Methods, devices, equipment, media, and products for cloud phone service quality scheduling

By analyzing the click interaction behavior of cloud phone applications, using a click interaction recognition model to identify application categories and configure differentiated QoS parameters, the problems of resource waste and poor user experience in cloud phone service quality scheduling are solved, achieving more efficient resource utilization and a stable user experience.

CN121644355BActive Publication Date: 2026-11-17CHINA MOBILE INTERNET CO LTD +1
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511652850.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-11-12
Publication Date
2026-11-17
Estimated Expiration
2045-11-12

AI Technical Summary

Technical Problem

Existing cloud phone service quality scheduling strategies cannot be configured differently according to the network resource requirements of different applications, resulting in resource waste and poor user experience.

Method used

By analyzing user click interaction behavior, the click interaction recognition model automatically identifies the category of cloud phone foreground applications, and performs differentiated QoS parameter configuration and dynamic adjustment based on the category, including video encoding parameters and jitter buffer settings.

Benefits of technology

It improved resource utilization efficiency, enhanced the user experience in different application scenarios, and ensured service quality stability when the network changes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121644355B_ABST
    Figure CN121644355B_ABST
Patent Text Reader

Abstract

This disclosure relates to a method, apparatus, device, medium, and product for scheduling QoS (Quality of Service) on cloud phones in the field of cloud computing technology. The method applied in the cloud includes: inputting real-time user click interaction behavior data into a click interaction recognition model, outputting the click interaction category of the cloud phone's foreground application; and adjusting QoS parameter configuration based on the click interaction category of the cloud phone's foreground application. By analyzing the user's click interaction behavior with the cloud phone based on the click interaction recognition model, the click interaction category of the cloud phone's foreground application is automatically identified, and differentiated QoS parameter configuration and dynamic adjustment are performed based on the application category. This differentiated scheduling strategy can better meet the needs of applications with different click interaction categories in the cloud phone scenario, improving resource utilization efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This disclosure relates to the field of cloud computing technology, and in particular to a method, apparatus, device, medium and product for scheduling cloud mobile phone service quality. Background Technology

[0002] With the development of cloud computing and virtualization technologies, cloud phone services have gradually become an emerging mobile application solution. A cloud phone is a mobile phone that migrates local storage, computing, and rendering to the cloud, and its operating system and various applications also run in the cloud. It can be controlled through multiple platforms such as applications (APPs), HTML5 (Hypertext Markup Language), and mini-programs. Users can install and run various applications on cloud phones, including games, video playback, news, etc., and quickly access these cloud resources on their mobile devices.

[0003] Current technology uses the same network transmission parameter settings for real-time footage from different apps running on cloud phones, employing a unified Quality of Service (QoS) scheduling strategy, such as resolution, frame rate, bitrate, and jitter buffer size. However, because different app categories have significantly different network resource requirements, a unified QoS configuration not only wastes resources but also results in a poor user experience. Summary of the Invention

[0004] To address the aforementioned technical issues, this disclosure provides a method, apparatus, device, medium, and product for cloud mobile phone service quality scheduling, achieving differentiated cloud mobile phone service quality scheduling.

[0005] A first aspect of this disclosure provides a method for scheduling cloud mobile phone service quality, applied in the cloud, the method comprising: Input real-time user click interaction data into the click interaction recognition model, and output the click interaction category of the cloud phone front-end application; Adjust the service quality parameter configuration based on the click interaction category of the cloud phone front-end application.

[0006] A second aspect of this disclosure provides a method for cloud mobile phone service quality scheduling, applied to a terminal, the method comprising: Receive information from the cloud that includes the click interaction categories of the cloud phone's foreground application; Set the initial jitter buffer parameters according to the click interaction category of the cloud phone foreground application.

[0007] A third aspect of this disclosure provides an apparatus for scheduling cloud mobile phone service quality, applied in the cloud, the apparatus comprising: The recognition module is configured to input real-time user click interaction behavior data into the click interaction recognition model and output the click interaction category of the cloud phone front-end application. The adjustment module is configured to adjust the service quality parameter configuration based on the click interaction category of the cloud phone front-end application.

[0008] A fourth aspect of this disclosure provides an apparatus for cloud mobile phone service quality scheduling, applied to a terminal, the apparatus comprising: The receiving module is configured to receive information sent from the cloud that includes the click interaction categories of the cloud phone's foreground application; The settings module is configured to set initial jitter buffer parameters based on the click interaction category of the cloud phone foreground application.

[0009] A fifth aspect of this disclosure provides an electronic device, including: At least one processor; Memory for storing the at least one processor-executable instruction; The at least one processor is used to execute the instructions to implement the above-described method.

[0010] A sixth aspect of this disclosure provides a computer-readable storage medium that, when instructions in the computer-readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform the methods described above.

[0011] A seventh aspect of this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the cloud mobile phone service quality scheduling method described above.

[0012] The above-mentioned at least one technical solution adopted in the embodiments of this disclosure can achieve the following beneficial effects: The embodiments of this disclosure analyze the click interaction behavior of users and cloud phones based on the click interaction recognition model, automatically identify the click interaction category of the cloud phone's foreground application, and perform differentiated QoS parameter configuration and dynamic adjustment based on the click interaction category of the application. This differentiated scheduling strategy can better meet the needs of applications with different click interaction categories in the cloud phone scenario and improve resource utilization efficiency. Attached Figure Description

[0013] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0014] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A schematic flowchart illustrating a method for scheduling cloud mobile phone service quality in the cloud, provided as an embodiment of this disclosure; Figure 2 A flowchart illustrating another method for scheduling cloud phone service quality in the cloud, provided as an embodiment of this disclosure; Figure 3 A flowchart illustrating a method for scheduling cloud mobile phone service quality applied to a terminal, provided in an embodiment of this disclosure; Figure 4 A schematic diagram of a device for scheduling cloud mobile phone service quality in the cloud, provided as an embodiment of this disclosure; Figure 5 A schematic diagram of a device for scheduling cloud mobile phone service quality in a terminal, provided in an embodiment of this disclosure; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure; Figure 7 This is a schematic diagram of the structure of an exemplary computer system provided in an embodiment of the present disclosure. Detailed Implementation

[0016] To better understand the above-mentioned objectives, features, and advantages of this disclosure, the solutions disclosed herein will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0017] Numerous specific details are set forth in the following description in order to provide a full understanding of this disclosure, but this disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some, and not all, of the embodiments of this disclosure.

