Real-time video stream transmission method and system for smart phone based on 5G network

Abstract

The invention relates to a real-time video stream transmission method for a smart phone based on a 5G network, and the method comprises the steps: obtaining the basic parameters of a real-time video stream and the network condition of an uploading end, and generating a first uploading strategy; constructing an uploading network-uploading monitoring data set; the upload analysis model analyzes the data set based on a machine self-learning algorithm to generate an upload network influence coefficient, corrects the first upload strategy to generate a second upload strategy, analyzes and processes the video stream data to obtain cache data, and sends the cache data to a video receiving end; acquiring video downloading request data, and generating a first downloading transmission strategy based on the second uploading strategy; and a downloading network-downloading monitoring data set is constructed, the downloading analysis model analyzes the data set to generate a downloading network influence coefficient, a first downloading transmission strategy is generated based on the influence coefficient and the video downloading request data, and a second downloading transmission strategy is generated through correction of a second uploading strategy. The method has the effect of improving the video transmission quality and the user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless image communication, and in particular to a real-time video stream transmission method and system for a smart phone based on a 5G network. Background Art

[0002] In recent years, as an indispensable communication tool in daily life, the use of smartphones has expanded from simple call and text messaging services to various data services and multimedia functions, especially the growing demand for video streaming services. With the commercial deployment of the fifth-generation mobile communication technology (5G network), users' requirements for the quality of real-time video streaming transmission have also increased.

[0003] At present, 5G networks have significant improvements in speed and latency compared to 4G, providing higher data transmission rates and lower latency. Since 5G networks can theoretically support larger data throughput, when a smartphone is in a 5G network state, the existing video streaming server will send data packets with large data content to the smartphone as the receiving end to ensure the quality of the video. However, since the 5G network operates based on a higher frequency band of radio spectrum, it is easily affected by building obstruction and distance attenuation, which reduces the signal transmission quality. As a result, 5G smartphones may encounter stability and continuity problems in some areas when receiving real-time video streaming transmission. At this time, the server still sends data packets with large data content, which reduces the rate at which the smartphone receives video data, resulting in a decrease in the transmission quality of the real-time video watched by the smartphone, thereby reducing the user's viewing experience, so it needs to be improved. Summary of the invention

[0004] In order to improve the transmission quality of real-time video viewed on smartphones, the present application provides a real-time video streaming transmission method and system for smartphones based on a 5G network.

[0005] In the first aspect, the above-mentioned invention objective of the present application is achieved through the following technical solutions: A real-time video streaming transmission method for a smart phone based on a 5G network, the method comprising the steps of: Acquire basic parameters of the real-time video stream and the network status of the upload end, and generate a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; Based on a preset public timeline, the network data and the upload monitoring data of the upload end are recorded, and based on the public timeline, a mapping relationship between the network data of the upload end and the upload monitoring data is formed to construct an upload network-upload monitoring data set, wherein the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization; The preset upload analysis model analyzes the upload network-upload monitoring data set based on a machine self-learning algorithm to generate an upload network influence coefficient, and the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading the video stream; Based on the predicted video upload efficiency, the first upload strategy is modified to generate a second upload strategy, and based on the second upload strategy, predicted cache data is calculated and data upload channel data is adjusted, wherein video stream data is received, and the video stream data uploaded to the server is parsed to obtain cache data cached in the server, and the cache data is used to send to the video receiving end; Acquire video download request data, and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of the receiving ends; Acquire the download network data and download monitoring data of the receiving end, and form a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct a download network-download monitoring data set; The preset download analysis model analyzes the download network-download monitoring data set based on a machine self-learning algorithm to generate a download network impact coefficient, and modifies the first download transmission strategy based on the download network impact coefficient to generate a second download transmission strategy.

[0006] By adopting the above technical solution and taking advantage of the high-speed transmission characteristics of the 5G network, the solution can achieve fast real-time video transmission. By recording the network data and monitoring data of the uploader and the receiver and building relevant data sets, the network status can be monitored and analyzed in real time. By using machine self-learning algorithms and real-time monitored network data, the transmission strategy can be dynamically adjusted to adapt to network changes, thereby adaptively selecting the best video upload and download strategy. By predicting the video upload efficiency and dynamically adjusting the transmission parameters, the video transmission process can be optimized in advance to improve the transmission efficiency and quality. The solution generates personalized transmission strategies based on video download request data such as the number and clarity of the receiver, so that different users can obtain customized video transmission services according to their needs, providing a better user experience.

[0007] In a preferred example, the present application may be further configured as follows: in the step of obtaining basic parameters of the real-time video stream and the network conditions of the upload end, and generating a first upload strategy based on the basic parameters of the video stream and the network conditions of the upload end, the steps are as follows: Obtain the content of the real-time video stream and label the video according to the content to build a content profile of the upload end; Obtain the historical upload network data of the real-time video stream and associate it with the corresponding video tags to build an upload network fluctuation dataset; The preset upload end network prediction model analyzes the upload end network fluctuation data set based on a machine self-learning algorithm to generate upload end network fluctuation prediction data corresponding to different video tags; The first uploading strategy is generated based on the upload end network fluctuation prediction data and the real-time video stream basic parameters.

