状态信息，确定正在运行应用的触发基准值；若预测持续时长大于或等于触发基准值，则生成执行策略；其中，通信预测策略包括执行策略，执行策略被配置为使得应用模块对正在运行应用进行预缓存。
在一些实施例中，正在运行应用的状态信息包括：正在运行应用的运行配置信息；
第二获取部分1401，还被配置为获取正在运行应用的历史运行配置信息和正在运行应用的当前运行配置信息；历史运行配置信息为与当前时刻相邻历史时间段内的运行配置信息；
第二确定部分1402，还被配置为根据历史运行配置信息和当前运行配置信息，确定正在运行应用的运行配置信息。
在一些实施例中，正在运行应用的状态信息包括：正在运行应用的当前运行业务类型和正在运行应用的运行配置信息；
第二确定部分1402，还被配置为根据正在运行应用的运行配置信息，以及预设配置区间与预设等级之间的映射关系，确定目标配置等级；目标配置等级表征运行配置信息所属的配置等级；根据正在运行应用的当前运行业务类型、目标配置等级，以及预设等级与预设基准值之间的映射关系，确定触发基准值。
在一些实施例中，正在运行应用的状态信息包括：正在运行应用的当前运行业务类型；
第二获取部分1401，还被配置为获取多个应用样本的卡顿特征；
第二确定部分1402，还被配置为根据多个应用样本的卡顿特征，确定各个应用样本的初始触发基准值；其中，初始触发基准值表征应用样本在预设运行条件下进入目标区域之后的正常运行时间；根据各个应用样本的初始触发基准值，确定正在运行应用的初始触发基准值；或者，根据正在运行应用的当前运行业务类型和各个应用样本的初始触发基准值，确定正在运行应用的初始触发基准值；将正在运行应用的初始触发基准值作为正在运行应用的触发基准值。
在一些实施例中，预测装置140包括第二接收部分1404；
第二接收部分1404，被配置为接收应用模块发送的针对执行策略的执行反馈信息；
发送部分1403，被配置为向服务器发送执行反馈信息；其中，执行反馈信息被配置为对通信质量地图进行更新。
在一些实施例中，通信预测策略是携带预测持续时长的通信质量告警信息；通信质量告警信息被配置为使得应用模块根据正在运行应用的状态信息，判断是否采取对应的执行策略。
在一些实施例中，第二接收部分1404，被配置为接收应用模块发送的在采取或不采取执行策略之后的应用反馈信息；
发送部分1403，被配置为向服务器发送应用反馈信息；其中，应用反馈信息被配置为对通信质量 地图进行更新。
在一些实施例中，第二获取部分1401，还被配置为通过定位获取当前位置；发送全量地图下载请求至服务器，得到服务器针对全量地图下载请求发送的通信质量地图；或者，发送增量地图下载请求至服务器，得到服务器针对增量地图下载请求发送的增量通信质量地图；基于增量通信质量地图，对获取的当前通信质量地图进行更新，得到通信质量地图。
需要说明的是，上述实施例提供的预测装置与预测装置侧的方法实施例属于同一构思，其具体实现过程及有益效果详见方法实施例，这里不再赘述。对于本装置实施例中未披露的技术细节，请参照本申请方法实施例的描述而理解。
基于上述图14所示的预测装置140，本申请实施例还提供一种终端设备，如图15所示，图15为本申请实施例提供的一种终端设备的可选的结构示意图，该终端设备150包括上述图14中的预测装置140和应用模块1405，预测装置140，被配置为获取当前位置，以及服务器发送的通信质量地图；通信质量地图被配置为提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息；根据通信质量地图和当前位置，确定通信预测策略；向终端设备中的应用模块发送通信预测策略；
应用模块1405，被配置为根据通信预测策略对正在运行应用进行通信预测。
在本申请实施例中，图16为本申请实施例提出的一种服务器组成结构示意图，如图16所示，本申请实施例提出的服务器160包括第一处理器1601、存储可执行计算机程序的第一存储器1602，第一处理器1601，用于执行第一存储器1602中存储的可执行计算机程序时，实现本申请实施例服务器侧提供的通信预测方法。
在一些实施例中，服务器160还可以包括第一通信接口1603，以及用于连接第一处理器1601、第一存储器1602和第一通信接口1603的第一总线1604。在本申请实施例中，第一总线1604用于连接第一通信接口1603、第一处理器1601以及第一存储器1602，实现这些器件之间的相互通信。
在本申请实施例中，图17为本申请实施例提出的一种终端设备组成结构示意图，如图17所示，本申请实施例提出的终端设备170包括第一处理器1701、存储可执行计算机程序的第一存储器1702，第一处理器1701，用于执行第一存储器1702中存储的可执行计算机程序时，实现本申请实施例预测装置侧提供的通信预测方法。
在一些实施例中，终端设备170还可以包括第一通信接口1703，以及用于连接第一处理器1701、第一存储器1702和第一通信接口1703的第一总线1704。在本申请实施例中，第一总线1704用于连接第一通信接口1703、第一处理器1701以及第一存储器1702，实现这些器件之间的相互通信。