[0018] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0019] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0020] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0021] Before explaining this disclosure, the terms used in this disclosure will be explained for ease of understanding.

[0022] Frame rate refers to the number of consecutive still images (i.e., "frames") played in a video stream per unit of time. Its core function is to determine the smoothness of the video. The unit is usually "frames per second" (fps).

[0023] Existing QoS scheduling strategies for cloud phones have certain limitations. On one hand, applications with different click-based interaction types have significantly different network resource requirements, which may lead to resource waste under a uniform QoS configuration. For example, running games on a cloud phone, a high-frequency interaction application, requires a high frame rate and low latency to ensure a smooth user experience, while playing videos on a cloud phone, a low-frequency interaction application, focuses on image quality and continuous playback. Under a uniform QoS configuration, high-frequency interaction applications may experience increased end-to-end latency due to excessive jitter buffering, while low-frequency interaction applications may waste bandwidth due to excessively high frame rates. On the other hand, because there is no optimization for the characteristics of different applications, the user experience is poor when using different applications. For example, in weak network scenarios, playing games on a cloud phone may result in noticeable latency and stuttering.

[0024] To address the aforementioned problems with existing QoS scheduling schemes, this disclosure proposes a cloud phone QoS scheduling method based on application identification. By analyzing the click interaction behavior of users and cloud phones, the method automatically identifies the click interaction categories of applications and performs differentiated QoS parameter configuration and dynamic adjustment based on these categories.

[0025] The following is combined Figures 1-7 This disclosure describes the methods, apparatus, equipment, media, and products for scheduling cloud mobile phone service quality provided in the embodiments.

[0026] Figure 1 This is a schematic flowchart illustrating a method for scheduling cloud phone service quality in the cloud, as provided in an embodiment of this disclosure. Figure 1 As shown, a method for scheduling cloud phone service quality includes: S101. Input the user's real-time click interaction behavior data into the click interaction recognition model, and output the click interaction category of the cloud phone front-end application; Click interaction data is collected based on how users operate the cloud phone, mainly including the frequency of user clicks, swipe length, and time.

[0027] A cloud phone foreground application refers to an application that is active on the cloud phone interface and is being directly operated and viewed by the user. In contrast to background applications, foreground applications occupy the main display area of ​​the cloud phone, receive user input, and display the results in real time.

[0028] Cloud phones, as a new type of application, are an extension and expansion of physical mobile phones. Due to their powerful cloud computing capabilities and perpetual online nature, users primarily use cloud phones for activities such as playing cloud games, watching videos, and browsing the web. To perform QoS scheduling for different applications, it is necessary to identify the click interaction categories of the applications.

[0029] S102. Adjust the service quality parameter configuration according to the click interaction category of the cloud phone front-end application.

[0030] Click interaction categories are categorized based on the degree of click interaction between the user and the application running on the cloud phone. These categories can include N subcategories, for example, the following three: high-frequency interaction; medium-frequency interaction; and low-frequency interaction. These three categories encompass typical scenarios of applications running on the cloud phone. For example, games are considered high-frequency interactions, where users frequently click on the cloud phone app's interface, generally requiring low latency and high frame rates; short video or news websites are considered medium-frequency interactions, where users swipe the interface to turn pages or refresh videos, generally requiring moderate latency and clarity; and video playback applications are considered low-frequency interactions, where users rarely continue interacting with the interface after clicking to play the video, generally requiring high clarity and tolerating a certain level of latency.

[0031] This embodiment of the disclosure analyzes the click interaction behavior of users and cloud phones based on a click interaction recognition model, automatically identifies the click interaction categories of the cloud phone's foreground applications, and performs differentiated QoS parameter configuration and dynamic adjustment based on the click interaction categories of the applications. This differentiated scheduling strategy can better meet the needs of applications with different click interaction categories in the cloud phone scenario and improve resource utilization efficiency.

[0032] Figure 2 A flowchart illustrating another method for scheduling cloud phone service quality in the cloud, as provided in this disclosure embodiment, is shown below. Figure 2 As shown, in some embodiments, the method for scheduling cloud phone service quality includes: S211. Input the user's real-time click interaction behavior data into the click interaction recognition model, and output the predicted probability of the cloud mobile phone front-end application corresponding to different click interaction categories.

[0033] In actual operation, real-time user click interaction data is collected, input into the trained click interaction recognition model, and the predicted probability of the click interaction behavior within that time period belonging to different click interaction categories is output.

[0034] In this embodiment, tagged user interaction behavior data is collected. Touch event objects are obtained by overriding the `onTouchEvent()` method in the Activity or View of the cloud phone app. Specifically, `getAction()` is used to obtain the action type of the touch event, such as `ACTION_DOWN`, `ACTION_MOVE`, `ACTION_UP`, etc. `getX()` and `getY()` are used to obtain the position coordinates of the touch event. The sliding distance can be calculated by calculating the coordinate difference between two touch events. The sliding speed can be calculated by calculating the time difference and coordinate difference between two touch events. `getPointerCount()` is used to obtain the number of pointers involved in the touch event.

[0035] It is understood that the above-mentioned click interaction recognition model is pre-trained; therefore, in some embodiments, before step S211, the cloud phone service quality scheduling method further includes: S210. Use the user's historical click interaction behavior data as training data to train the click interaction recognition model.