[0008] By adopting the above technical solution, the real-time video stream content is labeled and processed to understand the characteristics and content of the video, providing a more accurate reference for the subsequent upload network fluctuation prediction and strategy generation. By constructing an upload network fluctuation data set, the fluctuation of the upload network under different video tags can be analyzed to provide data support for the prediction and strategy generation of the upload network. The first upload strategy is generated based on the upload network fluctuation prediction data and the basic parameters of the video stream, and the preliminary upload strategy is determined according to the network situation and video characteristics to improve the efficiency and quality of video transmission.

[0009] In a preferred example, the present application may be further configured as follows: in the step of analyzing the upload network-upload monitoring data set based on the machine self-learning algorithm in the preset upload analysis model to generate the upload network influence coefficient, the steps are as follows: Obtaining a mapping relationship between the network data in the uploaded monitoring data and the uploaded monitoring data in the uploaded network-upload monitoring data set, wherein the mapping relationship of the uploaded network upload monitoring data set includes: network data-upload speed mapping, network data-delay mapping, network data-packet loss rate mapping, and network data-bandwidth utilization mapping; The upload analysis model analyzes the mapping relationship between the upload network and the upload monitoring data set to generate a network data-upload speed relationship sub-factor, a network data-delay mapping relationship sub-factor, a network data-packet loss rate mapping relationship sub-factor, and a network data-bandwidth utilization mapping relationship sub-factor; Calculate the influence weight corresponding to each mapping relationship in the upload network-upload monitoring data set based on the upload network-upload monitoring data set; An upload network influence coefficient is generated based on the influence weights and sub-factors of each mapping relationship between the upload network and the upload monitoring data set.

[0010] By adopting the above technical solution, according to the influence weights and sub-factors of each mapping relationship in the upload network-upload monitoring data set, combined with a preset algorithm, an upload network influence coefficient is generated, such as an upload speed influence coefficient, a delay influence coefficient, a packet loss rate influence coefficient, and a bandwidth utilization influence coefficient. By generating the upload network influence coefficient based on the influence weights and sub-factors of each mapping relationship in the upload network-upload monitoring data set, the influence of different network data on the upload speed can be comprehensively considered, providing a basis for subsequent upload strategy generation and adjustment.

[0011] In a preferred example, the present application may be further configured as follows: before the step of obtaining the video download request data and generating the first download transmission strategy for transmission to the receiving end based on the predicted cache data corresponding to the second upload strategy, the step includes: Recording historical download request data, and associating the historical download request data with the content portrait of the upload end based on the common timeline to construct a request data set; The preset download request prediction model analyzes the request data set to generate download request prediction data for predicting download request trends corresponding to different content portraits.

[0012] By adopting the above technical solution, the constructed request data set is analyzed using a pre-set download request prediction model. In an embodiment of the present application, the download request prediction model is provided with a decision tree self-learning algorithm and a neural network algorithm. Through this model, in a live video broadcast scenario, a decision tree model is used to predict the download request trends of viewers for different channels based on the historical behaviors of viewers for different live channels, such as viewing time, interaction frequency, etc. Different content portraits can be predicted, such as the download request trends corresponding to live channels. A deep learning model is used to learn the complex patterns of the download requests of viewers for different live channels, that is, to predict the download needs of viewers for different content. By analyzing the request data set through the prediction model, the download request trends corresponding to different content portraits can be predicted, which helps to understand the audience's needs for different content in advance and provide a basis for the subsequent generation and adjustment of transmission strategies.

[0013] In a preferred example, the present application may be further configured as follows: after the step of analyzing the request data set in a preset download request prediction model to generate download request prediction data for predicting download request trends corresponding to different content portraits, the following steps are included: Comparing the real-time download request data with the download request prediction data to generate prediction deviation data; The download request prediction data is corrected based on the prediction deviation data to generate corrected download request prediction data, and the corrected download request prediction data is used to correct the first download transmission strategy.

[0014] By adopting the above technical solution, the download request prediction data is corrected based on the prediction deviation data, the accuracy of the prediction is improved, and more accurate prediction data is provided for the correction of the first download transmission strategy.

[0015] In a preferred example, the present application may be further configured as follows: after the step of analyzing the download network-download monitoring data set based on the machine self-learning algorithm by the preset download analysis model to generate a download network influence coefficient, and modifying the first download transmission strategy based on the download network influence coefficient to generate a second download transmission strategy, the following steps are included: Acquire transmission bandwidth data in the second download transmission strategy; Preprocessing the video stream corresponding to the video download request data based on the bandwidth data to generate a processed video stream; Acquire sending feedback data, where the sending feedback data includes receiving waiting time and download cache time; The processing strategy for processing the video stream is adjusted based on the sent feedback data.

[0016] By adopting the above technical solution, the video stream is preprocessed, including compressing the video, adjusting the resolution or selecting a suitable encoding method to ensure that the video stream can be effectively transmitted under the available transmission bandwidth. The receiving waiting time indicates the time the audience waits for the video stream to start playing, and the download cache time indicates the buffering time of the video stream on the audience side. According to the feedback data of the receiving waiting time and the download cache time, the strategy for processing the video stream can be adjusted, such as adjusting the encoding method of the video, selecting a more suitable resolution or adjusting the transmission method, so as to improve the audience's viewing experience.

[0017] In a preferred example, the present application can be further configured as follows: the download request prediction model is pre-set with a decision tree self-learning algorithm and a neural network algorithm.

[0018] By adopting the above technical solutions, the decision tree divides the data set into different categories according to the conditions of the features, which is suitable for predicting discrete variables. The neural network model learns the complex relationship between input features and outputs through the connection of multiple layers of neurons, which is suitable for processing large-scale data and complex nonlinear relationships, and improving the accuracy of model calculations.

