TWI681664B - Detection method - Google Patents

Abstract

A detection method is to collect a video delivery record, and then integrate the video delivery record into playback events to analyze the number of times that the quality of the playback events is poor, wherein the playback event has all of network transmission requires, and the measurement method changes the bit-rate through the video delivery to count the number of times that the quality of the video is not good, so a network system management terminal can periodically integrate multiple play events to analyze the poor quality playback events through time segment, so that it can quickly and/or effectively gain reasons for poor quality of the video delivery.

Description

Detection method

The invention relates to a detection method, in particular to a detection method of network video and audio delivery quality.

The transmission volume of audio and video files on the network is quite large. Early audio and video files were downloaded to the demand side at one time through the supply address, mainly in the form of packets, but because the audio and video files sometimes have a large capacity, they are downloaded at the same time. There may not be a small number of demand ends. When each demand end downloads audio and video files from the supply address at the same time, it often causes the transmission speed to be too slow, resulting in poor transmission quality, resulting in troubles for each demand end, and the system maintenance end cannot actually understand the network audio and video The exact cause of the poor delivery quality makes it impossible to immediately and surely know the solution to the poor quality of the audio and video delivery, or even cause an increase in the labor cost of the system management and maintenance side.

In response to the above problems, the industry has proposed a segmented streaming method. As shown in FIG. 1, the content server 10 of the audio and video supply end divides the audio and video file into a plurality of audio and video clips M to be transmitted from the content server 10 of the audio and video supply end to the plurality of delivery servers 11 in the network system room, And then through the delivery server 11 to the demand side of the electronic device 12 (such as personal computers, smart phones or other) to achieve the effect of load balancing, to solve the single The transmission side may cause congestion.

However, the network transmission system operated by the conventional segmented streaming method is limited by the influence of the existing network transmission rate, so the video and audio quality presented by the electronic device 12 on the demand side will be unstable, and the network system The operator can only detect the current network status based on the notification on the demand side or the usage time notified on the demand side to estimate the cause of unstable audio and video quality, so only the delivery server set up by the network system operator can be used 11. Inspection of base stations or lines, and it is impossible to check the electronic device 12 on the demand side or the audio and video supply side (the original audio and video files and the content server), so that the true cause of the unstable audio and video quality cannot even be known, or even A large amount of manpower needs to be sent to check the line or machine, which leads to a significant increase in the cost of testing.

Therefore, how to overcome the above-mentioned problems of the conventional technology has become an urgent problem to be solved at present.

In order to solve the above-mentioned problems, the present invention provides a detection method, which includes: collecting audiovisual delivery records, wherein the audiovisual delivery records are a content server of a supply end transmitting at least one audiovisual information to at least one delivery server to The status of at least one end-use device, and the audio-visual information is a plurality of audio-visual segments divided from an audio-visual file; integrating the audio-visual delivery record into a playback event; and analyzing the number of times of poor quality in the playback event, where, The playback event has the total number of network transmission requirements, and the number of times of poor quality of the audio and video clips is counted through the video and audio delivery record replacement bit rate.

In the aforementioned detection method, the complex includes connecting the The audio quality measurement device of the delivery server collects the audio and video delivery records of each delivery server.

In the aforementioned detection method, the playback event includes the specification and source address of the user device, the name of the video clip, the transmission speed of the video information transmitted by the delivery server, and/or the target address of the delivery server .

In the aforementioned detection method, the number of times of poor quality is caused by the change in the bit rate caused by the transmission quality of each audio and video delivery record of the audio and video information transmitted by the delivery server to identify the total network transmission requirements Times of poor quality.

In the foregoing detection method, the complex includes outputting analysis data for the number of times of poor quality according to the playback event to generate a statistical report. Further, the method includes using the playback event and the data content of at least one database to learn the reason for the poor quality, wherein the data content of the database corresponds to the data content of the playback event.

In the aforementioned detection method, the number of times of poor quality in the playback event excludes or does not include the first decrease in the bit rate.

In summary, the detection method of the present invention mainly analyzes the characteristics of the bit rate of video and audio delivery, measures the delivery status of the delivery server through the video and audio quality measurer, and counts the video and audio information by section identification The number of times the delivery quality is not good, so it can regularly provide statistical reports to the system management end for inspection. Therefore, the detection method used by the present invention to integrate playback events to check the bit rate can not only systematically classify the audio and video clips for statistical analysis, but also shorten the detection time of the system management end. In order to greatly reduce the labor cost consumed by the system management end, the reason for the poor quality of the network audio and video delivery can be further quickly and/or effectively known.

