WO2018177165A1

WO2018177165A1 - Method and system for optimizing quality network pushed stream

Info

Publication number: WO2018177165A1
Application number: PCT/CN2018/079677
Authority: WO
Inventors: 黄德恒; 谭洪学
Original assignee: 上海七牛信息技术有限公司
Priority date: 2017-03-30
Filing date: 2018-03-20
Publication date: 2018-10-04
Also published as: CN107026856A

Abstract

Provided are a method and system for optimizing the quality of a network pushed stream, wherein the method comprises: collecting feature information about a pushed stream according to a pre-set period; adjusting, according to a current average sending duration of each frame of TCP and the current data throughput of a buffer region, a code rate of a next period, and taking adjusted code rates as target code rates and feeding same back to an encoder; adjusting, according to the current data throughput of the buffer region, a frame rate of the next period, and determining a time interval between two frames according to the adjusted frame rate; filtering, based on the determined time interval, audio/video data entering the encoder so as to generate audio/video data conforming to the adjusted frame rate; and encoding and transmitting the audio/video data according to the adjusted code rate and the adjusted frame rate. By means of the method and system for optimizing the quality of a network pushed stream provided in the present application, a code rate and a frame rate of the pushed stream can be adjusted smoothly.

Description

Optimization method and optimization system for network push flow quality

Technical field

The present application relates to the field of audio and video processing technologies, and in particular, to an optimization method and an optimization system for network push stream quality.

Background technique

In recent years, mobile video broadcasts have become more and more popular, and have been widely used in entertainment, education, gaming, e-sports, sports, etc., and various live broadcast products are highly competitive, and live broadcast products and services that can provide excellent user experience will eventually be here. Stand out from the competition.

The factors affecting the live viewing experience are generally divided into two dimensions, one is the picture clarity (ie, the picture quality), and the other is the fluency. Image quality is positively correlated with code rate and resolution, and fluency is positively correlated with frame rate and data bandwidth. When the bandwidth of the network is constant, the fluency has a negative correlation with the code rate, that is, when the code rate exceeds the actual network data throughput rate, the fluency is reduced. Ideally, when the anchor end and the viewer end have unlimited bandwidth, you only need to select the best image quality that a live broadcast device can bear for streaming, so that you can get a high-quality and smooth high-quality viewing experience. However, considering that the network environment in which the live broadcaster (ie, the anchor) is located varies greatly, packet loss, jitter, and delay often occur on the link from the live broadcast terminal to the push server, resulting in poor live streaming, resulting in a reduced viewing experience. . In this case, we must weigh the two dimensions of image quality and fluency to maintain a better viewing experience. A large number of user feedback indicates that using lower image quality and high fluency, with high image quality and frequent stagnation Compared with the live broadcast, the former can bring a better user viewing experience. At this time, a technical solution that can dynamically adjust the quality of the push flow according to the network quality of the live end is particularly necessary.

At present, the universal protocol standard for live streaming is RTMP (Real Time Messaging Protocol), which is based on TCP transmission. It is a connection-oriented and reliable transport layer protocol, and internally implements a congestion control strategy. In a network environment with high packet loss rate, the automatic retransmission strategy of TCP occupies a large part of the bandwidth, which leads to a decrease in network data throughput per unit time, resulting in a jam on the playback end. There are three main options for dealing with weak network traffic on the market today:

The anchor broadcasts the clarity of the push stream according to the feedback of the audience, that is, the push stream does not automatically judge the network processing;

Determine the network quality at the beginning of the live broadcast. If poor network conditions are detected, the live broadcast is started directly with a lower push flow configuration, and the configuration is always used throughout the live broadcast process;

The push stream quality is configured to preset two levels of high and low, and switches between the two levels according to the network condition during the live broadcast.