[0019] In the second aspect, the above invention objective of the present application is achieved through the following technical solutions: A real-time video streaming transmission device for a smart phone based on a 5G network, the device comprising: A first upload strategy generation module, used to obtain basic parameters of the real-time video stream and the network status of the upload end, and generate a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; An upload network-upload monitoring data set construction module is used to record the network data and upload monitoring data of the upload end based on a preset public time axis, and form a mapping relationship between the network data of the upload end and the upload monitoring data based on the public time axis to construct an upload network-upload monitoring data set, wherein the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization; An upload network influence coefficient generation module, used to analyze the upload network-upload monitoring data set based on a preset upload analysis model based on a machine self-learning algorithm to generate an upload network influence coefficient, wherein the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading a video stream; A second upload strategy generation module is used to modify the first upload strategy based on the predicted video upload efficiency to generate a second upload strategy, calculate the predicted cache data and adjust the data upload channel data based on the second upload strategy, wherein the video stream data is received, and the video stream data uploaded to the server is parsed to obtain the cache data cached in the server, and the cache data is used to send to the video receiving end; a first download transmission strategy generating module, configured to obtain video download request data, and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of receiving ends; A download network-download monitoring data set construction module, used to obtain the download network data and download monitoring data of the receiving end, and form a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct the download network-download monitoring data set; The second download transmission strategy generation module is used to analyze the download network-download monitoring data set based on a preset download analysis model and a machine self-learning algorithm to generate a download network impact coefficient, and to modify the first download transmission strategy based on the download network impact coefficient to generate a second download transmission strategy.

[0020] On the third aspect, the above-mentioned purpose of the present application is achieved through the following technical solutions: An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned method for real-time video streaming transmission for a smartphone based on a 5G network are implemented.

[0021] Fourthly, the above-mentioned purpose of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for real-time video streaming transmission for a smartphone based on a 5G network.

[0022] In summary, the present application includes at least one of the following beneficial technical effects: 1. Taking advantage of the high-speed transmission characteristics of the 5G network, the solution can achieve fast real-time video transmission. By recording the network data and monitoring data of the uploader and the receiver and building relevant data sets, the network status can be monitored and analyzed in real time. By using machine self-learning algorithms and real-time monitored network data, the transmission strategy can be dynamically adjusted to adapt to network changes, thereby adaptively selecting the best video upload and download strategy. By predicting the video upload efficiency and dynamically adjusting the transmission parameters, the video transmission process can be optimized in advance to improve the transmission efficiency and quality. The solution generates personalized transmission strategies based on the video download request data such as the number and clarity of the receiver, so that different users can obtain customized video transmission services according to their needs, providing a better user experience; 2. Label the content of real-time video streams to understand the characteristics and content of the video, and provide a more accurate reference for subsequent upload network fluctuation prediction and strategy generation. By constructing an upload network fluctuation data set, the fluctuation of the upload network under different video tags can be analyzed to provide data support for the prediction and strategy generation of the upload network. The first upload strategy is generated based on the upload network fluctuation prediction data and the basic parameters of the video stream, and the preliminary upload strategy is determined according to the network situation and video characteristics to improve the efficiency and quality of video transmission. 3. According to the influence weights and sub-factors of each mapping relationship in the upload network-upload monitoring data set, combined with the preset algorithm, the upload network influence coefficient is generated, such as the upload speed influence coefficient, the delay influence coefficient, the packet loss rate influence coefficient, and the bandwidth utilization influence coefficient. By generating the upload network influence coefficient based on the influence weights and sub-factors of each mapping relationship in the upload network-upload monitoring data set, the influence of different network data on the upload speed can be comprehensively considered, providing a basis for the subsequent generation and adjustment of the upload strategy; 4. Preprocess the video stream, including compressing the video, adjusting the resolution or selecting a suitable encoding method to ensure that the video stream can be effectively transmitted under the available transmission bandwidth. The receiving wait time indicates the time the audience waits for the video stream to start playing, and the download cache time indicates the buffering time of the video stream on the audience side. According to the feedback data of the receiving wait time and the download cache time, the strategy for processing the video stream can be adjusted, such as adjusting the video encoding method, selecting a more suitable resolution or adjusting the transmission method, to improve the audience's viewing experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a flowchart of step S10 in a real-time video streaming transmission method for a smartphone based on a 5G network in an embodiment of the present application; Figure 2 This is a principle block diagram of a real-time video streaming transmission device for a smartphone based on a 5G network in one embodiment of the present application; Figure 3 It is a schematic diagram of an electronic device in an embodiment of the present application. Attached photos

[0024] 1. First upload strategy generation module; 2. Upload network-upload monitoring data set construction module; 3. Upload network influence coefficient generation module; 4. Second upload strategy generation module; 5. First download transmission strategy generation module; 6. Download network-download monitoring data set construction module; 7. Second download transmission strategy generation module. DETAILED DESCRIPTION

[0025] The present application is further described in detail below in conjunction with the accompanying drawings.

[0026] In one embodiment, taking live video as an example, Figure 1 As shown, the present application discloses a real-time video streaming transmission method for a smartphone based on a 5G network, which specifically includes the following steps: S10: acquiring basic parameters of the real-time video stream and the network status of the upload end, and generating a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; Specifically, in the embodiment of the present application, for live video, the basic parameters of the real-time video stream include video resolution, encoding method, frame rate, etc., and the network conditions of the upload end include upload bandwidth, delay, packet loss rate, etc. Based on these parameters, a first upload strategy is generated, such as determining the initial video compression ratio and transmission protocol to adapt to the current network conditions.