10,101‧‧‧Content server

11,103‧‧‧Delivery server

12‧‧‧Electronic device

102‧‧‧Video information

104‧‧‧Video quality measuring instrument

9a~9d‧‧‧Use end device

M‧‧‧Video clip

S100~S600‧‧‧Step

S40~S51‧‧‧Step

Figure 1 is a network transmission system architecture diagram of a conventional segmented streaming method; Figure 2A is a flowchart of the detection method of the present invention; Figure 2B is a schematic diagram of the system architecture of the detection method of the present invention; Figures 2C-1 and 2C-2 are diagrams of the data content of the audio and video delivery records of the detection method of the present invention; Figure 2D is a diagram of the data contents of the two playback events of the detection method of the present invention; Figure 2E Figure 2D is the analysis result of the number of times of poor quality of Figure 2D; Figure 2F is a graph of the analysis results of the six playback events of the detection method of the present invention and the number of times of poor quality; 2G-1, 2G-2 and Figure 2G-3 is a chart of the database of the detection method of the present invention; Figures 3A-1 and 3A-2 are packet information tables of playback events of the detection method of the present invention; Figure 3B is a detection of the present invention The graph of the bit rate and packet sequence number of the method; Figure 4 is a flow chart of the analysis method of the present invention for analyzing whether the bit rate is updated; Figure 5 is a flow chart of another embodiment of the detection method of the present invention Figure.

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but are not used to limit the present invention.

It should be noted that the structure, ratio, size, etc. shown in the drawings of this specification are only used to match the contents disclosed in the specification, for those who are familiar with this skill to understand and read, not to limit the creation of this creation. The limited conditions, so it does not have the technical significance, any modification of structure, change of proportional relationship or adjustment of size, should not fall within the original without affecting the effect and the purpose of this creation. The technical content revealed by the creation must be within the scope. At the same time, the terms such as "上" and "一" quoted in this manual are only for the convenience of description, not to limit the scope of this creation can be implemented, and the relative relationship is changed or adjusted. Substantial changes in technical content should also be regarded as the scope of this creation.

Figures 2A and 2B are schematic diagrams of the method for detecting network audio and video delivery quality of the present invention. As shown in FIG. 2A, the detection method includes the following steps S100-S400.

In step S100, the video and audio delivery records within a predetermined time are collected. Specifically, as shown in FIGS. 2A and 2B, a network server room established by the system management end is used to install a plurality of delivery servers 103 (such as Edge type) therein, and a content server from an audio and video supply end 101 transmits at least one audio-visual information 102 to each of the delivery servers 103, and then transmits the audio-visual information 102 through the delivery server 103 to the user device that issues the demand command 9a, 9b, 9c, 9d (such as personal computers, smart phones, or other electronic devices, etc.), at this time, the audio and video of the delivery server 103 through an electrical or communication connection (such as wireless or wired) The quality measuring device 104 collects the audio and video delivery records of each delivery server 103. For example, the audio and video information 102 is a plurality of audio and video clips divided from an audio and video file, and the content size of the audio and video clips can be in units of seconds (such as audio and video clips in seconds) and subunits (such as audio and video clips in minutes) Or other time-length audio and video content.

In step S200, the playback event is integrated. Specifically, the video and audio delivery records of each delivery server 103 collected by the video and audio quality measuring device 104 are processed to form a plurality of packets, and then the packets are used to sort out the required playback events. In this embodiment, the packet defines a list of playback events, as shown in Figures 2C-1 and 2C-2, which lists a plurality of audio and video delivery records (such as numbers 1 to 13), and the audio and video delivery record system Contains the source address (IP) of the user device 9a-9d or the demand side (Subnet), the name of the video clip viewed during the time period (such as M3U8 file), the playback device of the demand side, the transmission speed (such as TS) and the The target address (IP) of the delivery server 103, etc., is used to obtain at least one playback event in the subsequent sorting and statistical sorting operation. For example, the first playback event (10 groups, numbers 1, 2, 4, 6, 7, 9-13) contains the source address [17.12.53.74], the playback device [IOS], and the name of the video being watched [Videol.m3u8] , And the second playback event (3 groups, number 3,5,8) includes the source address [210.44.73.189], the playback device [PC] and the name of the movie to watch [Moviel.m3u8], as shown in Figure 2D.