[0036] First, collect historical click interaction data of users in the cloud phone. Based on the way users operate the cloud phone, the main data collected include the frequency of user clicks, swipe length, and time. Based on this labeled click interaction data, a click interaction behavior model is trained.

[0037] In some embodiments, the click interaction recognition model is trained using a neural network model, then the click interaction recognition model includes: Input layer: Contains 4 nodes; Hidden layer: Contains 2 layers, with ReLU as the activation function; Output layer: Contains N nodes, corresponding to N predicted probabilities for N types of click interactions; The normalized exponential function is used as the activation function to output a probability distribution.

[0038] For example, for labeled training samples in a cloud phone, the input is: X = {number of clicks within a period, number of drags within a period, number of swipes within a period, and swipe distance within a period}, and the output is: Y = {probability of high-frequency interaction class, probability of medium-frequency interaction class, and probability of low-frequency interaction class}. A neural network model is constructed that can accurately predict the probability of the application controlled or viewed by the user belonging to different click interaction categories based on the user's click interaction behavior.

[0039] The aforementioned neural network model adjusts its hyperparameters based on training data, such as the number of layers, the number of nodes in the hidden layers, and the activation function. Specifically, the input layer, related to the model's input parameters, has 4 nodes; the hidden layer uses a 2-layer network with a Rectified Linear Unit (ReLU) as the activation function; the output layer has 3 nodes, corresponding to the three click interaction categories of Y in the output parameters, and uses the Softmax activation function to transform the output into a probability distribution, ensuring that the sum of the probabilities of each category is 1; the cross-entropy loss function is used during training.

[0040] It should be noted that when building a click interaction recognition model, not only neural network models can be used, but also other multi-classifiers, such as decision trees and random forests.

[0041] When training the click interaction recognition model, in order to ensure the consistency of the data dimensions input to the model within a unit time period, the training data is selected as follows: X = {number of clicks within the period, number of drags within the period, number of swipes within the period, and swipe distance within the period}, Y = {probability of high-frequency interaction class, probability of medium-frequency interaction class, and probability of low-frequency interaction class}. Among them, the number of clicks within the period is obtained by detecting the number of ACTION_UP events in the touch events. The number of drags and swipes within the period is mainly determined by comparing the absolute values ​​of the horizontal and vertical distances when scrolling is detected, to determine whether it is a drag (horizontal scrolling) or a swipe (vertical scrolling). The swipe distance within the period can also be obtained by getting the position coordinates of the touch event using getX() and getY(). By calculating the coordinate difference between two touch events, the swipe distance can be calculated.

[0042] When the click interaction recognition model is trained, the corresponding model parameters are obtained and stored in the application.

[0043] When users actually use cloud phones, their click interaction data is collected in real time. The model input parameters are periodically constructed, and the period can be determined based on actual usage needs to balance computing resource requirements and application demands. For example, data is collected every 4 seconds, X = {number of clicks, number of drags, number of swipes, and swipe distance within the period}, and input into the click interaction recognition model. The model output Y = {probability of high-frequency interaction type, probability of medium-frequency interaction type, and probability of low-frequency interaction type}, thus obtaining the predicted probability of belonging to each click interaction category.

[0044] S212. Match the cloud phone front-end application with the application library to obtain the preset probability of the cloud phone front-end application corresponding to different click interaction categories.

[0045] The cloud phone's foreground application is matched with the application library to find the preset probabilities of different click interaction categories for the applications stored in the application library.

[0046] The application library is pre-built and stores information including the application's package name and the preset probability of the application corresponding to different click interaction categories.

[0047] The above method also includes: S201, establishing an application library.

[0048] The application library includes the package name information of the applications and the preset probability of the applications corresponding to different click interaction categories.

[0049] Step S201, establishing the application library includes: S2011. Based on the user's click interaction data of the application, calculate the preset probability that the application belongs to different click interaction categories; An application library is established by analyzing mainstream applications in cloud phones. The preset probabilities of belonging to different click interaction categories are statistically analyzed based on actual user operating habits. For a given application, the probability of it belonging to N click interaction categories is calculated according to different user habits. For example, if the click interaction categories are: {high-frequency interaction category, medium-frequency interaction category, low-frequency interaction category}, then the probabilities of belonging to these three click interaction categories are calculated as: {probability of high-frequency interaction category, probability of medium-frequency interaction category, probability of low-frequency interaction category}.

[0050] S2012. Store the application's package name information and the preset probability of the application belonging to different click interaction categories in the application library.

[0051] A record in the application library named {packageName, p1, p2, p3} indicates that the application's package name is packageName, and the preset probabilities of belonging to the {high-frequency interaction class, medium-frequency interaction class, and low-frequency interaction class} are p1, p2, and p3, respectively, where p1 + p2 + p3 = 1. For example, a record in the application library named {package_game, 0.8, 0.15, 0.05} indicates that the game with the package name package_game has preset probabilities of belonging to the {high-frequency interaction class, medium-frequency interaction class, and low-frequency interaction class} of 0.8, 0.15, and 0.05, respectively.

[0052] The above step S212 includes: S2121, matching the obtained package name information of the cloud phone front-end application with the application library; In this embodiment, system-level permissions can be obtained on the cloud device. The package name of the currently foreground application on the cloud phone is obtained periodically using the UsageStatsManager in the Android operating system. Specifically, the UsageStatsManager is used to query the application's usage over a recent period, i.e., a List. <usagestats>appList= usageStatsManager.queryUsageStats(UsageStatsManager.INTERVAL_DAILY, time - First, sort by time and retrieve the package name of the most recently used foreground application, i.e., `String currentApp = recentStats.getPackageName()`. This method retrieves the package name of the foreground application currently running on the cloud phone.

[0053] Once the package name information of the cloud phone's foreground application is obtained, it can be matched with the package names stored in the application library.

[0054] S2122. If the package name information is matched in the application library, output the preset probability of the cloud phone front-end application corresponding to different click interaction categories; if the package name information is not matched in the application library, output the probability of the cloud phone front-end application corresponding to different click interaction categories as 1 / N, where there are N click interaction categories and N is a natural number.