[0027] S20: Recording the network data and the uploaded monitoring data of the upload end based on a preset common time axis, and forming a mapping relationship between the network data and the uploaded monitoring data of the upload end based on the common time axis to construct an uploaded network-uploaded monitoring data set; Specifically, the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization rate. In the live video scenario, the network data of the upload end, such as upload speed, delay, packet loss rate, etc., can be recorded at regular intervals and mapped with the monitoring data during the actual video upload process to form an upload network-upload monitoring data set. By constructing an upload network-upload monitoring data set, we can gain an in-depth understanding of the performance indicators of the upload end network and provide data support for the subsequent upload network impact coefficient analysis.

[0028] S30: The preset upload analysis model analyzes the upload network-upload monitoring data set based on a machine self-learning algorithm to generate an upload network influence coefficient, wherein the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading the video stream; S40: Based on the predicted video upload efficiency, the first upload strategy is modified to generate a second upload strategy, and based on the second upload strategy, predicted cache data is calculated and data upload channel data is adjusted, wherein video stream data is received, and the video stream data uploaded to the server is parsed to obtain cache data cached in the server, and the cache data is used to send to the video receiving end; Specifically, according to the predicted video upload efficiency, the first upload strategy is adjusted, such as adjusting the video compression ratio and selecting an appropriate transmission protocol, to generate the second upload strategy, calculate the predicted cache data and adjust the data upload channel data. By correcting the upload strategy, it is better to adapt to network changes, improve video upload efficiency and stability, and ensure real-time transmission of video data.

[0029] S50: Obtain video download request data, and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of receiving ends; Specifically, the video download request data of the receiving end is obtained, including the number of receiving ends and the demand for video clarity, and a first download transmission strategy for transmission to the receiving end is generated based on the predicted cache data corresponding to the second upload strategy. The first download transmission strategy generated based on the second upload strategy can better meet the video download demand of the receiving end, and improve the efficiency of video transmission and user experience.

[0030] S60: Obtain download network data and download monitoring data of the receiving end, and form a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct a download network-download monitoring data set; Specifically, the download network data and download monitoring data of the receiving end, such as download speed, delay, packet loss rate, etc., are recorded, and a download network-download monitoring data set is formed for subsequent analysis.

[0031] S70: The preset download analysis model analyzes the download network-download monitoring data set based on a machine self-learning algorithm to generate a download network impact coefficient, and modifies the first download transmission strategy based on the download network impact coefficient to generate a second download transmission strategy.

[0032] Specifically, a download network influence coefficient is generated, a first download transmission strategy is generated based on the influence coefficient and video download request data, and the first download transmission strategy is modified according to the second upload strategy to generate a second download transmission strategy.

[0033] By generating the download network impact coefficient and adjusting the download transmission strategy, the video receiving end transmission process is optimized, and the video download efficiency and the receiving end viewing experience are improved.

[0034] In summary, compared with the existing technology, the outstanding features of this solution include: this solution makes full use of the high bandwidth, low latency and high reliability of the 5G network, optimizes the 5G network environment, and realizes efficient real-time video transmission. Compared with the traditional 4G network, the 5G network can provide faster transmission speed and lower latency, providing better basic conditions for real-time video transmission; adaptive optimization is achieved by analyzing and monitoring the network data of the uploader and the receiver, and real-time optimization is performed according to the network conditions and user needs by dynamically adjusting the upload and download strategies to improve the stability and quality of video transmission. Compared with the traditional fixed strategy, this solution can make flexible adjustments according to the actual situation to adapt to different network environments and user needs. The network conditions and needs of the uploader and the receiver are comprehensively considered, and the end-to-end video transmission optimization is realized by optimizing the upload and download strategies. By generating the upload network influence coefficient and the download network influence coefficient, combining the video download request data and the predicted cache data, and comprehensively considering various factors in the upload and download process, the overall video transmission efficiency and user experience are improved.

[0035] In S10: obtaining basic parameters of the real-time video stream and the network status of the upload end, and generating a first upload strategy based on the basic parameters of the video stream and the network status of the upload end, the following steps are included: S11: Obtain the content of the real-time video stream and label the video according to the content, so as to construct a content profile of the upload end; Specifically, for the live video broadcast scenario, the content of the real-time video stream can be parsed to extract the video tag information, such as video theme, scene, characters, etc., and these tags can be processed to build a content portrait on the upload end. It is further explained that classification and labeling can be based on existing classifications in the live broadcast platform, such as games, outdoor, dance, etc.

[0036] By labeling the real-time video stream content, we can understand the characteristics and content of the video, and provide a more accurate reference for subsequent upload-side network fluctuation prediction and strategy generation.

[0037] S12: Obtain historical upload-end network data of the real-time video stream and associate it with the corresponding video tag to construct an upload-end network fluctuation dataset; Specifically, we collect historical upload network data, such as upload speed, delay, packet loss rate, etc., and associate these data with the corresponding video tags to build an upload network fluctuation dataset. By building an upload network fluctuation dataset, we can analyze the fluctuation of the upload network under different video tags and provide data support for the prediction and strategy generation of the upload network.