在本申请实施例中，上述处理器(第一处理器1601和第一处理器1701)可以为特定用途集成电路(Application Specific Integrated Circuit，ASIC)、数字信号处理器(Digital Signal Processor，DSP)、数字信号处理装置(Digital Signal Processing Device，DSPD)、可编程逻辑装置(ProgRAMmable Logic Device，PLD)、现场可编程门阵列(Field ProgRAMmable Gate Array，FPGA)、中央处理器(Central Processing Unit，CPU)、控制器、微控制器、微处理器中的至少一种。可以理解地，对于不同的设备，用于实现上述处理器功能的电子器件还可以为其它，本申请实施例不作具体限定。
存储器(第一存储器1602和第一存储器1702)用于存储可执行计算机程序和数据，该可执行计算机程序包括计算机操作指令，存储器(第一存储器1602和第一存储器1702)可能包含高速RAM存储器，也可能还包括非易失性存储器，例如，至少两个磁盘存储器。在实际应用中，上述存储器(第一存储器1602和第一存储器1702)可以是易失性存储器(volatile memory)，例如随机存取存储器(Random-Access Memory，RAM)；或者非易失性存储器(non-volatile memory)，例如只读存储器(Read-Only Memory，ROM)，快闪存储器(flash memory)，硬盘(Hard Disk Drive，HDD)或固态硬盘(Solid-State Drive，SSD)；或者上述种类的存储器的组合，并向处理器(第一处理器1601和第一处理器1701)提供可执行计算机程序和数据。
另外，在本实施例中的各功能模块可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能模块的形式实现。
集成的单元如果以软件功能模块的形式实现并非作为独立的产品进行销售或使用时，可以存储在一个计算机可读取存储介质中，基于这样的理解，本实施例的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)或processor(处理器)执行本实施例方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(Read Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
本申请实施例提供一种计算机可读存储介质，存储有计算机程序，用于被处理器执行时实现如上任一实施例所述的通信预测方法。
示例性的，本实施例中的一种通信预测方法对应的程序指令可以被存储在光盘，硬盘，U盘等存储介质上，当存储介质中的与一种通信预测方法对应的程序指令被一电子设备读取或被执行时，可以实现如上述任一实施例所述的通信预测方法。
本领域内的技术人员应明白，本申请实施例可提供为方法、系统、或计算机程序产品。因此，本申请可采用硬件实施例、软件实施例、或结合软件和硬件方面的实施例的形式。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器和光学存储器等)上实施的计算机程序产品的形式。
本申请是参照根据本申请实施例的方法、设备(系统)、和计算机程序产品的实现流程示意图和/或方框图来描述的。应理解可由计算机程序指令实现流程示意图和/或方框图中的每一流程和/或方框、以及实现流程示意图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在实现流程示意图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在实现流程示意图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在实现流程示意图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
以上所述，仅为本申请的较佳实施例而已，并非用于限定本申请的保护范围。
工业实用性
本申请实施例提供了一种通信预测方法、服务器、预测装置、终端设备和存储介质。根据本申请实施例提供的方案，获取多个属性维度上的多个通信质量数据；每个通信质量数据对应多个属性维度下的属性信息；多个属性维度表征不同场景或应用的维度；根据多个通信质量数据和预设参数门限值，确定目标通信质量数据和目标通信质量数据对应的目标时间信息。目标通信质量数据对应的区域表示弱信号区域，根据预设参数门限值将需要关注的弱信号区域筛选出来，弱信号区域携带的相关特性(例如，应用、空间位置、弱信号时间信息等)，可以适配不同应用和不同场景。根据目标通信质量数据在多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。根据有限的测试生成通信质量地图，即可针对不同的应用、不同的场景(例如，地铁出行场景、室内场景、偏远场景、上次班高峰场景等)，采用该通信质量地图判断是否启动通信质量预测，提高了通信质量地图的通用性。