[0055] For example, when N=3, there are three click interaction categories: {high-frequency interaction, medium-frequency interaction, and low-frequency interaction}. The package name information of the cloud phone foreground application is matched with the package names stored in the application library. If an application is matched, the preset probabilities p1, p2, and p3 corresponding to the {high-frequency interaction, medium-frequency interaction, and low-frequency interaction} are output. If no application is matched, the default application is used, and the default membership probabilities 1 / 3, 1 / 3, and 1 / 3 are output.

[0056] S213. Determine the click interaction category of the cloud phone front-end application based on the predicted probability corresponding to different click interaction categories and the preset probability corresponding to different click interaction categories.

[0057] Step S213 includes: S2131, there are N types of click interaction categories, the predicted probability of the Nth click interaction category calculated by the click interaction recognition model is weighted and averaged with the preset probability of the Nth click interaction category matched in the application library to obtain the probability of the corresponding Nth click interaction category, where N is a natural number; Based on the recognition results from the aforementioned application library and the click interaction recognition model, a weighted average is used to obtain the final recognition result. ; in, This represents the probability of belonging to the nth click interaction category. This represents the weight of the k-th model. This represents the probability that the click interaction belongs to the nth click interaction category in the kth model.

[0058] For example, there are three types of click interaction categories: high-frequency interaction, medium-frequency interaction, and low-frequency interaction. The predicted probability output by the click interaction recognition model is {0.7, 0.25, 0.05}, and the preset probability obtained from the application library is {0.8, 0.15, 0.05}, which is equivalent to having 2 models, i.e., k=2.

[0059] The probability P1 of belonging to the high-frequency interaction class is then P1 = w 1×0.7+ w 2×0.8; The probability P2 of belonging to the mid-frequency interaction class is then P2 = w 1×0.25+ w 2×0.15; The probability P3 of belonging to the low-frequency interaction class is then P3 = w 1×0.05+ w 2×0.05; in, w 1 represents the weight of the click interaction recognition model; w 2 represents the weight of the application library.

[0060] S2132. Based on the N probabilities belonging to N click interaction categories calculated above, the click interaction category corresponding to the maximum probability value is determined as the click interaction category of the cloud phone front-end application.

[0061] Based on step S2131 As a result, by using the softmax method, the highest probability is selected as the final recognition result, that is, the click interaction category corresponding to the highest probability value is determined as the click interaction category of the cloud phone front-end application.

[0062] It should be noted that in the click interaction recognition model, a neural network model of click interaction behavior is trained using supervised learning based on collected data such as the frequency, swipe length, and time of user click interactions with labels. During actual operation, real-time user click interaction behavior data is collected and input into the click interaction recognition model. The model outputs the probability that the click interaction behavior within that time period belongs to one of three application types: {high-frequency interaction type, medium-frequency interaction type, and low-frequency interaction type}. Finally, based on a weighted average, the results of the two recognition methods are combined to obtain the final click interaction category of the application. In this embodiment, the application library and click interaction recognition model can be managed through the cloud phone's resource management platform, facilitating batch updates of the cloud phone's application library and click interaction recognition model.

[0063] In this embodiment, by using an application matching library established in the cloud phone and analyzing the user's click interaction behavior with the cloud phone, the click interaction categories of applications are automatically identified, and differentiated QoS parameters are configured and dynamically adjusted based on these categories. According to the identified click interaction categories, video encoding parameters and receiver jitter buffering parameters are automatically and differentiated. For example, high-frequency interaction applications prioritize low latency and high frame rates, while low-frequency interaction applications prioritize image quality and stability. This differentiated scheduling strategy can better meet the needs of applications with different click interaction categories in the cloud phone scenario, improving resource utilization efficiency.

[0064] In some embodiments, in order to ensure that the transmitted video data stream does not exceed the capacity limit of the network link, it is necessary to detect the network bandwidth in real time. Otherwise, network packet loss, network congestion, and other issues will occur, ultimately causing problems such as video stuttering on the terminal.

[0065] Then, step S102 includes: adjusting the transmission parameters of the transmitted video to a preset parameter threshold according to the click interaction category of the cloud phone front-end application and the capacity of the network link.

[0066] In some embodiments, a congestion control-based bandwidth probing algorithm is used, specifically Real-time Transport Protocol (RTP) packets sent by the cloud to transmit video streams. Each video packet contains information such as a timestamp and sequence number. The terminal receives the RTP packets and calculates the corresponding packet loss rate and reception latency gradient, feeding this information back to the cloud via Real-time Transport Control Protocol (RTCP) feedback packets. Based on the packet loss rate and latency information, the cloud estimates the available network bandwidth and controls the transmission of video packets by adjusting the congestion window.

[0067] For example, when the click interaction category is a high-frequency interaction category, the resolution is reduced first, and then the frame rate is reduced to a first frame rate threshold; when the click interaction category is a low-frequency interaction category, the frame rate is reduced first to a second frame rate threshold, and if the bandwidth continues to decrease while reducing the frame rate, the resolution is reduced; when the click interaction category is a medium-frequency interaction category, if the bandwidth continues to decrease, the frame rate is reduced first to a third frame rate threshold, and if the bandwidth continues to decrease, the resolution is reduced.

[0068] Among them, the first frame rate threshold > the third frame rate threshold > the second frame rate threshold.

[0069] Based on the click interaction categories of the cloud phone's front-end application and the capacity of the network link, the cloud encoding strategy is automatically adjusted. That is, the frame rate, resolution, bit rate, etc. of video encoding are adjusted according to the bandwidth detection results and the video resource requirements of the click interaction categories.

[0070] Before proceeding, initial encoding parameters need to be configured. Subsequently, the encoding parameters can be automatically adjusted based on the click interaction categories of the cloud phone's front-end application and the capacity of the network link.