[0038] S13: The preset upload-end network prediction model analyzes the upload-end network fluctuation data set based on a machine self-learning algorithm to generate upload-end network fluctuation prediction data corresponding to different video tags; Specifically, using a machine learning algorithm, in an embodiment of the present application, time series analysis and regression analysis are used to analyze the upload network fluctuation data set, train a prediction model, and generate upload network fluctuation prediction data corresponding to different video tags based on the model. For example, for a type of outdoor live broadcast where the geographical variation of the tag is large during the live broadcast, the network fluctuation is more obvious.

[0039] By analyzing the upload network prediction model, we can predict the fluctuation of the upload network under different video tags and provide prediction data for the subsequent upload strategy generation.

[0040] S14: generating the first uploading strategy based on the upload end network fluctuation prediction data and the real-time video stream basic parameters; Specifically, according to the network fluctuation prediction data of the upload end and the basic parameters of the real-time video stream, such as video resolution, frame rate, etc., combined with the preset strategy generation algorithm, a first upload strategy is generated. In the embodiment of the present application, the first upload strategy includes an initial video compression ratio, a transmission protocol, etc. By generating the first upload strategy based on the network fluctuation prediction data of the upload end and the basic parameters of the video stream, a preliminary upload strategy determination can be made according to the network situation and video characteristics to improve the efficiency and quality of video transmission.

[0041] In S30: the preset upload analysis model analyzes the upload network-upload monitoring data set based on a machine self-learning algorithm to generate an upload network influence coefficient, the steps include the following: S31: Obtaining a mapping relationship between the network data in the uploaded monitoring data and the uploaded monitoring data in the uploaded network-upload monitoring data set, wherein the mapping relationship between the uploaded network-upload monitoring data set includes: network data-upload speed mapping, network data-delay mapping, network data-packet loss rate mapping, and network data-bandwidth utilization mapping; By obtaining the mapping relationship between the network data in the upload monitoring data set and the upload monitoring data, we can understand the relationship between different network data and upload speed, and provide basic data for the subsequent upload network impact coefficient analysis.

[0042] S32: the upload analysis model analyzes the mapping relationship between the upload network and the upload monitoring data set to generate a network data-upload speed relationship sub-factor, a network data-delay mapping relationship sub-factor, a network data-packet loss rate mapping relationship sub-factor, and a network data-bandwidth utilization mapping relationship sub-factor; By analyzing the mapping relationship between the upload network and the upload monitoring data set through the upload analysis model, the correlation between different network data and the upload speed can be determined, and the sub-factors between the network data and the upload speed can be extracted, providing a basis for the subsequent calculation of the upload network impact coefficient.

[0043] S33: calculating the influence weight corresponding to each mapping relationship in the upload network-upload monitoring data set based on the upload network-upload monitoring data set; Specifically, based on the data and analysis results of each mapping relationship in the upload network-upload monitoring data set, the impact weight corresponding to each mapping relationship is calculated, such as the impact weight of upload speed, the impact weight of delay, the impact weight of packet loss rate, and the impact weight of bandwidth utilization.

[0044] By calculating the influence weights of each mapping relationship in the upload network-upload monitoring data set, the influence of different mapping relationships on the upload speed can be determined, providing a weight reference for the subsequent generation of the upload network influence coefficient.

[0045] S34: generating an upload network influence coefficient based on the influence weights and sub-factors of each mapping relationship between the upload network and the upload monitoring data set; Specifically, according to the influence weights and sub-factors of each mapping relationship in the upload network-upload monitoring data set, combined with a preset algorithm, an upload network influence coefficient is generated, such as an upload speed influence coefficient, a delay influence coefficient, a packet loss rate influence coefficient, and a bandwidth utilization influence coefficient. By generating an upload network influence coefficient based on the influence weights and sub-factors of each mapping relationship in the upload network-upload monitoring data set, the impact of different network data on the upload speed can be comprehensively considered, providing a basis for subsequent upload strategy generation and adjustment.

[0046] Before the step of obtaining the video download request data and generating the first download transmission strategy for transmission to the receiving end based on the predicted cache data corresponding to the second upload strategy in S50, the following steps are included: S501: Record historical download request data, and associate the historical download request data with a content portrait of the upload end based on the common timeline to construct a request data set; Specifically, in the live video broadcast scenario, historical download request data is recorded, including the audience's download requests for different content profiles (such as different live broadcast channels). By associating the historical download request data with the content profile of the upload end (such as the characteristics of the live broadcast channel), a request data set is constructed to facilitate subsequent download request prediction and analysis.

[0047] By recording historical download request data and associating it with content portraits, we can obtain the download request trends of viewers for different content portraits, providing basic data for subsequent download request predictions.

[0048] S502: Analyzing the request data set using a preset download request prediction model to generate download request prediction data for predicting download request trends corresponding to different content profiles; The constructed request data set is analyzed using a pre-set download request prediction model (such as a model based on machine learning or statistical analysis). In the embodiment of the present application, the download request prediction model is provided with a decision tree self-learning algorithm and a neural network algorithm. Through this model, in the video live broadcast scenario, a decision tree model is used to predict the download request trend of the audience for different channels based on the audience's historical behavior on different live channels (such as viewing time, interaction frequency, etc.). The download request trend corresponding to different content portraits (such as live channels) can be predicted, and a deep learning model (convolutional neural network) is used to learn the complex pattern of the audience's download request for different live channels, that is, to predict the audience's download demand for different content. By analyzing the request data set through the prediction model, the download request trend corresponding to different content portraits can be predicted, which helps to understand the audience's demand for different content in advance and provide a basis for the subsequent transmission strategy generation and adjustment.

