WO2011150643A1

WO2011150643A1 - Method and device for on-line adaptation of media content

Info

Publication number: WO2011150643A1
Application number: PCT/CN2010/079913
Authority: WO
Inventors: 张鹏; 赵华
Original assignee: 华为技术有限公司
Priority date: 2010-12-17
Filing date: 2010-12-17
Publication date: 2011-12-08
Also published as: CN102217278B; CN102217278A

Abstract

A method and device for on-line adaptation of media content are disclosed, wherein the method includes: before a non-first user request arrives, predicting the media content required by the user; obtaining the predicted media content; when the user request arrives, if the media content actually required by the user and the predicted media content are the same, returning the obtained media content to the user. The present invention can reduce the delay of a response to a user request caused by an on-line adaptation processing.

Description

Online media content adaptation method and device

The present invention relates to the field of streaming media technologies, and in particular, to a method and device for online adaptation of media content.

Background technique

With the development of network communication and multimedia technology, the application of streaming media technology has become increasingly widespread. The so-called streaming media refers to the media format that is played on the Internet and the intranet by means of streaming. Through the streaming media technology, the user can provide an online playing service, and the client can watch the online after the multimedia file is completely downloaded, which brings great convenience to the user.

Due to the increasing format of media content, and the different network connections and terminal devices of users (for example, mobile phones, personal PCs, etc.), content providers often need to provide different versions of the same media content to meet user needs and adapt to bandwidth changes. And adapting the media content to the code rate and resolution according to the change of the network bandwidth and the user terminal. A disadvantage of the pre-storage method is that more storage resources are needed to store copies of media content at different bit rates/resolutions, which are costly to store, and have limited adaptability to network bandwidth and changes in terminal devices.

In order to better adapt to different user access network environments, the prior art may use a method of generating media content copies of different code rates/resolutions online. For example, the user may be generated by online adaptation on the edge nodes. Quality media content is required to achieve smooth quality changes.

However, this method of online adaptation requires online adaptation of the media content, and the adaptation process is usually triggered after receiving the user request, and after receiving the media content that meets the requirements, the response is returned to the user. Obviously, the time required for this adaptation process will increase the response delay requested by the user. For example, in a user request, the requested media content is a slice with a duration of 2 seconds and a rate of 512 kbps. In the prior art method, after receiving the user request, first query whether there is a cache in the cache. The user requests the appropriate media content. If not, the corresponding media content needs to be requested from the original server, or the media content in the cache is transcoded, and then the response is returned to the user. In any case, there is a delay in the response to the user request. Summary of the invention

The embodiment of the invention provides a method and a device for online media content adaptation, which can reduce the response delay caused by the online adaptation process to the user request.

The embodiment of the invention provides the following solutions:

An online adaptation method for media content, including:

Predicting the media content the user needs before the non-first user request arrives;

Obtaining the predicted media content required by the user;

When the user request arrives, if the media content actually requested by the user is the same as the media content required by the predicted user, the obtained media content is returned to the user.

A media content online adaptation device, comprising:

a prediction unit, configured to predict a media content required by the user before the non-first user request arrives; and an obtaining unit, configured to acquire the media content required by the predicted user;

And a response unit, configured to: when the user request arrives, return the obtained media content to the user if the media content actually requested by the user is the same as the media content required by the predicted user.

With the embodiment of the present invention, the media content required by the user can be predicted before the user requests, and if the prediction is accurate, the response to the user request can be directly returned to the user. Obviously, the response delay of the user request can be greatly reduced as compared to the manner in which the corresponding media content is prepared after receiving the user request. DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present invention, and other drawings can be obtained from those skilled in the art without any inventive labor.

1 is a flowchart of a method according to an embodiment of the present invention; FIG. 3 is a schematic diagram of a device according to an embodiment of the present invention; 4 is a flowchart of another apparatus according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of an acquiring unit in an apparatus according to an embodiment of the present invention. detailed description

BRIEF DESCRIPTION OF THE DRAWINGS The technical solutions in the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without the creative work are all within the scope of the present invention.

S101: predicting media content required by the user before the non-first user request arrives;

S102: Obtain the predicted media content required by the user;

S103: When the user request arrives, if the media content actually requested by the user is the same as the media content required by the predicted user, the obtained media content is returned to the user.

First of all, it should be noted that in the process of playing the same media file, the client usually needs to send a user request to the media server multiple times, and each time a part of the media file is sent. Among them, a part of the media content requested by the user each time is called a slice of the media file, in other words, the client requests one slice of the media file each time. When the client sends a user request, it carries the information such as the bit rate, code rate, or resolution of the requested fragment. The server needs to return the media content of the corresponding bit rate for the user request.