In step S300, the video and audio quality measuring device 104 is used to analyze the number of times the video and audio clips received at the demand end are of poor quality. Specifically, single A playback event has the number of total network transmission requests (as shown in Figure 2D), and the feature of changing the bit rate (BitRate) through audio and video delivery can count the number of times that these audio and video clips are poor (such as subsequent 3A and 3B), so in this step S300, all playback events on the playback event list will be processed and completed. For example, the second playback event shown in Figure 2D in the TS field in Figures 2C-1 and 2C-2 is 5M, and there is no problem of poor transmission quality, so the number of poor quality and the rate of poor quality Both are 0, as shown in Figure 2E; on the other hand, the TS field shown in Figure 2D in Figures 2C-1 and 2C-2 is 1M, 5M, and 8M. The problem of poor transmission quality occurs, so the number of poor quality requirements shown in Figure 2E is 3 and the poor rate is 30.0%. Therefore, by changing the bit rate caused by the quality of each audio and video delivery, to identify the number of times of poor quality in the audio and video delivery requirements.

In step S400, the video quality measuring device 104 outputs a statistical report. Specifically, according to the attributes of each field of the playback events, an analysis data is formed by integrating and sorting for the poor times. As shown in FIG. 2F, the audio and video quality measurer 104 knows the impact on the receiving of audio and video by multiple playback events and the information in the database (as shown in FIGS. 2G-1 to 2G-3). The reasons for the quality of the clips, such as which computer room (as shown in Figure 2G-1), which delivery server (as shown in Figure 2G-2), and which channel of the audio and video supply side (such as 2G-3 (Pictured) or other factors.

Figures 3A-1 and 3A-2 are the aforementioned packet information tables of the first playback event, and Figure 3B is a diagram of the relationship between the bit rate and the packet sequence number (or network transmission requirements) of the present invention to determine Transmission quality, as shown in Figure 3B The three times shown are of poor quality (such as packet sequence numbers 5 to 7).

In this embodiment, based on the operating mode of most of the existing players, according to the HLS (HTTP Live Streaming) protocol, the user device 9a-9d downloads and obtains multiple audio and video clips (the audio and video clips) from the delivery server 103 It can be M3U8 file), the player or playback software of the user device 9a-9d will request the video clip with the appropriate bit rate according to the device specifications and its operating system, but before the video clip has never been played, it is generally played In the initial stage, the device will first play the audio and video clips at the highest bit rate to evaluate the current network bandwidth, device specifications and operating system of the user device 9a-9d, and subsequent audio and video clips will be requested for suitability. Play at the current bit rate.

Specifically, if the first bit rate drop occurs (as shown in the packet sequence numbers 2 to 3 in Figure 3A-1), it will be recorded as the player of the device 9a-9d adapting to the current environment The changes made will not be recorded as the reason for the poor audio and video delivery quality, so the first bit rate drop will not be included in the effective network demand for the audio and video delivery quality. Then, in the packet sequence number 5 again, the bit rate drop occurred again, this time is not the first bit rate drop, so the AV quality measurer 104 records the status as "poor quality" and starts Accumulate the number of times of poor quality network transmission requirements until the packet sequence number 7 is accumulated a total of three times. After that, in the packet sequence number 8, a bit rate increase occurs, and the audio-visual quality measurer 104 will change the state to "normal" to stop accumulating the number of times of poor quality network transmission requirements.

Furthermore, the detection method further includes a threshold value adjusted by a custom state, which is Based on whether the percentage of the difference between the bit rate of the previous rise or fall exceeds this threshold, the adjustment of the status stage is decided. Specifically, as shown in FIG. 3B, if the threshold value is 30%, the change point of the second bit rate will be 5M to 1M (that is, the packet sequence number 4 to 5), and the difference between the two is 4M. That is, the difference between the two is 80% (which is greater than 30%), and the audio-visual quality measurer 104 will adjust the status information. Similarly, the third bit rate change point is 1M to 8M (that is, packet sequence numbers 7 to 8), the difference between the two is 7M, that is, the difference between the two is 700% (which is greater than 30%), Therefore, the video quality measuring device 104 will also adjust the status information. On the other hand, if the percentage difference between the two is less than or equal to 30% (if the bit rate change point is 5M to 4M, the difference between the two is 1M, that is, the percentage of the difference between the two is 20%), then the audio-visual quality The measuring device 104 does not adjust the status information.

Therefore, as shown in FIG. 4, an analysis flow chart of the video quality measuring device 104 to determine whether the bit rate needs to be updated is organized.

In this embodiment, for a single playback event, the collected audio and video delivery records (network transmission demand records) will be sorted and analyzed first (step S40). When the demand record is entered (step S41), Will find the record of its bit rate (step S42), if there is already a record, it will judge whether the video delivery is of poor quality according to the bit rate discrimination method shown in Figure 3B (step S43) To determine whether the current status changes to normal or poor. If it is judged that the video and audio delivery is in a state of poor quality, it is judged whether the bit rate of this time has increased from the previous bit rate (step S44), that is, whether the bit rate is increased according to the current bit rate, so If it is, it means that the state of poor quality has ended. Change to normal, to get rid of the state of poor quality (step S45), and if not (that is, if there is no increase in the bit rate from the previous time), the number of times of poor quality network transmission requirements will be accumulated (step S46) And update the current bit rate (step S47).