The above scheme can alleviate the problem that the live terminal push flow is not smooth caused by network fluctuation or weak network to a certain extent, but there are also a series of defects, the specific analysis is as follows:

Relying on the anchor to select the stream definition according to the feedback of the audience, it is too dependent on the experience of the anchor, and the feedback information of the audience cannot accurately determine the root cause of the live card, the network of the anchor is not good or the network of the viewer itself may cause the live broadcast. Negative feedback of experience;

Without considering the factors of network fluctuations, the network may be in a state of fluctuation at the beginning of the live broadcast, resulting in low-quality streaming, and the network is stable during the live broadcast. The live broadcast can only always use low quality and cannot achieve the most viewing experience. optimization. In addition, it may be possible to use high-quality streaming at the beginning, and network fluctuations during the live broadcast process will also cause frequent jams;

Adopting the high and low quality two-quality image configuration to switch back and forth according to the network situation, will bring two problems. The first problem is that the playback end can clearly perceive the jitter of the image quality. Secondly, when switching from low image quality to high image quality frequently, the network load may be instantaneously increased, so that the transmitted code rate exceeds the actual bandwidth, causing the jam, and in general, cannot be done. Switch to smooth.

It should be noted that the above description of the technical background is only for the purpose of facilitating a clear and complete description of the technical solutions of the present application, and is convenient for understanding by those skilled in the art. The above technical solutions are not considered to be well known to those skilled in the art simply because these aspects are set forth in the background section of this application.

Summary of the invention

An object of the embodiments of the present application is to provide an optimization method and an optimization system for network push stream quality, which can smoothly adjust a code rate and a frame rate of a push stream.

To achieve the above objective, the present application provides an optimization method for network push stream quality, where the method includes: collecting feature information of a push stream according to a preset period, where the feature information includes an average code rate, an average frame rate, and a buffer. Area data throughput, average TCP transmission time per frame; adjust the code rate of the next period according to the current average TCP transmission duration per frame and the current buffer data throughput, and use the adjusted code rate as the target rate feedback. To the encoder; adjusting the frame rate of the next cycle according to the current buffer data throughput, and determining a time interval between the two frames according to the adjusted frame rate; and entering the encoder based on the determined time interval The audio and video data is filtered to generate audio and video data that conforms to the adjusted frame rate; the audio and video data is encoded and transmitted according to the adjusted code rate and the adjusted frame rate.

Further, adjusting the code rate of the next period according to the current TCP average transmission time per frame and the current buffer data throughput includes: when the current average TCP transmission time per frame is greater than the first preset risk value, Decrease the current code rate according to the preset code rate step, and use the reduced code rate as the code rate of the next cycle; when the current average TCP transmission time per frame is less than or equal to the first preset risk value, Determining whether the average TCP transmission duration of each frame is greater than a second preset risk value; if the current TCP average transmission duration per frame is greater than a second preset risk value and the current buffer data occupancy rate is greater than a preset threshold, The current code rate is decreased according to the preset code rate step size, and the dropped code rate is used as the code rate of the next cycle.

Further, the method further includes: determining whether the last rate rate adjustment is a falling code rate, and if yes, increasing the preset code rate step size to a preset value, and decreasing the current current value according to the increased preset value Code rate.

Further, adjusting the code rate of the next cycle according to the current average TCP transmission time per frame and the current buffer data throughput includes: determining the relationship between the amount of data sent by the current buffer and the amount of data swallowed, when When the amount of data to be spit is greater than or equal to the amount of data swallowed, the counter value of the counter is incremented by one; when the amount of spit data is less than the amount of swallowed data, the counter value is decremented by one; according to the size of the counter in the current counter, the adjustment is made. The code rate of one cycle.

Further, adjusting the code rate of the next period according to the size of the current counter in the counter includes: when the count value in the counter is less than the first determination threshold, decreasing the current code rate according to the preset code rate step, and The falling code rate is used as the code rate of the next cycle, and the counter value of the counter is reset to zero; when the counter value in the counter is greater than the second determination threshold, the current code is increased according to the preset code rate step size. Rate, and the increased code rate is taken as the code rate of the next cycle, and the counter value of the counter is reset to zero.