Claims (45)

  1. 一种通信预测方法，应用于服务器，所述方法包括：
    获取多个属性维度上的多个通信质量数据；每个所述通信质量数据对应所述多个属性维度下的属性信息；所述多个属性维度表征不同场景或应用的维度；
    根据多个所述通信质量数据和预设参数门限值，确定目标通信质量数据和所述目标通信质量数据对应的目标时间信息；
    根据所述目标通信质量数据在所述多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，所述通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。
  2. 根据权利要求1所述的方法，其中，
    所述多个属性维度包括以下至少两个：空间位置、时间段、应用类型、运营商和通信制式。
  3. 根据权利要求1所述的方法，其中，所述获取多个属性维度上的多个通信质量数据，包括：
    获取多个属性维度上的多个初始的通信质量数据；
    基于多个所述初始的通信质量数据，确定各个初始的通信质量数据对应的多个所述通信质量数据。
  4. 根据权利要求1-3任一项所述的方法，其中，所述根据多个所述通信质量数据和预设参数门限值，确定目标通信质量数据和所述目标通信质量数据对应的目标时间信息，包括：
    对多个所述通信质量数据按照属性信息一致性进行分组，确定至少一组通信质量数据；各组通信质量数据之间的属性信息不同，每组通信质量数据内各个通信质量数据的属性信息一致，且时间上连续；
    基于所述预设参数门限值，对所述各组通信质量数据分别进行抽取，确定所述各组通信质量数据分别对应的目标通信质量数据和所述目标通信质量数据对应的目标时间信息。
  5. 根据权利要求4所述的方法，其中，所述预设参数门限值包括：预设门限值；
    所述基于所述预设参数门限值，对所述各组通信质量数据分别进行抽取，确定所述各组通信质量数据分别对应的目标通信质量数据和所述目标通信质量数据对应的目标时间信息，包括：
    从所述每组通信质量数据中，确定小于或等于预设门限值的多个第一数据，并获取所述各个第一数据对应的多个时间点；
    若所述多个第一数据大于或等于预设样本数量，则将所述多个第一数据确定为所述目标通信质量数据；
    基于所述各个第一数据对应的多个时间点，确定所述目标时间信息；所述目标时间信息包括目标信号持续时长和目标信号开始时间。
  6. 根据权利要求5所述的方法，其中，所述预设参数门限值还包括：预设比值和预设时间段；
    所述从所述每组通信质量数据中，确定小于或等于预设门限值的多个第一数据，并获取所述各个第一数据对应的多个时间点之后，所述方法还包括：
    若所述多个第一数据大于或等于所述预设样本数量，则确定所述多个第一数据占所述每组通信质量数据的数据总数量的数据占比；
    从所述多个第一数据中，抽取所述数据占比大于或等于所述预设比值的多个第二数据，并基于所述各个第一数据对应的多个时间点，获取所述各个第二数据对应的多个时间点；
    根据所述各个第二数据对应的多个时间点，确定所述各个第二数据对应的各个候选持续时长；
    在所述多个第二数据中，将候选持续时长大于或等于所述预设时间段的多个第三数据，确定为所述目标通信质量数据，并基于所述各个第二数据对应的多个时间点，获取各个第三数据对应的多个时间点；
    基于所述各个第三数据对应的多个时间点，确定所述多个第三数据对应的目标信号持续时长和目标信号开始时间；所述目标时间信息包括所述目标信号持续时长和所述目标信号开始时间。
  7. 根据权利要求6所述的方法，其中，所述预设参数门限值还包括：预设偏差；
    所述根据所述各个第二数据对应的多个时间点，确定所述各个第二数据对应的各个候选持续时长之后，所述方法还包括：
    确定候选持续时长大于或等于所述预设时间段的多个第三数据之间的目标偏差；
    在所述多个第三数据中，确定所述目标偏差小于或等于所述预设偏差的多个目标数据；
    将所述多个目标数据作为所述目标通信质量数据，并获取所述各个目标数据对应的多个时间点；
    根据所述各个目标数据对应的多个时间点，确定所述多个目标数据对应的目标信号持续时长和目标信号开始时间；所述目标时间信息包括所述目标信号持续时长和目标信号开始时间。
  8. 根据权利要求7所述的方法，其中，所述目标偏差包括：第一日期偏差和第一数据偏差；所述预设偏差包括：预设日期波动差和预设数据波动差；
    所述在所述多个第三数据中，确定所述目标偏差小于或等于所述预设偏差的多个目标数据，包括：
    将所述多个第三数据按照日期进行分组，确定在不同日期下所述多个第三数据之间的所述第一日期偏差；
    在所述多个第三数据中，确定所述第一日期偏差小于或等于所述预设日期波动差的多个第四数据；
    确定所述各个第四数据的所述第一数据偏差；