For the first received user request to play a media file, the server may first prepare the corresponding media content and then return a response to the user. While returning the response, before the second user request arrives, the media content required by the second user request can be predicted and the predicted media content can be obtained, so that when the second user request arrives, The predicted media content parameters (for convenience of description, the following are all introduced by bit rate), compared with the bit rate of the media content carried in the second user request, if the same, the prediction is accurate, And directly using the predicted media content that has been obtained to return the response information for the second user request. While returning the response information for the second user request, the media content requested by the third user request may be predicted and the predicted media content may be obtained, so that after the third user request arrives, The bit rate of the predicted media content can be compared with the bit rate of the media content carried in the third user request. If the same, the predicted media content that has been obtained is directly used to return a request for the third user. Response information, and so on.

In other words, in the embodiment of the present invention, in addition to the user request received for the first time, the media content required by the user may be predicted before the user requests, and if the prediction is accurate, the user may directly return the request for the request. response. Obviously, the response delay of the user request can be greatly reduced as compared to the manner in which the corresponding media content is prepared after receiving the user request.

If the bit rate of the media content predicted by the user request is different from the bit rate of the media content actually carried in the user request, the media actually carried in the request may be re-prepared. The content of the content matches the media content, and then returns a response to the user request, while predicting the media content needed in the next user request.

Here's how to predict the media content your users need. For ease of understanding, the media file and its bit rate are briefly introduced. In the streaming media technology application, the client receives the media content for decoding and playing, and the smoothness of the playback is related to the state of the client (including bandwidth, hardware and software configuration, network link status, etc.) and the bit rate used when transmitting the media content. . For the same bit rate, if the state of the client is different, the fluency of the media files played on the client is different. In other words, from the perspective of the smoothness of playback, the ability of different client states to receive the bit rate is different under the premise of ensuring smooth play. When the actual bit rate exceeds the maximum bit rate that the client's current state can withstand, the client may not play smoothly.

Of course, the bit rate is not as small as possible, because the bit rate reduction is usually at the cost of reducing the amount of information transmitted per unit time, which means for the client, the clarity, resolution or audio of the video. The quality of the playback is declining. It can be seen that if the user selects a suitable bit rate according to the specific conditions of the client and comprehensively considers the smoothness and clarity of the playback, a higher quality playback effect can be obtained for the client.

Therefore, in the embodiment of the present invention, the media content required by the user in the next user request may be predicted according to the current network bandwidth condition and/or the virtual buffer status of the user player. In practical applications, the network bandwidth condition and/or the virtual buffer status of the user player and the bit rate of the media content may be established according to experience or other means. For specific prediction, you can first obtain the current network bandwidth condition and/or the virtual buffer status of the user player, and then according to the advance The established correspondence acquires the bit rate of the corresponding media content, and uses it as the predicted bit rate of the media content required by the user. The current network bandwidth condition and the virtual buffer status of the user player may be reported to the server by the client, or may be obtained through a separate mechanism to obtain the current network bandwidth condition and the virtuality of the user player. Buffer status information.

Of course, the above method for predicting the media content required by the user is not unique. In practical applications, other methods can also be used for prediction. For example, the prediction may be performed according to the current media content of the user and the previously requested media content, that is, the rule of the media content required by the user is obtained from the historical user request, and the media content required by the user in the next user request is predicted by using the rule. .

After predicting the bit rate of the media content desired by the user, the corresponding media content can be prepared in advance. Specifically, when obtaining the predicted media content required by the user, the server may first query the cache of the server. If the user has previously requested the part of the media content, and has not been deleted or has not been replaced by other content, the server caches. This part of the media content is already stored. At this time, it is equivalent to the media content that meets the requirements in the cache. After receiving the user request, if the predicted result is the same as the bit rate of the media content requested by the user, the media content is directly taken out from the cache as a response of the user. Return to the user.

If the query finds that there is no media content matching the predicted media content in the cache, the method may be used to obtain the prediction by using the method requested by the upper server or by transcoding the media content existing in the cache. The media content that the user needs.

The method of requesting the upper-level server is equivalent to making a request to the original server where the media content is located, and the original server stores multiple copies of the same media content, and each copy has the same content but different bits. Rate; When it is predicted that a fragment with a certain bit rate is needed, the fragment of the bit rate can be requested from the original server, and the original server will return the corresponding fragment.

The method of transcoding the media content existing in the cache is equivalent to when the fragment that needs to have a certain bit rate is predicted, the fragment of the bit rate does not exist in the cache, but the other bit rate is used. Fragmentation, at this time, other bit rates can be converted to the required bit rate by transcoding to obtain the desired media content.