[0049] By predicting the demand trend of video streams, the transmission strategy is adjusted to set the transmission channel and video format that meet the video stream needs in advance, thereby optimizing the video stream transmission efficiency while not wasting server resources.

[0050] After the step S502: a preset download request prediction model analyzes the request data set to generate download request prediction data for predicting download request trends corresponding to different content profiles, the following steps are included: S5021: Compare the real-time download request data with the download request prediction data to generate prediction deviation data; S5022: Correcting the download request prediction data based on the prediction deviation data to generate corrected download request prediction data, wherein the corrected download request prediction data is used to correct the first download transmission strategy.

[0051] In the embodiment of the present application, an example is given: in a live video broadcast scenario, assuming that the prediction model predicts that the download request volume of a live broadcast channel in the next hour is 1000 times, and the real-time download request data shows that the actual request volume is 1200 times. Then the prediction deviation data is the difference between the actual request volume and the predicted request volume, that is, 1200 - 1000 = 200 times, and the prediction deviation data is 200 times. The prediction value is corrected according to the prediction deviation data, and the predicted request volume is adjusted to the original prediction value plus the prediction deviation value, that is, 1000 + 200 = 1200 times, and the corrected download request prediction data is obtained.

[0052] Specifically, the prediction deviation data represents the difference between the actual download request volume and the predicted request volume, a positive value represents that the actual request volume is higher than the predicted value, and a negative value represents that the actual request volume is lower than the predicted value.

[0053] Correction based on positive deviation: If the predicted deviation data is positive, it means that the actual request volume is higher than the predicted value, and you can consider increasing the allocation of download resources or increasing the transmission speed to meet the needs of the audience. In the embodiment of the present application, by increasing the bandwidth of the server or increasing the number of concurrent transmissions of the video source to provide more download resources, or adjusting the transmission strategy to increase the transmission speed.

[0054] Correction based on negative deviation: If the predicted deviation data is negative, it means that the actual request volume is lower than the predicted value. Consider reducing the allocation of download resources or reducing the transmission speed to avoid wasting resources. In the embodiment of the present application, the bandwidth of the server is reduced or the number of concurrent transmissions of the video source is reduced to save resources, and the speed is appropriately reduced according to the actual needs of the audience.

[0055] It should be noted that when revising the first download transmission strategy, multiple factors should be considered comprehensively, such as network bandwidth, server load, audience demand, etc. Different strategies can be adopted to balance resource utilization and user experience according to actual conditions and system performance.

[0056] The download request prediction data is corrected based on the prediction deviation data to improve the accuracy of the prediction and provide more accurate prediction data for the correction of the first download transmission strategy.

[0057] After the step S70: the preset download analysis model analyzes the download network-download monitoring data set based on the machine self-learning algorithm to generate a download network influence coefficient, and modifies the first download transmission strategy based on the download network influence coefficient to generate a second download transmission strategy, the following steps are included: S701: Acquire transmission bandwidth data in the second download transmission strategy; S702: Preprocessing the video stream corresponding to the video download request data based on the bandwidth data to generate a processed video stream; S703: Acquire sending feedback data, where the sending feedback data includes receiving waiting time and download cache time; S704: Adjust the processing strategy for processing the video stream based on the sent feedback data.

[0058] In an embodiment of the present application, it is assumed that the second download transmission strategy determines that the available transmission bandwidth in a certain time period is 10 Mbps. According to the 10 Mbps transmission bandwidth data, the video stream is preprocessed, including compressing the video, adjusting the resolution, or selecting a suitable encoding method to ensure that the video stream can be effectively transmitted under the available transmission bandwidth. The receiving waiting time indicates the time the audience waits for the video stream to start playing, and the download cache time indicates the buffering time of the video stream at the audience end. According to the feedback data of the receiving waiting time and the download cache time, the strategy for processing the video stream can be adjusted, such as adjusting the encoding method of the video, selecting a more suitable resolution, or adjusting the transmission method, so as to improve the audience's viewing experience.

[0059] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0060] In one embodiment, a real-time video stream transmission device for a smartphone based on a 5G network is provided, and the real-time video stream transmission device for a smartphone based on a 5G network corresponds one-to-one to the real-time video stream transmission method for a smartphone based on a 5G network in the above-mentioned embodiment. Figure 2 As shown, the real-time video streaming transmission device for a smartphone based on a 5G network includes: A first upload strategy generating module 1, used to obtain basic parameters of the real-time video stream and the network status of the upload end, and generate a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; Upload network-upload monitoring data set construction module 2, used to record the network data and upload monitoring data of the upload end based on a preset public time axis, and form a mapping relationship between the network data of the upload end and the upload monitoring data based on the public time axis to construct an upload network-upload monitoring data set, wherein the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization; An upload network influence coefficient generating module 3 is used to analyze the upload network-upload monitoring data set based on a preset upload analysis model based on a machine self-learning algorithm to generate an upload network influence coefficient, and the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading a video stream; A second upload strategy generating module 4 is used to modify the first upload strategy based on the predicted video upload efficiency to generate a second upload strategy, calculate the predicted cache data and adjust the data upload channel data based on the second upload strategy, wherein the video stream data is received, and the video stream data uploaded to the server is parsed to obtain the cache data cached in the server, and the cache data is used to send to the video receiving end; A first download transmission strategy generating module 5, configured to obtain video download request data, and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of receiving ends; A download network-download monitoring data set construction module 6 is used to obtain the download network data and download monitoring data of the receiving end, and form a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct a download network-download monitoring data set; The second download transmission strategy generation module 7 is used to analyze the download network-download monitoring data set based on a preset download analysis model and a machine self-learning algorithm to generate a download network impact coefficient, and to modify the first download transmission strategy based on the download network impact coefficient to generate a second download transmission strategy.