    在所述多个第四数据中，确定所述第一数据偏差小于或等于所述预设数据波动差的所述多个目标数据。
  9. 根据权利要求7所述的方法，其中，所述目标偏差包括第二日期偏差和第二数据偏差，所述预设偏差包括：预设日期波动差和预设数据波动差；
    所述在所述多个第三数据中，确定所述目标偏差小于或等于所述预设偏差的多个目标数据，包括：
    确定所述多个第三数据的所述第二数据偏差；
    在所述多个第三数据中，确定所述第二数据偏差小于或等于所述预设数据波动差的多个第五数据；
    将所述多个第五数据按照日期进行分组，确定在不同日期下所述多个第五数据之间的所述第二日期偏差；
    在所述多个第五数据中，确定所述第二日期偏差小于或等于所述预设日期波动差的所述多个目标数据。
  10. 根据权利要求1-3任一项所述的方法，其中，所述根据所述目标通信质量数据在所述多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图之后，所述方法还包括：
    接收所述终端设备发送的基于所述通信质量地图的反馈信息；
    根据所述反馈信息，对所述通信质量地图进行更新，生成更新后的通信质量地图；所述更新后的通信质量地图用于所述终端设备对不同应用在不同场景下进行下一次通信质量预测。
  11. 根据权利要求10所述的方法，其中，所述根据所述反馈信息，对所述通信质量地图进行更新，生成更新后的通信质量地图，包括：
    根据所述反馈信息，对所述预设参数门限值进行调整，得到调整后的参数门限值；
    基于所述调整后的参数门限值，重新确定更新后的通信质量地图，完成对所述通信质量地图的更新；
    或者，
    根据所述反馈信息，对所述通信质量地图中与正在运行应用相关的信息进行更新，得到所述更新后的通信质量地图。
  12. 一种通信预测方法，应用于终端设备中的预测装置，所述方法包括：
    获取当前位置，以及服务器发送的通信质量地图；所述通信质量地图用于提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息；
    根据所述通信质量地图和所述当前位置，确定通信预测策略；
    向终端设备中的应用模块发送所述通信预测策略，供所述应用模块根据所述通信预测策略对正在运行应用进行通信预测。
  13. 根据权利要求12所述的方法，其中，所述根据所述通信质量地图和所述当前位置，确定通信预测策略，包括：
    根据所述当前位置和所述通信质量地图，确定预测区域；其中，所述预测区域为即将进入的目标区域；
    根据所述通信质量地图，确定所述预测区域的边界位置、进入所述预测区域的起始位置、进入所述预测区域的起始时刻，以及处于所述预测区域的预测持续时长；
    若所述当前位置与所述起始位置之间的距离，小于或等于所述边界位置与所述起始位置之间的距离，或者，到达时刻在所述起始时刻之后，则根据所述预测持续时长，确定所述通信预测策略；
    其中，所述到达时刻表征从所述当前位置到达所述起始位置的时间。
  14. 根据权利要求13所述的方法，其中，所述方法还包括：
    获取正在运行应用的状态信息；
    根据所述正在运行应用的状态信息，确定正在运行应用的触发基准值；
    所述根据所述预测持续时长，确定所述通信预测策略，包括：
    若所述预测持续时长大于或等于所述触发基准值，则生成执行策略；其中，所述通信预测策略包括所述执行策略，所述执行策略用于使得所述应用模块对所述正在运行应用进行预缓存。
  15. 根据权利要求14所述的方法，其中，所述正在运行应用的状态信息包括：所述正在运行应用的运行配置信息；
    所述获取正在运行应用的状态信息，包括：
    获取所述正在运行应用的历史运行配置信息和所述正在运行应用的当前运行配置信息；所述历史运行配置信息为与当前时刻相邻历史时间段内的运行配置信息；
    根据所述历史运行配置信息和所述当前运行配置信息，确定所述正在运行应用的运行配置信息。
  16. 根据权利要求14或15所述的方法，其中，所述正在运行应用的状态信息包括：所述正在运行应用的当前运行业务类型和所述正在运行应用的运行配置信息；
    所述根据所述正在运行应用的状态信息，确定正在运行应用的触发基准值，包括：
    根据所述正在运行应用的运行配置信息，以及预设配置区间与预设等级之间的映射关系，确定目标配置等级；所述目标配置等级表征所述运行配置信息所属的配置等级；
    根据所述正在运行应用的当前运行业务类型、所述目标配置等级，以及预设等级与预设基准值之间的映射关系，确定所述触发基准值。
  17. 根据权利要求14或15所述的方法，其中，所述正在运行应用的状态信息包括：所述正在运行应用的当前运行业务类型；
    所述根据所述正在运行应用的状态信息，确定正在运行应用的触发基准值之前，所述方法还包括：