Further, the method further includes: determining whether the last rate rate adjustment is a falling code rate, and if yes, increasing the preset code rate step size to a preset value, and decreasing the current current value according to the increased preset value The code rate is determined whether the last code rate adjustment is an increase code rate, and if so, the preset code rate step is increased to a preset value, and the current code rate is increased according to the increased preset value.

Further, the push stream is generated in the following manner: the pre-acquired audio and video data is encoded by an encoder, and the encoded audio and video data is encapsulated and transmitted through a real-time message transmission protocol RTMP to generate the Push flow.

To achieve the above objective, an embodiment of the present application further provides an optimization system for network push quality, where the system includes: a flow data collection control module, configured to collect feature information of the push flow according to a preset period, The feature information includes an average bit rate, an average frame rate, a buffer data throughput, and a TCP average transmission time per frame. The adaptive code rate control module is configured to: according to the current average TCP transmission time per frame and the current buffer data throughput, The code rate of the next cycle is adjusted, and the adjusted code rate is fed back to the encoder as the target code rate; the dynamic frame rate control module is configured to adjust the frame rate of the next cycle according to the current buffer data throughput. And determining a time interval between the two frames according to the adjusted frame rate; filtering the audio and video data entering the encoder based on the determined time interval to generate audio and video data conforming to the adjusted frame rate; and transmitting control The module is configured to encode and transmit audio and video data according to the adjusted code rate and the adjusted frame rate.

Further, the adaptive rate control module specifically includes: a first determining unit, configured to: when the current average transmission time of each frame of the TCP is greater than the first preset risk value, decrease the current code according to a preset code rate step size Rate, and the reduced code rate is used as the code rate of the next cycle; the second determining unit is configured to determine, when the current average TCP transmission time per frame is less than or equal to the first preset risk value, Whether the average TCP transmission duration of each frame is greater than the second preset risk value; the third determining unit is configured to: if the current average TCP transmission duration of each frame is greater than the second preset risk value, and the current buffer data occupancy rate is greater than the pre-predetermined When the threshold is set, the current code rate is decreased according to the preset code rate step, and the dropped code rate is used as the code rate of the next cycle.

Further, the adaptive rate control module specifically includes: a throughput determining unit, configured to determine a relationship between a quantity of data to be buffered and a quantity of data to be swallowed in the current buffer, and when the amount of data to be spit is greater than or equal to the amount of data to be swallowed When the amount of spit data is less than the amount of swallowed data, the counter value is decremented by one; the rate adjusting unit is configured to adjust the next period according to the size of the counter in the current counter. Code rate.

The embodiment of the present invention utilizes the matching effect of the code rate and the frame rate adjustment, and effectively controls the amount of audio and video data transmission under the condition of network fluctuation, thereby alleviating network congestion and bringing a smooth viewing experience to the playing end; When the network quality is restored, it can ensure that the bit rate frame rate of audio and video is pushed in the set optimal quality configuration, and can be performed in a smoother manner when switching the frame rate configuration of different code rates, and will not Bringing a sharp sense of picture quality fluctuations to the viewer.

Specific embodiments of the present application are disclosed in detail with reference to the following description and accompanying drawings, in which <RTIgt; It should be understood that the embodiments of the present application are not limited in scope. The embodiments of the present application include many variations, modifications, and equivalents within the scope of the appended claims.

Features described and/or illustrated with respect to one embodiment may be used in one or more other embodiments in the same or similar manner, in combination with, or in place of, features in other embodiments. .

It should be emphasized that the term "comprising" or "comprises" or "comprising" or "comprising" or "comprising" or "comprising" or "comprises"

DRAWINGS

The drawings are included to provide a further understanding of the embodiments of the present application, and are set forth in the accompanying drawings. Obviously, the drawings in the following description are only some of the embodiments of the present application, and those skilled in the art can obtain other drawings according to the drawings without any inventive labor. In the drawing:

FIG. 1 is a flowchart of a method for optimizing network push quality according to an embodiment of the present application;

2 is a schematic diagram of network traffic quality optimization in an embodiment of the present application;

3 is a flowchart of code rate adjustment in an embodiment of the present application;

FIG. 4 is a functional block diagram of an optimization system for network push quality according to an embodiment of the present application.

detailed description

In order to make those skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present application. The embodiments are only a part of the embodiments of the present application, and not all of them. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope should be included in the scope of the present application.