[0061] For the specific definition of the real-time video streaming transmission device for smartphones based on 5G networks, please refer to the definition of the real-time video streaming transmission method for smartphones based on 5G networks above, which will not be repeated here. Each module in the above-mentioned real-time video streaming transmission device for smartphones based on 5G networks can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the electronic device in the form of hardware, or can be stored in the memory of the electronic device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0062] In one embodiment, an electronic device is provided. The electronic device may be a server, and its internal structure diagram may be as follows: Figure 3As shown. The electronic device includes a processor, a memory, a network interface and a database connected via a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the electronic device is used to store a database. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a real-time video streaming method for a smartphone based on a 5G network is implemented.

[0063] In one embodiment, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: Acquire basic parameters of the real-time video stream and the network status of the upload end, and generate a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; Based on a preset public timeline, the network data and the upload monitoring data of the upload end are recorded, and based on the public timeline, a mapping relationship between the network data of the upload end and the upload monitoring data is formed to construct an upload network-upload monitoring data set, wherein the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization; The preset upload analysis model analyzes the upload network-upload monitoring data set based on a machine self-learning algorithm to generate an upload network influence coefficient, and the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading the video stream; Based on the predicted video upload efficiency, the first upload strategy is modified to generate a second upload strategy, and based on the second upload strategy, predicted cache data is calculated and data upload channel data is adjusted, wherein video stream data is received, and the video stream data uploaded to the server is parsed to obtain cache data cached in the server, and the cache data is used to send to the video receiving end; Acquire video download request data, and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of the receiving ends; Acquire the download network data and download monitoring data of the receiving end, and form a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct a download network-download monitoring data set; The preset download analysis model analyzes the download network-download monitoring data set based on a machine self-learning algorithm to generate a download network impact coefficient, and modifies the first download transmission strategy based on the download network impact coefficient to generate a second download transmission strategy.

[0064] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Acquire basic parameters of the real-time video stream and the network status of the upload end, and generate a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; Based on a preset public timeline, the network data and the upload monitoring data of the upload end are recorded, and based on the public timeline, a mapping relationship between the network data of the upload end and the upload monitoring data is formed to construct an upload network-upload monitoring data set, wherein the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization; The preset upload analysis model analyzes the upload network-upload monitoring data set based on a machine self-learning algorithm to generate an upload network influence coefficient, and the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading the video stream; Based on the predicted video upload efficiency, the first upload strategy is modified to generate a second upload strategy, and based on the second upload strategy, predicted cache data is calculated and data upload channel data is adjusted, wherein video stream data is received, and the video stream data uploaded to the server is parsed to obtain cache data cached in the server, and the cache data is used to send to the video receiving end; Acquire video download request data, and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of the receiving ends; Acquire the download network data and download monitoring data of the receiving end, and form a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct a download network-download monitoring data set; The preset download analysis model analyzes the download network-download monitoring data set based on a machine self-learning algorithm to generate a download network impact coefficient, and modifies the first download transmission strategy based on the download network impact coefficient to generate a second download transmission strategy.

[0065] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0066] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0067] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A real-time video streaming transmission method for a smart phone based on a 5G network, characterized in that: The method comprises the steps of: obtaining basic parameters of the real-time video stream and the network status of the upload end, and generating a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; Based on a preset public timeline, the network data and the upload monitoring data of the upload end are recorded, and based on the public timeline, a mapping relationship between the network data of the upload end and the upload monitoring data is formed to construct an upload network-upload monitoring data set, wherein the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization; The preset upload analysis model analyzes the upload network-upload monitoring data set based on a machine self-learning algorithm to generate an upload network influence coefficient, and the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading the video stream; Based on the predicted video upload efficiency, the first upload strategy is modified to generate a second upload strategy, and based on the second upload strategy, predicted cache data is calculated and data upload channel data is adjusted, wherein video stream data is received, and the video stream data uploaded to the server is parsed to obtain cache data cached in the server, and the cache data is used to send to the video receiving end; Acquire video download request data, and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of the receiving ends; Acquire the download network data and download monitoring data of the receiving end, and form a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct a download network-download monitoring data set; The preset download analysis model analyzes the download network-download monitoring data set based on a machine self-learning algorithm to generate a download network impact coefficient, and modifies the first download transmission strategy based on the download network impact coefficient to generate a second download transmission strategy.

2. A method for transmitting real-time video streams for a smart phone based on a 5G network according to claim 1, characterized in that: In the step of obtaining basic parameters of the real-time video stream and the network status of the upload end, and generating a first upload strategy based on the basic parameters of the video stream and the network status of the upload end, the following steps are included: Obtain the content of the real-time video stream and label the video according to the content to build a content profile of the upload end; Obtain the historical upload network data of the real-time video stream and associate it with the corresponding video tags to build an upload network fluctuation dataset; The preset upload end network prediction model analyzes the upload end network fluctuation data set based on a machine self-learning algorithm to generate upload end network fluctuation prediction data corresponding to different video tags; The first uploading strategy is generated based on the upload end network fluctuation prediction data and the real-time video stream basic parameters.