[0071] In the initial encoding parameter configuration, the bandwidth requirements of mainstream applications in cloud phone usage scenarios are considered, and the configuration is based on the video encoding capabilities provided by the cloud phone server and the click interaction category of the application. For example, refer to Table 1 to set the initial parameters.

[0072] Table 1

[0073] The encoding parameters are automatically adjusted based on the click interaction categories of the cloud phone's foreground application and the capacity of the network link.

[0074] In this embodiment, for interactive applications with different frequencies, the encoding parameters are automatically adjusted when the detected network bandwidth changes. This ensures that the transmitted parameters adapt to network changes in a timely manner, guaranteeing the service quality of the application. The adjustment method considers both network changes and controls the frequency of parameter adjustments, avoiding instability in video quality caused by frequent adjustments to encoding parameters.

[0075] For example, high-frequency interactive applications prioritize maintaining low latency and a stable interactive experience. When network changes occur, such as a detected decrease in network bandwidth, the amount of video data is reduced to lower latency, sacrificing image quality to ensure response time. Based on the bandwidth detection results, if a bandwidth decrease occurs within four consecutive 1-second detection cycles, the encoding parameters are adjusted. First, the resolution is reduced from 1080p to 720p, and then the frame rate is reduced. This is because high-frequency interactive applications have high frame rate requirements; for example, in gaming scenarios, frame rate is crucial to the gaming experience. Therefore, in high-frequency interactive application scenarios, the frame rate is reduced to a minimum of 30fps.

[0076] For applications with low-frequency interactions, maintaining image quality should be prioritized, such as online video playback, which requires a certain level of clarity but can tolerate some latency. In this case, the frame rate should be reduced first; for click-based interaction scenarios, the minimum frame rate limit is 18fps. If the detected bandwidth continues to decrease while reducing the frame rate, the resolution should be further reduced.

[0077] For mid-frequency interactive applications, a trade-off is struck between the two: when the detected bandwidth continuously decreases, the frame rate is reduced first, with a minimum frame rate of 20fps; if the detected bandwidth continues to decrease, the resolution is then reduced.

[0078] When setting the encoding bitrate, the bitrate is linearly adjusted based on network bandwidth detection results; that is, if the bandwidth is halved, the bitrate is halved accordingly. Simultaneously, when the bitrate decreases, the QP (Quantization Parameter) value is appropriately increased to further reduce the encoded output bitrate. This embodiment further utilizes real-time tracking of network bandwidth to dynamically adjust encoding parameters, ensuring the encoding bitrate is always synchronized with the network bandwidth, thus balancing image quality and transmission stability.

[0079] In this embodiment, based on the bandwidth detection results and the video resource requirements of the application type, the encoding parameters are automatically adjusted when the detected network bandwidth changes for interactive applications of different frequencies. This ensures that the transmitted parameters adapt to network changes in a timely manner, guaranteeing the service quality of the application. Furthermore, because the network status can be monitored in real time, QoS parameters are adjusted in real time when network conditions fluctuate or the user's click interaction type on the cloud phone changes, ensuring the best user experience in various network environments.

[0080] In some embodiments, the method for scheduling cloud phone service quality further includes: S214. Send the click interaction category of the cloud phone front-end application to the terminal.

[0081] The cloud transmits the encoded video data to the terminal via a data channel; The cloud also uses the application parameter interaction channel to send the application recognition results, such as the click interaction category of the cloud phone's front-end application and video encoding parameter information, to the terminal. Sending video streams: Sending video packets to the terminal via the data channel.

[0082] First, the encoded video frames are divided into packets based on their size. Each video packet is marked with information such as the sequence number and the frame number to which it belongs.

[0083] In order to adapt to changes in the network and prevent bursts of encoded output from overwhelming the network and causing congestion and packet loss during the transmission of video packets, a smooth transmission strategy is adopted. That is, every 5ms, the number of video packets that can be sent is calculated based on the estimated bandwidth value, and then smoothly transmitted to the network.

[0084] Click-interaction category information interaction: Its main function is to send the application identification results—the click-interaction category of the cloud phone front-end application and video encoding parameter information—from the cloud to the terminal, so that the cloud and the terminal can share information.

[0085] In this embodiment, an application parameter interaction channel is established between the terminal and the cloud to update the click interaction category and video parameters. This parameter interaction channel can be implemented via WebSocket, HTTP, or other methods to ensure that application parameter information can be accurately communicated between the terminal and the cloud.

[0086] In this embodiment, the defined click interaction category information parameter format comprises two parts: a message header and a message body. The message header includes {message type, sender ID, and sending timestamp}. The message type identifies the message type, such as a click interaction category change notification or video encoding parameter update. The sender ID identifies the sender, and the sending timestamp identifies the message sending time. The message body includes {click interaction category, video encoding parameters, etc.}. The click interaction category uses an 8-bit enumeration type to represent the click interaction category, such as a high-frequency interaction category. The video encoding parameters include parameters such as resolution, frame rate, bitrate, and quantization parameter (QP) values. When a change in the click interaction category in the network is detected, the relevant click interaction category information is assembled and sent to the terminal through the application parameter interaction channel.

[0087] Figure 3 This is a flowchart illustrating a method for scheduling cloud mobile phone service quality applied to a terminal, as provided in an embodiment of this disclosure. Figure 3 As shown, a method for scheduling cloud mobile phone service quality (QoS) applied to terminals includes: S301. Receive information sent from the cloud containing the click interaction categories of the cloud phone front-end application; The terminal receives the video stream: It assembles the received video packets into frames and stores them in a jitter buffer. The terminal assembles frames based on the frame number and sequence number of each video packet. If a video packet in a frame is lost in the network, the terminal triggers the packet loss retransmission logic, causing the cloud to resend the lost video packet. After receiving a video frame, the terminal can assemble the frames and extract the acquisition timestamp information for subsequent network tracking.