    获取多个应用样本的卡顿特征；
    根据所述多个应用样本的卡顿特征，确定各个应用样本的初始触发基准值；其中，所述初始触发基准值表征所述应用样本在预设运行条件下进入目标区域之后的正常运行时间；
    所述根据所述正在运行应用的状态信息，确定正在运行应用的触发基准值，包括：
    根据所述各个应用样本的初始触发基准值，确定所述正在运行应用的初始触发基准值；或者，根据所述正在运行应用的当前运行业务类型和所述各个应用样本的初始触发基准值，确定所述正在运行应用的初始触发基准值；
    将所述正在运行应用的初始触发基准值作为所述正在运行应用的触发基准值。
  18. 根据权利要求14所述的方法，其中，所述向终端设备中的应用模块发送所述通信预测策略之后，所述方法还包括：
    接收所述应用模块发送的针对所述执行策略的执行反馈信息；
    向所述服务器发送所述执行反馈信息；其中，所述执行反馈信息用于对所述通信质量地图进行更新。
  19. 根据权利要求13所述的方法，其中，
    所述通信预测策略是携带预测持续时长的通信质量告警信息；
    所述通信质量告警信息用于使得所述应用模块根据正在运行应用的状态信息，判断是否采取对应的执行策略。
  20. 根据权利要求19所述的方法，其中，所述向终端设备中的应用模块发送所述通信预测策略之后，所述方法还包括：
    接收所述应用模块发送的在采取或不采取所述执行策略之后的应用反馈信息；
    向所述服务器发送所述应用反馈信息；其中，所述应用反馈信息用于对所述通信质量地图进行更新。
  21. 根据权利要求12-15任一项所述的方法，其中，所述获取当前位置，以及服务器发送的通信质量地图，包括：
    通过定位获取所述当前位置；
    发送全量地图下载请求至所述服务器，得到所述服务器针对所述全量地图下载请求发送的所述通信质量地图；
    或者，
    发送增量地图下载请求至所述服务器，得到所述服务器针对所述增量地图下载请求发送的增量通信质量地图；基于所述增量通信质量地图，对获取的当前通信质量地图进行更新，得到所述通信质量地图。
  22. 一种服务器，所述服务器包括：
    第一获取部分，被配置为获取多个属性维度上的多个通信质量数据；每个所述通信质量数据对应所述多个属性维度下的属性信息；所述多个属性维度表征不同场景或应用的维度；
    第一确定部分，被配置为根据多个所述通信质量数据和预设参数门限值，确定目标通信质量数据和所述目标通信质量数据对应的目标时间信息；根据所述目标通信质量数据在所述多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图；其中，所述通信质量地图用于终端设备对不同应用在不同场景下进行通信质量预测。
  23. 根据权利要求22所述的服务器，其中，所述第一获取部分，还被配置为获取多个属性维度上的多个初始的通信质量数据；
    所述第一确定部分，还被配置为基于多个所述初始的通信质量数据，确定各个初始的通信质量数据对应的多个所述通信质量数据。
  24. 根据权利要求22或23所述的服务器，其中，所述第一确定部分，还被配置为对多个所述通信质量数据按照属性信息一致性进行分组，确定至少一组通信质量数据；各组通信质量数据之间的属性信息不同，每组通信质量数据内各个通信质量数据的属性信息一致，且时间上连续；基于所述预设参数门限值，对所述各组通信质量数据分别进行抽取，确定所述各组通信质量数据分别对应的目标通信质量数据和所述目标通信质量数据对应的目标时间信息。
  25. 根据权利要求24所述的服务器，其中，所述预设参数门限值包括：预设门限值；
    所述第一确定部分，还被配置为从所述每组通信质量数据中，确定小于或等于预设门限值的多个第一数据，并获取所述各个第一数据对应的多个时间点；若所述多个第一数据大于或等于预设样本数量，则将所述多个第一数据确定为所述目标通信质量数据；基于所述各个第一数据对应的多个时间点，确定所述目标时间信息；所述目标时间信息包括目标信号持续时长和目标信号开始时间。
  26. 根据权利要求25所述的服务器，其中，所述预设参数门限值还包括：预设比值和预设时间段；
    所述第一确定部分，还被配置为从所述每组通信质量数据中，确定小于或等于预设门限值的多个第一数据，并获取所述各个第一数据对应的多个时间点之后，若所述多个第一数据大于或等于所述预设样本数量，则确定所述多个第一数据占所述每组通信质量数据的数据总数量的数据占比；从所述多个第一数据中，抽取所述数据占比大于或等于所述预设比值的多个第二数据，并基于所述各个第一数据对应的多个时间点，获取所述各个第二数据对应的多个时间点；根据所述各个第二数据对应的多个时间点，确定所述各个第二数据对应的各个候选持续时长；