FIG. 1 is a flowchart of a method for optimizing network push quality according to an embodiment of the present application. Although the processes described below include multiple operations occurring in a particular order, it should be clearly understood that these processes can include more or fewer operations that can be performed sequentially or in parallel (eg, using a parallel processor or a multi-threaded environment). . As shown in FIG. 1, the method may include:

Step S1: collecting feature information of the push stream according to a preset period, where the feature information includes an average code rate, an average frame rate, a buffer data throughput, and an average TCP transmission duration per frame.

In this embodiment, data collection of a current system may be performed for a period of time to obtain historical data of the push stream. Specifically, referring to FIG. 2, the source audio and video files can be sampled by the audio and video collection module, so that audio and video data can be collected from the source audio and video files. In the actual application process, the audio and video collection module can perform audio and video data collection according to a preset sampling period.

In this embodiment, the collected audio and video data can enter the encoder and be encoded with the encoding format and code rate set in the encoder. After the encoded audio and video frames are encapsulated, they can be transmitted through a Real Time Messaging Protocol (RTMP) in the transmission control module to generate the push stream. The audio and video data to be transmitted can be temporarily stored in the transmission buffer in the transmission control module, so that the audio and video data can be transmitted in real time.

In this embodiment, the flow data collection control module may collect the feature information of the push flow according to a preset period, where the feature information may include an average code rate, an average frame rate, a buffer data throughput, and an average TCP transmission time per frame. .

Step S2: Adjust the code rate of the next cycle according to the current average TCP transmission duration and the current buffer data throughput, and feed back the adjusted code rate to the encoder as the target code rate.

In the present embodiment, in order to optimize the push flow quality, the collected current feature information may be analyzed, so that it is possible to predict whether the code rate and the frame rate of the push flow need to be adjusted in the next cycle. Specifically, the code rate of the next cycle may be adjusted according to the current TCP average transmission time per frame and the current buffer data throughput. Referring to FIG. 3, when the average TCP transmission duration of each frame is greater than the first preset risk value, it indicates that a frame loss in the buffer may be about to occur, and the current code may be decreased according to the preset code rate step. Rate, and the code rate after the fall is taken as the code rate of the next cycle.

When the average TCP transmission duration of each frame is less than or equal to the first preset risk value, it may continue to determine whether the current average TCP transmission duration per frame is greater than a second preset risk value. The second preset risk value may be smaller than the first preset risk value.

If the current average transmission time of the TCP per frame is greater than the second preset risk value and the data occupation rate of the current buffer is greater than the preset threshold, the frame loss in the buffer may also occur, and the preset code rate may be used. The step size decreases the current code rate, and the dropped code rate is taken as the code rate of the next cycle.

In the present embodiment, the code rate in each cycle can be analyzed, and the code rate of the next cycle is predicted in real time. In the actual application process, when the network fluctuates, it may cause the code rate of several consecutive cycles to decrease. In this case, in order to be able to reduce the number of times the code rate is lowered, the rate at which the rate is lowered can be increased accordingly. Specifically, it can be determined whether the last rate adjustment performed is a falling code rate, and if yes, indicating that the current network condition is in a continuous decline of the code rate, so that the preset code rate step can be increased to a preset value. , thereby increasing the rate at which the rate is decreased, and decreasing the current code rate according to the increased preset value.