[0088] The terminal also receives information from the cloud containing the click interaction categories of the cloud phone's foreground application.

[0089] S302. Set the initial jitter buffer parameters according to the click interaction category of the cloud phone foreground application.

[0090] The terminal receives data from the application parameter interaction channel and parses out the relevant data. Based on the actual parameter content, the jitter buffer parameters of the receiving end are set. For example, different initial values ​​for the jitter buffer length are set for different click interaction categories, as shown in Table 2.

[0091] Table 2

[0092] In some embodiments, the terminal calculates network jitter: The terminal tracks the jitter of the downlink network in order to objectively and accurately measure the jitter in the network and avoid the impact of packet loss and retransmission.

[0093] In this embodiment, network jitter is measured from the perspective of received video frames. This method is more in line with the adjustment logic of the terminal's jitter buffer. After removing some outliers, a time smoothing method is used to characterize the network jitter. The specific process is as follows: 1) The terminal counts the received acquisition time difference samples within a period of T, denoted as RC (ReceiveCapture).

[0094] Each RC sample represents the time difference between the acquisition and reception of a video frame, reflecting the transmission time of that video frame over the network. The period T can be selected from 500ms, 1s, etc.

[0095] 2) Calculate the smoothed sample (LowestRC, MaxJitter) within one period.

[0096] LowestRC represents the minimum RC value within a period. MaxJitter represents the maximum difference between two consecutive RC values ​​within a period, after removing outliers and the larger value other than three times the standard deviation.

[0097] 3) The smoothing values ​​SmoothLowestRC and SmoothJitter are calculated using a single exponential smoothing method. SmoothLowestRC is smoothed using LowestRC within each cycle, and SmoothJitter is smoothed using MaxJitter within each cycle. The smoothed SmoothLowestRC ensures that a stable reception time difference reference is maintained to prevent drastic changes, while MaxJitter measures the range that the jitter buffer can tolerate.

[0098] SmoothLowestRC =alpha×LowestRC+(1-alpha)×Pre_SmoothLowestRC; SmoothJitter= alpha×MaxJitter+(1-alpha)×Pre_SmoothJitter; The alpha parameter can be set to 0.8, and Pre_SmoothLowestRC and Pre_SmoothJitter represent the smoothed SmoothLowestRC and SmoothJitter calculated in the previous cycle, respectively.

[0099] Jitter buffer: The terminal's jitter buffer is a buffer used to handle network jitter, and its main function is to provide a smooth playback experience for video data. This mainly includes setting an initial jitter buffer value based on the type of click interaction, and adjusting it according to network jitter conditions.

[0100] (1) Set the initial network jitter buffer The initial value of the network jitter buffer is set according to the click interaction category and requirements. Different initial jitter buffer values ​​are set for different application scenarios. The terminal parses the click interaction category information received from the cloud, obtains the click interaction category, and sets the initial jitter buffer value for different click interaction categories. At the same time, the initial decoding time difference decodeTimeDiff is calculated. decodeTimeDiff measures whether the video frame can be pushed to the video engine for decoding. If the difference between the current time and the frame acquisition time is greater than decodeTimeDiff, it means that the video frame can leave the jitter buffer and be pushed to the video engine for decoding and rendering.

[0101] The initial decodeTimeDiff calculation principle is as follows: Based on the initial jitter buffer value, after the initial jitter buffer has filled with video frames, the initial decodeTimeDiff is calculated.

[0102] Initial decodeTimeDiff = Current time - timestamp of the first frame captured in the jitter buffer (2) Update the jitter buffer value based on the terminal network tracking results. In this embodiment of the disclosure, the value of the jitter buffer is dynamically adjusted based on the network jitter statistics results tracked by the terminal network, so that it can dynamically adapt to changes in the network.

[0103] The adjustment method is as follows: the decoding time difference decodeTimeDiff is adjusted according to the difference AjustDiff between decodeTimeDiff and network tracking calculation (SmoothLowestRC + SmoothJitter). In order to avoid frame-cutting logic caused by drastic changes in the jitter buffer, which may lead to playback adjustments, this embodiment of the disclosure also sets a threshold for adjusting the jitter buffer each time, so that the jitter buffer can smoothly cope with network changes.

[0104] AjustDiff = decodeTimeDiff - (SmoothLowestRC + SmoothJitter) For each received video frame, an RC sample is updated. When the timing for updating the smoothing values ​​SmoothLowestRC and SmoothJitter is reached, the relevant jitter buffer adjustment logic is triggered. If AjustDiff > 0, the jitter buffer is reduced by AjustValue, i.e., decodeTimeDiff decreases by AjustValue. If AjustDiff < 0, the jitter buffer is increased by AjustValue, i.e., decodeTimeDiff increases by AjustValue.

[0105] That is, AjustValue = min(abs(AjustDiff), Ajust_thredshould), Here, abs(AjustDiff) represents the absolute value of AdjustDiff, and Ajust_thredshould represents the threshold for one adjustment, which can be selected as 3ms, 5ms, etc.

[0106] In this implementation, the terminal tracks network link jitter in real time and adjusts the jitter buffer according to network jitter and application handover interaction type information to provide differentiated QoS services for different application types.

[0107] The above description is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements, optimizations and modifications can be made without departing from the principle of the present invention, and these should also be considered within the scope of protection of the present invention.

[0108] Figure 4 This is a schematic diagram of a device for scheduling cloud mobile phone service quality in the cloud, provided as an embodiment of the present disclosure. Figure 4 As shown, the device 400 includes: The recognition module 401 is configured to input real-time user click interaction behavior data into the click interaction recognition model and output the click interaction category of the cloud phone front-end application. The adjustment module 402 is configured to adjust the service quality parameter configuration according to the click interaction category of the cloud phone front-end application.

[0109] In some embodiments, the identification module 401 is configured to input real-time user click interaction behavior data into the click interaction identification model and output the predicted probability of the cloud phone front-end application corresponding to different click interaction categories.