In contrast, in step S43, if there is no current state of poor quality, it is judged whether the bit rate has decreased (step S48), so if not, the process proceeds to step S47 to update the current bit rate, and if yes, it means Knowing the network transmission requirements of poor quality, the operation of step S49 is further executed to determine whether it is the first time that a drop has occurred. Therefore, in step S49, if not, it is necessary to accumulate the number of times of poor quality to proceed to step S47 to update the current bit rate; if so, proceed directly to step S47 to update the current bit rate.

On the other hand, in step S42, if there is no record of any bit rate for the video delivery, the current bit rate can be directly updated (step S47).

Finally, after step S47, it will go to step S50 to determine the end of the playback event, that is, to detect whether it is the last network transmission requirement of the playback event, so when the playback event ends (step S51), Obtain and output a statistical report of the analysis of this playback event (including the number of times of poor quality network transmission requirements) (as shown in Figure 2F). If it is determined in step S50 that the play event has not ended, then return to the initial step S40 to continue the detection of the play event.

As shown in FIG. 5, the statistical report of step S400 described in FIG. 2A can be calculated according to a custom-selected output time period (such as step S500) to create a correspondence table (such as step S600). Specifically, when the system administrator enters When performing line measurement, before step S400 shown in Figure 2A, select the output time segment according to the information of step S300 (as shown in Figures 2C to 2E), and then look up the pre-established correspondence table according to the packet record ( (Such as the database in Figures 2G-1 to 2G-3) to generate the report information to be aggregated, for example, based on the corresponding source IP, M3U8 file type, device used by the player, delivery server 103, and obtained through the analysis process Knowing the number of poor quality, the number of requests for network transmission (Request) and the rate of poor quality, etc., to create a correspondence table, so that the system administrator can quickly grasp the status of the delivery quality of the audiovisual information 102.

In summary, the system administrator of the network operator can easily know the network transmission status of the user device 9a, 9b, 9c, 9d through the output of the statistical report, so the user device 9a, 9b can be known, Factors of unstable audio and video quality presented by 9c, 9d, such as the network operator’s computer room, the network operator’s delivery server, the audio and video supply channel, the system specifications of the user device 9a, 9b, 9c, 9d or Other factors, which can greatly reduce the cost of inspection, can further quickly and/or effectively learn the reason for the poor quality of network audio and video delivery, and is different from the previous method of using Hash (hashing method) to identify anonymous URL URLs. The detection method can further improve its cache efficiency through effectiveness analysis and regression judgment.

The above embodiments are used to exemplify the principles and effects of the present invention, rather than to limit the present invention. Anyone who is familiar with this skill can modify the above embodiments without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the rights of the present invention should be as listed in the scope of patent application mentioned later.

S100~S400‧‧‧Step

Claims (6)

A detection method includes: collecting audio and video delivery records, wherein the audio and video delivery records are the status of a content server of a supply end transmitting at least one audio and video information to at least one user device via at least one delivery server, and the audio and video The information is a plurality of audio and video segments divided from an audio and video file; integrate the audio and video delivery record into a playback event; and analyze the number of times of poor quality in the playback event, and output analysis based on the playback event for the number of times of poor quality Data to generate a corresponding statistical report, where the playback event has the total number of network transmission requirements, and uses the audio and video delivery record to change the bit rate to count the number of times that the audio and video clips are of poor quality; where, by The frequency of the poor quality of the audio-visual clips is analyzed through performance analysis and regression judgment, and the statistical report is generated by comparison to know the reason for the poor quality of the audio-video clips. The detection method as described in item 1 of the patent application scope further includes collecting an audio and video delivery record of each of the delivery servers by an audio or video quality measuring device connected to the delivery server by an electrical or communication. The detection method as described in item 1 of the patent application scope, wherein the playback event includes the specification and source address of the device at the user end, the name of the audio-visual clip, the transmission speed at which the delivery server transmits the audio-visual information, and/or Or the target address of the delivery server. The detection method as described in item 1 of the patent application scope, wherein the number of times of poor quality is transmitted by the delivery server every time the audiovisual information is transmitted A change in bit rate caused by the transmission quality of a video delivery record to identify the number of times of poor quality in the total network transmission requirements. The detection method as described in item 1 of the patent application scope, including including the playback event and the data content of at least one database to learn the reason for the poor quality, wherein the data content of the database corresponds to Play event data content. The detection method as described in item 1 of the patent application scope, wherein the number of times of poor quality in the playback event excludes or does not include the first decrease in the bit rate.

Images