In this embodiment, the buffer data throughput can be analyzed to determine if the code rate of the next cycle needs to be adjusted. Specifically, the relationship between the amount of data to be buffered and the amount of data to be swallowed in the current buffer may be determined. When the amount of data to be spit is greater than or equal to the amount of data to be swallowed, the counter may be incremented by one; When the amount of data is entered, the counter value can be decremented by one. In this way, according to the size of the counter value in the counter, the magnitude relationship between the amount of data to be spouted and the amount of swallowed data in the previous cycles can be known, so that the code rate of the next cycle can be adjusted according to the size of the counter value in the current counter.

Specifically, when the count value in the counter is less than the first determination threshold, it indicates that the buffer is consecutively the data swallowing amount is greater than the data spitting amount. For example, the first determination threshold may be -3, so that it can be indicated that the buffer is more than three times in succession, and the data swallowing amount is greater than the data ejection amount. In this case, the risk of frame loss is generated. Therefore, the current code rate can be decreased according to the preset code rate step, and the dropped code rate is used as the code rate of the next cycle, and the counter value of the counter is reset to zero. .

When the count value in the counter is greater than the second determination threshold, it indicates that the buffer is consecutively more than the data swallowing amount. For example, the second determination threshold may be 3, so that it can be indicated that the buffer is more than three times in a row, and the data discharge amount is greater than the data swallowing amount. In this case, the network status is good and the bandwidth is sufficient. Therefore, the current code rate can be increased according to the preset code rate step, and the increased code rate is used as the code rate of the next cycle, and the counter value of the counter is returned. zero.

In the present embodiment, when determining the buffer data throughput, it can be determined whether the last rate rate adjustment is a falling code rate, and if so, it indicates that the current network condition is continuously decreasing in the code rate, and thus Adding the preset code rate step size to a preset value to increase the rate of the rate reduction, and lowering the current code rate according to the increased preset value;

In addition, it can be determined whether the last rate adjustment performed is an increase code rate, and if so, it indicates that the current network condition is continuously rising in the code rate, so the preset code rate step size can be increased to a preset value. To increase the rate at which the rate increases, and increase the current code rate according to the increased preset value.

In this way, by analyzing the current average TCP transmission duration and the current buffer data throughput, the transmission rate of the next cycle can be predicted in real time, and the code rate can be flexibly switched. Since the analysis is performed every cycle, the bit rate switching is prevented from jumping too much, and the smooth viewing experience of the user is ensured.

Step S3: Adjusting the frame rate of the next cycle according to the current buffer data throughput, and determining a time interval between the two frames according to the adjusted frame rate; and entering the tone of the encoder based on the determined time interval. The video data is filtered to generate audio and video data that conforms to the adjusted frame rate.

In this embodiment, the current buffer data throughput can be analyzed in a manner similar to that in step S2, so that the frame rate of the next cycle can be adjusted. The adjusted frame rate may indicate a time interval between two frames through which audio and video data entering the encoder may be filtered to generate audio and video data that conforms to the adjusted frame rate.

It should be noted that, in this embodiment, the process of rate adjustment may be performed in an adaptive rate control module, and the process of frame rate adjustment may be performed in a dynamic frame rate control module. In various comparison processes, the threshold of each comparison may be determined according to a TCP transmission duration prediction operator or a buffer size prediction operator. The adaptive rate control module and the dynamic frame rate control module may be managed by the adaptive stream quality control module. The adaptive code rate module and the dynamic frame rate control module described above can be individually turned on or off, and the user can determine the application form of the technical solution according to his own needs. If the usage scenario requires maintaining a high image quality push flow as much as possible, the dynamic frame rate control module can be used to preferentially adjust the frame rate to control the amount of data entering the encoder per unit time, thereby controlling the amount of data transmitted; High scenes are preferred for adaptive rate control modules.

Step S4: encoding and transmitting the audio and video data according to the adjusted code rate and the adjusted frame rate.

In the embodiment, the adjusted frame rate can adjust the audio and video data collected by the audio and video collection module, and the adjusted code rate can be applied to the encoder to adjust the data processed in the encoder. Thus, according to the adjusted code rate and the adjusted frame rate, encoding and transmission of audio and video data can be performed.