[0110] In some embodiments, the device 400 further includes: The matching module 403 is configured to match the cloud phone front-end application with the application library to obtain the preset probability of the cloud phone front-end application corresponding to different click interaction categories; The determining module 404 is configured to determine the click interaction category of the cloud phone foreground application based on the predicted probabilities corresponding to different click interaction categories and the preset probabilities corresponding to different click interaction categories. The determining module 404 is further configured to: Since there are N click interaction categories, the predicted probability of the Nth click interaction category calculated by the click interaction recognition model is weighted and averaged with the preset probability of the Nth click interaction category matched in the application library to obtain the probability corresponding to the Nth click interaction category, where N is a natural number; based on the N probabilities belonging to the N click interaction categories obtained above, the click interaction category corresponding to the highest probability value is determined as the click interaction category of the cloud phone foreground application. The click interaction recognition model includes: an input layer containing 4 nodes; a hidden layer containing 2 layers, with ReLU as the activation function; an output layer containing N nodes, corresponding to N predicted probabilities of the N click interaction categories; and a normalized exponential function as the activation function, outputting a probability distribution.

[0111] The device 400 further includes a training module configured to train the click interaction recognition model using historical click interaction behavior data of the user as training data.

[0112] Figure 5 This is a schematic diagram of a device for scheduling cloud mobile phone service quality in a terminal, provided in an embodiment of this disclosure. Figure 5 As shown, the device 500 includes: The receiving module 501 is configured to receive information sent from the cloud that includes the click interaction categories of the cloud phone front-end application; The setting module 502 is configured to set initial jitter buffer parameters based on the click interaction category of the cloud phone foreground application.

[0113] The specific implementation process of the functions and roles of each module in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0114] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure, such as... Figure 6 As shown, this disclosure also provides an electronic device 600, which includes at least one processor 601 and a memory 602 coupled to the processor 601. The memory 602 is used to store at least one processor 601 executable instructions, wherein the at least one processor 601 is used to execute the instructions to implement the steps of the method described above in this disclosure.

[0115] The processor 601 described above can also be called a Central Processing Unit (CPU), which can be an integrated circuit chip with signal processing capabilities. Each step in the method described in this embodiment can be implemented by the integrated logic circuitry in the processor 601 or by software instructions. The processor 601 can be a general-purpose processor, a digital signal processor (DSP), an ASIC, a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method in this embodiment can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can be located in the memory 602, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The processor 601 reads information from the memory 602 and, in conjunction with its hardware, completes the steps of the method described above.

[0116] Figure 7 This is a schematic diagram of an exemplary computer system provided by an embodiment of the present disclosure. Various operations / processes according to embodiments of the present disclosure, implemented via software and / or firmware, can be transmitted from a storage medium or network to a computer system with a dedicated hardware architecture, for example... Figure 7 The computer system 700 shown is equipped with the programs that constitute the software. When various programs are installed, the computer system is able to perform various functions, including those described above.

[0117] Computer system 700 is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0118] like Figure 7 As shown, the computer system 700 includes a computing unit 701, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. The RAM 703 may also store various programs and data required for the operation of the computer system 700. The computing unit 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0119] Multiple components in the computer system 700 are connected to the I / O interface 705, including: an input unit 706, an output unit 707, a storage unit 708, and a communication unit 709. The input unit 706 can be any type of device capable of inputting information into the computer system 700. The input unit 706 can receive input numerical or character information and generate key signal inputs related to user settings and / or function control of the electronic device. The output unit 707 can be any type of device capable of presenting information and may include, but is not limited to, a monitor, speaker, video / audio output terminal, vibrator, and / or printer. The storage unit 708 may include, but is not limited to, a hard disk and an optical disk. The communication unit 709 allows the computer system 700 to exchange information / data with other devices via a network such as the Internet, and may include, but is not limited to, modems, network interface cards, infrared communication devices, wireless communication transceivers, and / or chipsets, such as Bluetooth™ devices, Wi-Fi devices, WiMax devices, cellular communication devices, and / or the like.

[0120] The computing unit 701 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 701 performs the various methods and processes described above. For example, in some embodiments, the methods described in the embodiments of this disclosure can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 708. In some embodiments, part or all of the computer program can be loaded and / or installed on an electronic device via ROM 702 and / or communication unit 709. In some embodiments, the computing unit 701 can be configured to perform the methods described in the embodiments of this disclosure by any other suitable means (e.g., by means of firmware).

[0121] This disclosure provides a computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the methods described in this disclosure.

[0122] Computer-readable storage media can be volatile memory, such as random-access memory (RAM); or non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD); or devices that include one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc.

[0123] It should be noted that the computer-readable storage medium described in this disclosure can be a computer-readable signal medium, a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), or any suitable combination thereof.

[0124] Embodiments of this disclosure provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the cloud mobile phone service quality scheduling method described above.

[0125] In embodiments of this disclosure, computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include, but are not limited to, object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar languages. The program code can be executed entirely on a computer, partially on a computer, as a standalone software package, partially on a computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0126] The modules, components, or units described in the embodiments of this disclosure can be implemented in software or hardware. The names of the modules, components, or units do not necessarily constitute a limitation on the module, component, or unit itself.

[0127] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary hardware logic components that can be used include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), etc.