3. The method for transmitting real-time video stream for a smart phone based on a 5G network according to claim 1, characterized in that: In the step of analyzing the upload network-upload monitoring data set based on the machine self-learning algorithm to generate the upload network influence coefficient, the following steps are included: Obtaining a mapping relationship between the network data in the uploaded monitoring data and the uploaded monitoring data in the uploaded network-upload monitoring data set, wherein the mapping relationship between the uploaded network-upload monitoring data set includes: network data-upload speed mapping, network data-delay mapping, network data-packet loss rate mapping, and network data-bandwidth utilization mapping; The upload analysis model analyzes the mapping relationship between the upload network and the upload monitoring data set to generate a network data-upload speed relationship sub-factor, a network data-delay mapping relationship sub-factor, a network data-packet loss rate mapping relationship sub-factor, and a network data-bandwidth utilization mapping relationship sub-factor; Calculate the influence weight corresponding to each mapping relationship in the upload network-upload monitoring data set based on the upload network-upload monitoring data set; An upload network influence coefficient is generated based on the influence weights and sub-factors of each mapping relationship between the upload network and the upload monitoring data set.

4. The method for transmitting real-time video stream for a smart phone based on a 5G network according to claim 1, characterized in that: Before the step of obtaining the video download request data and generating a first download transmission strategy for transmission to the receiving end based on the predicted cache data corresponding to the second upload strategy, the following steps are included: Recording historical download request data, and associating the historical download request data with the content portrait of the upload end based on the common timeline to construct a request data set; The preset download request prediction model analyzes the request data set to generate download request prediction data for predicting download request trends corresponding to different content portraits.

5. The method for transmitting real-time video stream for a smart phone based on a 5G network according to claim 1, characterized in that: Before the step of obtaining the video download request data and generating a first download transmission strategy for transmission to the receiving end based on the predicted cache data corresponding to the second upload strategy, the following steps are included: Recording historical download request data, and associating the historical download request data with the content portrait of the upload end based on the common timeline to construct a request data set; The preset download request prediction model analyzes the request data set to generate download request prediction data for predicting download request trends corresponding to different content portraits.

6. The method for transmitting real-time video stream for a smart phone based on a 5G network according to claim 1, characterized in that: After the step of analyzing the download network-download monitoring data set based on the machine self-learning algorithm by the preset download analysis model to generate a download network influence coefficient, and correcting the first download transmission strategy based on the download network influence coefficient to generate a second download transmission strategy, the following steps are included: Acquire transmission bandwidth data in the second download transmission strategy; Preprocessing the video stream corresponding to the video download request data based on the bandwidth data to generate a processed video stream; Acquire sending feedback data, where the sending feedback data includes receiving waiting time and download cache time; The processing strategy for processing the video stream is adjusted based on the sent feedback data.

7. The method for transmitting real-time video stream for a smart phone based on a 5G network according to claim 4, characterized in that: The download request prediction model is pre-set with a decision tree self-learning algorithm and a neural network algorithm.

8. A real-time video stream transmission device for a smartphone based on a 5G network, applied to a real-time video stream transmission method for a smartphone based on a 5G network as claimed in any one of claims 1 to 7, characterized in that: It comprises a first upload strategy generation module (1), which is used to obtain basic parameters of the real-time video stream and the network status of the upload end, and generate a first upload strategy based on the basic parameters of the video stream and the network status of the upload end; An upload network-upload monitoring data set construction module (2) is used to record the network data and upload monitoring data of the upload end based on a preset common time axis, and to form a mapping relationship between the network data of the upload end and the upload monitoring data based on the common time axis to construct an upload network-upload monitoring data set, wherein the upload monitoring data includes video upload speed, delay, packet loss rate, and bandwidth utilization rate; An upload network influence coefficient generation module (3), used to analyze the upload network-upload monitoring data set based on a preset upload analysis model and a machine self-learning algorithm to generate an upload network influence coefficient, wherein the upload network influence coefficient is used to calculate the predicted video upload efficiency corresponding to the upload network when uploading a video stream; A second upload strategy generating module (4) is used to modify the first upload strategy based on the predicted video upload efficiency to generate a second upload strategy, calculate the predicted cache data and adjust the data upload channel data based on the second upload strategy, wherein the video stream data is received, and the video stream data uploaded to the server is parsed to obtain the cache data cached in the server, and the cache data is used to send to the video receiving end; A first download transmission strategy generating module (5) is used to obtain video download request data and generate a first download transmission strategy for transmission to a receiving end based on the predicted cache data corresponding to the second upload strategy, wherein the video download request data includes the number and definition of receiving ends; A download network-download monitoring data set construction module (6), used for acquiring the download network data and download monitoring data of the receiving end, and forming a mapping relationship between the network data of the receiving end and the download monitoring data based on the common time axis to construct the download network-download monitoring data set; The second download transmission strategy generating module (7) is used to analyze the download network-download monitoring data set based on a preset download analysis model and a machine self-learning algorithm to generate a download network influence coefficient, and to modify the first download transmission strategy based on the download network influence coefficient to generate a second download transmission strategy.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of a real-time video streaming transmission method for a smartphone based on a 5G network are implemented as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a real-time video streaming transmission method for a smartphone based on a 5G network are implemented as described in any one of claims 1 to 7.

Images