Referring to FIG. 4, an embodiment of the present application further provides an optimization system for network push quality, where the system includes:

The flow data collection control module 100 is configured to collect feature information of the push flow according to a preset period, where the feature information includes an average code rate, an average frame rate, a buffer data throughput, and an average TCP transmission time per frame;

The adaptive code rate control module 200 is configured to adjust the code rate of the next cycle according to the current average TCP transmission duration and the current buffer data throughput, and feed back the adjusted code rate as the target code rate to the coding. Device

The dynamic frame rate control module 300 is configured to adjust a frame rate of a next cycle according to a current buffer data throughput, and determine a time interval between the two frames according to the adjusted frame rate; Filtering audio and video data entering the encoder to generate audio and video data conforming to the adjusted frame rate;

The transmission control module 400 is configured to perform encoding and transmission of audio and video data according to the adjusted code rate and the adjusted frame rate.

In the embodiment of the present application, the adaptive code rate control module 200 specifically includes:

a first determining unit, configured to: when the current average transmission time of the TCP per frame is greater than the first preset risk value, decrease the current code rate according to the preset code rate step, and use the reduced code rate as the next period Code rate

a second determining unit, configured to determine, when the current average TCP transmission duration of each frame is less than or equal to the first preset risk value, whether the current average TCP transmission duration per frame is greater than a second preset risk value;

a third determining unit, configured to: if the current average TCP transmission time length of each frame is greater than a second preset risk value, and the current buffer data occupation rate is greater than a preset threshold, decrease the current code rate according to the preset code rate step size, The dropped code rate is taken as the code rate of the next cycle.

The throughput determining unit is configured to determine a relationship between the amount of data to be sent and the amount of data to be swallowed in the current buffer. When the amount of data to be spit is greater than or equal to the amount of data to be swallowed, the counter is incremented by one; When the amount of data is less than the amount of data, the counter value is decremented by one;

The code rate adjusting unit is configured to adjust the code rate of the next cycle according to the size of the count value in the current counter.

The above description of various embodiments of the present application is provided to those skilled in the art for the purpose of description. It is not intended to be exhaustive or to limit the invention to the single disclosed embodiments. As described above, various alternatives and variations of the present application will be apparent to those skilled in the art. Thus, while a few alternative embodiments have been discussed in detail, other embodiments will be apparent or apparent to those skilled in the art. The present application is intended to cover all alternatives, modifications, and variations of the present invention, as well as other embodiments that fall within the spirit and scope of the application.

Claims

An optimization method for network push quality, characterized in that the method comprises:

Collecting feature information of the push stream according to a preset period, where the feature information includes an average code rate, an average frame rate, a buffer data throughput, and an average TCP transmission time per frame;

Adjusting the code rate of the next cycle according to the current average TCP transmission time per frame and the current buffer data throughput, and feeding the adjusted code rate as the target code rate to the encoder;

Adjusting the frame rate of the next cycle according to the current buffer data throughput, and determining a time interval between the two frames according to the adjusted frame rate; and performing audio and video data entering the encoder based on the determined time interval Filtering to generate audio and video data that conforms to the adjusted frame rate;

The audio and video data is encoded and transmitted according to the adjusted code rate and the adjusted frame rate.
The network push quality optimization method according to claim 1, wherein adjusting the code rate of the next cycle according to the current average TCP transmission duration and the current buffer data throughput includes:

When the average TCP transmission duration of each frame is greater than the first preset risk value, the current code rate is decreased according to the preset code rate step, and the decreased code rate is used as the code rate of the next cycle;

Determining, if the current average TCP transmission duration of each frame is less than or equal to the first preset risk value, determining whether the current average TCP transmission duration per frame is greater than a second preset risk value;

If the average TCP transmission duration of each frame is greater than the second preset risk value and the current buffer data occupancy rate is greater than a preset threshold, the current code rate is decreased according to the preset code rate step size, and the decreased code rate is obtained. As the code rate of the next cycle.
The method for optimizing network push quality according to claim 2, wherein the method further comprises:

Determining whether the last rate rate adjustment is a falling code rate, and if so, increasing the preset code rate step size to a preset value, and decreasing the current code rate according to the increased preset value.
The network push quality optimization method according to claim 1, wherein adjusting the code rate of the next cycle according to the current average TCP transmission duration and the current buffer data throughput includes:

Determining the relationship between the amount of data sent by the current buffer and the amount of data swallowed. When the amount of data to be spit is greater than or equal to the amount of data being swallowed, the counter is incremented by one; when the amount of spit is less than the amount of data being swallowed, Decrease the counter value by one;

The code rate of the next cycle is adjusted according to the size of the count value in the current counter.
The network push quality optimization method according to claim 4, wherein adjusting the code rate of the next cycle according to the size of the current counter in the counter comprises:

When the count value in the counter is less than the first determination threshold, the current code rate is decreased according to the preset code rate step, and the dropped code rate is used as the code rate of the next cycle, and the counter value of the counter is simultaneously Return to zero;

When the count value in the counter is greater than the second determination threshold, the current code rate is increased according to the preset code rate step, and the increased code rate is used as the code rate of the next cycle, and the counter value of the counter is simultaneously Return to zero.
The method for optimizing network push quality according to claim 5, wherein the method further comprises:

Determining whether the last rate rate adjustment is a falling code rate, and if so, increasing the preset code rate step size to a preset value, and decreasing the current code rate according to the increased preset value;

Determining whether the last rate rate adjustment is an increase code rate, and if so, increasing the preset code rate step size to a preset value, and increasing the current code rate according to the increased preset value.
The method for optimizing network push quality according to claim 1, wherein the push stream is generated as follows:

The pre-acquired audio and video data is encoded by an encoder, and the encoded audio and video data is packetized and transmitted through a real-time message transmission protocol (RTMP) to generate the push stream.
An optimization system for network push quality, characterized in that the system comprises:

The flow data collection control module is configured to collect feature information of the push flow according to a preset period, where the feature information includes an average code rate, an average frame rate, a buffer data throughput, and an average TCP transmission time per frame;

The adaptive rate control module is configured to adjust the code rate of the next cycle according to the current average TCP transmission duration and the current buffer data throughput, and feed the adjusted code rate as the target code rate to the encoder. ;

a dynamic frame rate control module, configured to adjust a frame rate of a next cycle according to a current buffer data throughput, and determine a time interval between the two frames according to the adjusted frame rate; The audio and video data entering the encoder is filtered to generate audio and video data that conforms to the adjusted frame rate;

The transmission control module is configured to encode and transmit audio and video data according to the adjusted code rate and the adjusted frame rate.
The network traffic quality optimization system according to claim 8, wherein the adaptive code rate control module specifically includes:

a first determining unit, configured to: when the current average transmission time of the TCP per frame is greater than the first preset risk value, decrease the current code rate according to the preset code rate step, and use the reduced code rate as the next period Code rate

a second determining unit, configured to determine, when the current average TCP transmission duration of each frame is less than or equal to the first preset risk value, whether the current average TCP transmission duration per frame is greater than a second preset risk value;

a third determining unit, configured to: if the current average TCP transmission time length of each frame is greater than a second preset risk value, and the current buffer data occupation rate is greater than a preset threshold, decrease the current code rate according to the preset code rate step size, The dropped code rate is taken as the code rate of the next cycle.
The network traffic quality optimization system according to claim 8, wherein the adaptive code rate control module specifically includes:

The throughput determining unit is configured to determine a relationship between the amount of data to be sent and the amount of data to be swallowed in the current buffer. When the amount of data to be spit is greater than or equal to the amount of data to be swallowed, the counter is incremented by one; When the amount of data is less than the amount of data, the counter value is decremented by one;

The code rate adjusting unit is configured to adjust the code rate of the next cycle according to the size of the count value in the current counter.