    所述第一获取部分，还被配置为在所述多个第二数据中，将候选持续时长大于或等于所述预设时间段的多个第三数据，确定为所述目标通信质量数据，并基于所述各个第二数据对应的多个时间点，获取各个第三数据对应的多个时间点；
    所述第一确定部分，还被配置为基于所述各个第三数据对应的多个时间点，确定所述多个第三数据对应的目标信号持续时长和目标信号开始时间；所述目标时间信息包括所述目标信号持续时长和所述目标信号开始时间。
  27. 根据权利要求26所述的服务器，其中，所述预设参数门限值还包括：预设偏差；
    所述第一确定部分，还被配置为根据所述各个第二数据对应的多个时间点，确定所述各个第二数据对应的各个候选持续时长之后，确定候选持续时长大于或等于所述预设时间段的多个第三数据之间的目标偏差；在所述多个第三数据中，确定所述目标偏差小于或等于所述预设偏差的多个目标数据；
    所述第一获取部分，还被配置为将所述多个目标数据作为所述目标通信质量数据，并获取所述各个目标数据对应的多个时间点；
    所述第一确定部分，还被配置为根据所述各个目标数据对应的多个时间点，确定所述多个目标数据对应的目标信号持续时长和目标信号开始时间；所述目标时间信息包括所述目标信号持续时长和目标信号开始时间。
  28. 根据权利要求27所述的服务器，其中，所述目标偏差包括：第一日期偏差和第一数据偏差；所述预设偏差包括：预设日期波动差和预设数据波动差；
    所述第一确定部分，还被配置为将所述多个第三数据按照日期进行分组，确定在不同日期下所述多个第三数据之间的所述第一日期偏差；在所述多个第三数据中，确定所述第一日期偏差小于或等于所述预设日期波动差的多个第四数据；确定所述各个第四数据的所述第一数据偏差；在所述多个第四数据中，确定所述第一数据偏差小于或等于所述预设数据波动差的所述多个目标数据。
  29. 根据权利要求27所述的服务器，其中，所述目标偏差包括第二日期偏差和第二数据偏差，所述预设偏差包括：预设日期波动差和预设数据波动差；
    所述第一确定部分，还被配置为确定所述多个第三数据的所述第二数据偏差；在所述多个第三数据中，确定所述第二数据偏差小于或等于所述预设数据波动差的多个第五数据；
    将所述多个第五数据按照日期进行分组，确定在不同日期下所述多个第五数据之间的所述第二日期偏差；在所述多个第五数据中，确定所述第二日期偏差小于或等于所述预设日期波动差的所述多个目标数据。
  30. 根据权利要求22或23所述的服务器，其中，所述服务器还包括第一接收部分和更新部分；
    所述第一接收部分，被配置为根据所述目标通信质量数据在所述多个属性维度上的目标属性信息，以及目标时间信息，确定通信质量地图之后，接收所述终端设备发送的基于所述通信质量地图的反馈信息；
    所述更新部分，被配置为根据所述反馈信息，对所述通信质量地图进行更新，生成更新后的通信质量地图；所述更新后的通信质量地图用于所述终端设备对不同应用在不同场景下进行下一次通信质量预测。
  31. 根据权利要求30所述的服务器，其中，所述第一获取部分，还被配置为根据所述反馈信息，对所述预设参数门限值进行调整，得到调整后的参数门限值；
    所述第一确定部分，还被配置为基于所述调整后的参数门限值，重新确定更新后的通信质量地图，完成对所述通信质量地图的更新；
    或者，
    所述第一获取部分，还被配置为根据所述反馈信息，对所述通信质量地图中与正在运行应用相关的信息进行更新，得到所述更新后的通信质量地图。
  32. 一种预测装置，所述预测装置包括：
    第二获取部分，被配置为获取当前位置，以及服务器发送的通信质量地图；所述通信质量地图用于提供在低于预设通信质量的情况下，不同应用在不同场景下对应的属性信息和时间信息；
    第二确定部分，被配置为根据所述通信质量地图和所述当前位置，确定通信预测策略；
    发送部分，用于向终端设备中的应用模块发送所述通信预测策略，供所述应用模块根据所述通信预测策略对正在运行应用进行通信预测。
  33. 根据权利要求32所述的预测装置，其中，所述第二确定部分，还被配置为根据所述当前位置和所述通信质量地图，确定预测区域；其中，所述预测区域为即将进入的目标区域；根据所述通信质量地图，确定所述预测区域的边界位置、进入所述预测区域的起始位置、进入所述预测区域的起始时刻，以及处于所述预测区域的预测持续时长；若所述当前位置与所述起始位置之间的距离，小于或等于所述边界位置与所述起始位置之间的距离，或者，到达时刻在所述起始时刻之后，则根据所述预测持续时长，确定所述通信预测策略；其中，所述到达时刻表征从所述当前位置到达所述起始位置的时间。
  34. 根据权利要求33所述的预测装置，其中，所述第二获取部分，还被配置为获取正在运行应用的状态信息；
    所述第二确定部分，还被配置为根据所述正在运行应用的状态信息，确定正在运行应用的触发基准值；若所述预测持续时长大于或等于所述触发基准值，则生成执行策略；其中，所述通信预测策略包括所述执行策略，所述执行策略用于使得所述应用模块对所述正在运行应用进行预缓存。