[0128] It should be noted that, in this document, terms such as "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0129] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.< / usagestats>

Claims

1. A method for scheduling cloud mobile phone service quality, applied in the cloud, characterized in that, The method includes: The user's real-time click interaction data is input into the click interaction recognition model, and the click interaction category of the cloud phone front-end application is output; wherein, the click interaction category includes the following three types: high-frequency interaction category; medium-frequency interaction category; low-frequency interaction category; Receive RTCP feedback and obtain the current network link capacity based on packet loss rate and latency gradient; Adjusting service quality parameter configuration based on the click interaction category of the cloud phone front-end application includes: adjusting the transmission parameters of the transmitted video to a preset parameter threshold based on the click interaction category of the cloud phone front-end application and the capacity of the network link. When the click interaction category of the cloud phone foreground application is a high-frequency interaction, the resolution is first reduced, and then the frame rate is reduced to the first frame rate threshold. When the click interaction category of the cloud phone foreground application is a low-frequency interaction, the frame rate is first reduced to the second frame rate threshold. While reducing the frame rate, if the bandwidth continues to decrease, the resolution is reduced. When the click interaction category of the cloud phone foreground application is a medium-frequency interaction, when the bandwidth continues to decrease, the frame rate is first reduced to the third frame rate threshold. If the bandwidth continues to decrease, the resolution is reduced.

2. The method according to claim 1, characterized in that, The process of inputting real-time user click interaction data into the click interaction recognition model and outputting the click interaction categories of the cloud phone front-end application includes: Input real-time user click interaction data into the click interaction recognition model, and output the predicted probability of the cloud mobile phone front-end application corresponding to different click interaction categories.

3. The method according to claim 2, characterized in that, The method further includes: The cloud phone foreground application is matched with the application library to obtain the preset probability of the cloud phone foreground application corresponding to different click interaction categories; The click interaction category of the cloud phone front-end application is determined based on the predicted probability corresponding to different click interaction categories and the preset probability corresponding to different click interaction categories.

4. The method according to claim 3, characterized in that, Determining the click interaction category of the cloud phone foreground application based on the predicted probability corresponding to different click interaction categories and the preset probability corresponding to different click interaction categories includes: There are N types of click interaction categories. The predicted probability of the Nth click interaction category calculated by the click interaction recognition model is weighted and averaged with the preset probability of the Nth click interaction category matched in the application library to obtain the probability of the corresponding Nth click interaction category, where N is a natural number. Based on the N probabilities belonging to N click interaction categories obtained from the above calculations, the click interaction category corresponding to the highest probability value is determined as the click interaction category of the cloud phone front-end application.

5. The method according to claim 4, characterized in that, The click interaction recognition model includes: Input layer: Contains 4 nodes; Hidden layer: Contains 2 layers, with ReLU as the activation function; Output layer: Contains N nodes, corresponding to N predicted probabilities for N types of click interactions; The normalized exponential function is used as the activation function to output a probability distribution.

6. The method according to claim 1, characterized in that, The method further includes: The click interaction recognition model is trained using the user's historical click interaction behavior data as training data.

7. The method according to claim 3, characterized in that, The step of matching the cloud phone foreground application with the application library to obtain the preset probability of the cloud phone foreground application corresponding to different click interaction categories includes: The package name information of the cloud phone foreground application is matched with the application library; If the package name information is matched in the application library, the preset probability of the cloud phone front-end application corresponding to different click interaction categories is output. If the package name information is not matched in the application library, the probability that the cloud phone front-end application corresponds to different click interaction categories is 1 / N, where there are N click interaction categories and N is a natural number.

8. The method according to claim 7, characterized in that, The method further includes: An application library is established, which includes the package name information of the applications and the preset probability of the applications corresponding to different click interaction categories.

9. The method according to claim 8, characterized in that, The establishment of the application library includes: Based on user click interaction data of the application, the preset probability of the application belonging to different click interaction categories is calculated; Store the application's package name information and the preset probability of the application belonging to different click interaction categories in the application library.

10. The method according to claim 1, characterized in that, The method further includes: Information containing the click interaction categories of the cloud phone foreground application is sent to the terminal.

11. A method for scheduling cloud mobile phone service quality, applied to a terminal, characterized in that, The method includes: Receive information from the cloud that includes the click interaction categories of the cloud phone's foreground application; Set the initial jitter buffer parameters according to the click interaction category of the cloud phone foreground application.

12. A device for scheduling cloud mobile phone service quality, applied in the cloud, characterized in that, The device includes: The recognition module is configured to input real-time user click interaction behavior data into the click interaction recognition model and output the click interaction category of the cloud phone front-end application; wherein, the click interaction category includes the following three types: high-frequency interaction category; medium-frequency interaction category; and low-frequency interaction category. The capacity acquisition module is configured to receive RTCP feedback and obtain the capacity of the current network link based on the packet loss rate and latency gradient. The adjustment module is configured to adjust the service quality parameter configuration according to the click interaction category of the cloud phone front-end application, including: adjusting the transmission parameters of the transmitted video to a preset parameter threshold according to the click interaction category of the cloud phone front-end application and the capacity of the network link; Specifically, when the click interaction category of the cloud phone foreground application is a high-frequency interaction, the resolution is first reduced, and then the frame rate is reduced to a first frame rate threshold; when the click interaction category of the cloud phone foreground application is a low-frequency interaction, the frame rate is first reduced to a second frame rate threshold, and if the bandwidth continues to decrease while reducing the frame rate, the resolution is reduced; when the click interaction category of the cloud phone foreground application is a medium-frequency interaction, if the bandwidth continues to decrease, the frame rate is first reduced to a third frame rate threshold, and if the bandwidth continues to decrease, the resolution is reduced.

13. A device for scheduling cloud mobile phone service quality, applied to a terminal, characterized in that, The device includes: The receiving module is configured to receive information sent from the cloud that includes the click interaction categories of the cloud phone's foreground application; The settings module is configured to set initial jitter buffer parameters based on the click interaction category of the cloud phone foreground application.

14. An electronic device, characterized in that, include: At least one processor; Memory for storing the at least one processor-executable instruction; The at least one processor is configured to execute the instructions to implement the method as described in any one of claims 1-10 or 11.

15. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the method as described in any one of claims 1-10 or 11.

16. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the cloud mobile phone service quality scheduling method as described in any one of claims 1-10 or 11.

Citation Information

Patent Citations

  • User terminal device and method for dynamically regulating size of shake buffer area

    CN102238294A

  • Personalized configuration method for cloud mobile phone and related equipment

    CN120915873A