  35. 根据权利要求34所述的预测装置，其中，所述正在运行应用的状态信息包括：所述正在运行应用的运行配置信息；
    所述第二获取部分，还被配置为获取所述正在运行应用的历史运行配置信息和所述正在运行应用的当前运行配置信息；所述历史运行配置信息为与当前时刻相邻历史时间段内的运行配置信息；
    所述第二确定部分，还被配置为根据所述历史运行配置信息和所述当前运行配置信息，确定所述正在运行应用的运行配置信息。
  36. 根据权利要求34或35所述的预测装置，其中，所述正在运行应用的状态信息包括：所述正在运行应用的当前运行业务类型和所述正在运行应用的运行配置信息；
    所述第二确定部分，还被配置为根据所述正在运行应用的运行配置信息，以及预设配置区间与预设等级之间的映射关系，确定目标配置等级；所述目标配置等级表征所述运行配置信息所属的配置等级；根据所述正在运行应用的当前运行业务类型、所述目标配置等级，以及预设等级与预设基准值之间的映射关系，确定所述触发基准值。
  37. 根据权利要求34或35所述的预测装置，其中，所述正在运行应用的状态信息包括：所述正在运行应用的当前运行业务类型；
    所述第二获取部分，还被配置为根据所述正在运行应用的状态信息，确定正在运行应用的触发基准值之前，获取多个应用样本的卡顿特征；
    所述第二确定部分，还被配置为根据所述多个应用样本的卡顿特征，确定各个应用样本的初始触发基准值；其中，所述初始触发基准值表征所述应用样本在预设运行条件下进入目标区域之后的正常运行时间；根据所述各个应用样本的初始触发基准值，确定所述正在运行应用的初始触发基准值；或者，根据所述正在运行应用的当前运行业务类型和所述各个应用样本的初始触发基准值，确定所述正在运行应用的初始触发基准值；将所述正在运行应用的初始触发基准值作为所述正在运行应用的触发基准值。
  38. 根据权利要求34所述的预测装置，其中，所述预测装置还包括第二接收部分；
    所述第二接收部分，被配置为向终端设备中的应用模块发送所述通信预测策略之后，接收所述应用模块发送的针对所述执行策略的执行反馈信息；
    发送部分，被配置为向所述服务器发送所述执行反馈信息；其中，所述执行反馈信息用于对所述通信质量地图进行更新。
  39. 根据权利要求33所述的预测装置，其中，
    所述通信预测策略是携带预测持续时长的通信质量告警信息；
    所述通信质量告警信息用于使得所述应用模块根据正在运行应用的状态信息，判断是否采取对应的执行策略。
  40. 根据权利要求39所述的预测装置，其中，所述第二接收部分，被配置为向终端设备中的应用模块发送所述通信预测策略之后，接收所述应用模块发送的在采取或不采取所述执行策略之后的应用反馈信息；
    所述发送部分，还被配置为向所述服务器发送所述应用反馈信息；其中，所述应用反馈信息用于对所述通信质量地图进行更新。
  41. 根据权利要求32-35任一项所述的预测装置，其中，所述第二获取部分，还被配置为通过定位获取所述当前位置；
    所述发送部分，还被配置为发送全量地图下载请求至所述服务器，得到所述服务器针对所述全量地图下载请求发送的所述通信质量地图；
    或者，
    所述发送部分，还被配置为发送增量地图下载请求至所述服务器，得到所述服务器针对所述增量地图下载请求发送的增量通信质量地图；基于所述增量通信质量地图，对获取的当前通信质量地图进行更新，得到所述通信质量地图。
  42. 一种终端设备，其中，终端设备包括如权利要求32所述的预测装置和应用模块；
    所述应用模块，被配置为根据所述通信预测策略对正在运行应用进行通信预测。
  43. 一种服务器，所述服务器包括：
    第一存储器，用于存储可执行计算机程序；
    第一处理器，用于执行所述第一存储器中存储的可执行计算机程序时，实现权利要求1-11任一项所述的方法。
  44. 一种终端设备，所述终端设备包括：
    第二存储器，用于存储可执行计算机程序；
    第二处理器，用于执行所述第二存储器中存储的可执行计算机程序时，实现权利要求12-21任一项所述的方法。
  45. 一种计算机可读存储介质，存储有计算机程序，用于被处理器执行时，实现权利要求1-11或12-21任一项所述的方法。
PCT/CN2023/099477 2022-08-30 2023-06-09 通信预测方法、服务器、预测装置、终端设备和存储介质 WO2024045761A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211060668.6A CN117676655A (zh) 2022-08-30 2022-08-30 通信预测方法、服务器、预测装置、终端设备和存储介质
CN202211060668.6 2022-08-30

Publications (1)

Publication Number Publication Date
WO2024045761A1 true WO2024045761A1 (zh) 2024-03-07

Family

ID=90079513

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/099477 WO2024045761A1 (zh) 2022-08-30 2023-06-09 通信预测方法、服务器、预测装置、终端设备和存储介质

Country Status (2)

Country Link
CN (1) CN117676655A (zh)
WO (1) WO2024045761A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100128698A1 (en) * 2008-11-21 2010-05-27 Kentaro Ishizu Communication Terminal Device, Communication System and Method of Selecting Base Station Thereof
US20140200038A1 (en) * 2013-01-16 2014-07-17 Apple Inc. Location Assisted Service Capability Monitoring
WO2021196117A1 (en) * 2020-04-02 2021-10-07 Qualcomm Incorporated Preloading application data based on network status prediction and data access behavior prediction
WO2021250925A1 (ja) * 2020-06-12 2021-12-16 パナソニック株式会社 無線通信装置およびスケジューリング方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100128698A1 (en) * 2008-11-21 2010-05-27 Kentaro Ishizu Communication Terminal Device, Communication System and Method of Selecting Base Station Thereof
US20140200038A1 (en) * 2013-01-16 2014-07-17 Apple Inc. Location Assisted Service Capability Monitoring
WO2021196117A1 (en) * 2020-04-02 2021-10-07 Qualcomm Incorporated Preloading application data based on network status prediction and data access behavior prediction
WO2021250925A1 (ja) * 2020-06-12 2021-12-16 パナソニック株式会社 無線通信装置およびスケジューリング方法

Also Published As

Publication number Publication date
CN117676655A (zh) 2024-03-08

Similar Documents

Publication Publication Date Title
US11902092B2 (en) Systems and methods for latency-aware edge computing
Zhao et al. Mobility prediction-assisted over-the-top edge prefetching for hierarchical VANETs
US10212551B2 (en) Location-based buffer management systems and methods
CN108009688B (zh) 聚集事件预测方法、装置及设备
CN107734350B (zh) 直播系统及直播方法
US20200358673A1 (en) Service management method and related device
CN109992427B (zh) Dpi关联规则回填处理方法、装置、设备及介质
CN114174767A (zh) 使用众包网络数据进行路线规划
WO2016150494A1 (en) Methods and apparatus for evaluating communication network resource along a navigational route
CN109389833A (zh) 一种车辆拥堵预警方法和装置
US11974193B2 (en) Data processing method and apparatus, server, and computer-readable storage medium
JP2020102838A (ja) 車両マクロクラウドにおける改善された無線通信
WO2019164518A1 (en) Method and system for automated dynamic network slice deployment using artificial intelligence
US9154984B1 (en) System and method for estimating network performance
CN111787256B (zh) 警前录像的管理方法、装置、介质及电子设备
CN111629336B (zh) 一种目标区域内人数确定方法、装置、设备及存储介质
WO2024045761A1 (zh) 通信预测方法、服务器、预测装置、终端设备和存储介质
Liotou et al. Enriching HTTP adaptive streaming with context awareness: A tunnel case study
EP3413595B1 (en) Processing method, device and system for service optimization
CN115208955A (zh) 一种资源请求处理的方法、装置、计算机设备及介质
JP6737286B2 (ja) 通信予測装置および通信予測方法、並びにコンピュータ・プログラムが格納された記録媒体
KR20140061629A (ko) 클라우드 서버 모니터링 시스템 및 방법
CN115333917A (zh) 一种cdn异常检测方法及装置
CN107147694B (zh) 一种信息处理方法和装置
US20230144248A1 (en) Dynamic quality of service traffic steering in a multi-access edge computing environment

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23858800

Country of ref document: EP

Kind code of ref document: A1