What needs to be explained here is the number of shards and the size of the shards for the same media file. It is usually predetermined by the server, and each slice has its own slice identifier. For copies of different bitrates of the same media file, the sharding mode between the individual splicings is the same, and the sharding identities of the shards are also corresponding. For example, a media file has three copies: copy A, copy B, copy C, each of which has M shards, and the shard identifiers for each shard in each copy are 1, 2, 3... ... M, then the content of the same slice in the three copies is the same, that is, when the client plays the slice of the same slice in the three copies (assuming smooth Under the premise of playing), the content that the user sees is the same, but the definition will be different.

When sending a user request, the client carries the fragmentation identifier of the fragment in addition to the bit rate of the fragment; and the fragments of the media file are sequentially arranged, so, for predicting the media required by the next user request In terms of content, only the required bit rate needs to be predicted. For the fragmentation identifier, only the next fragmentation needs to be taken on the basis of the fragmentation identifier in the current user request.

For example, if the current request is a slice with a slice ID of 2, and the slice identifier of the next slice of the slice is 3, the slice identifier requested by the next user should be 3; and the next user is predicted. If the bit rate of the requested fragment is 512 kbps, the server can find the fragment with the bit rate of 512 kbps in the fragment with the fragment identifier of 3 in the cache. If it does not exist, it can request the branch from the upper server. The slice is identified as a slice with a bit rate of 512 kbps, or a slice of another bit rate existing in the buffer is identified as a slice of 3, and is converted into a slice having a bit rate of 512 kbps.

In a specific implementation, the media content that is predicted by the user may be obtained by using any of the foregoing methods. In order to further reduce the response delay to the user request, in the embodiment of the present invention, whether the same content already exists in the cache may be used. However, factors such as fragmentation of different bit rates, transcoding delay, request delay of the upper server request, transcoding load, user level, etc., determine whether to use the request of the upper server or use the cache. The existing media content is transcoded to obtain the predicted media content required by the user.

Specifically, if there is media content in the cache that is higher than the predicted bit rate of the media content required by the user, the user's level is transcoded, and the high bit rate media content existing in the cache is transcoded into the predicted content. If the transcoding delay of the media content required by the user is less than the delay of the request from the upper server, the media content that is predicted by the user is obtained by transcoding the media content existing in the cache; otherwise, the superior is used. The way the server requests, the predicted media content required by the user. For example, the fragmentation identifier of the next fragment of the currently requested fragment is 3, and the bit rate of the media content required by the next user request is predicted according to the network bandwidth condition and the virtual buffer status of the user player. At 512 kbps, the steps to obtain the predicted media content required by the user can be as shown in Figure 2:

S201: determining whether there is a fragmentation identifier in the cache, and a media content having a bit rate of 512 kbps exists, if yes, ending the acquisition process; otherwise, proceeding to step S202;

S202: determining whether there is a fragment identifier in the cache is 3, a media content having a bit rate higher than 512 kbps exists, if yes, proceeding to step S203, otherwise proceeding to step S208;

As mentioned earlier, the reduction in bit rate is usually at the expense of reducing the amount of information transmitted per unit time, that is, low bit rate media content is transmitted per unit time relative to high bit rate media content. The amount of information will be relatively small. Therefore, for the fragmentation of the same content, it is relatively easy to convert from a high bit rate to a low bit rate, as long as some information is deleted from the high bit rate media content; otherwise, if it is from Converting a low bit rate to a high bit rate may mean that operations such as restoration of information are required, which is relatively complicated. Therefore, in order to further reduce the response delay to the user request, it is possible to perform the conversion from the high bit rate to the low bit rate only when there is media content higher than the predicted bit rate in the cache, that is, the buffer is cached. The media content of other bit rate existing in the conversion manner is obtained, and the predicted media content required by the user is obtained. Otherwise, the predicted media content required by the user is obtained in a manner requested from the upper server.

S203: determining whether the user's level can be transcoded, if yes, proceeding to step S204, no, and proceeding to step S208;

S204: Calculate a transcoding delay required for transcoding from a high bit rate to a 512 kbps slice according to an average transcoding delay and a current transcoding load condition;

The current transcoding load refers to the transcoding load of the transcoder. Since the server may provide streaming media services to multiple users at the same time, the transcoder may perform multiple transcoding processes at the same time. When the transcoding load is high, The time it takes to complete a new transcoding task will be longer.

S205: determining whether the transcoding delay is less than the delay required by the upper server request, if not, proceeding to step S206, otherwise proceeding to step S208;

S206: Submit a transcoding task to the transcoder; For example, if the bit rate of the high bit rate media content existing in the cache is lmbps, the transcoding task submitted to the transcoder is: the slice is identified as 3, the media content of the bit rate is lmbps, and the bit rate is transcoded. Media content of 512kbps.

S207: The transcoder writes the transcoded media content to a cache of the server;

S208: The upper server requests the media content whose slice identifier is 3 and the bit rate is 512 kbps. Of course, the above method of selecting media content is not unique, and the system can be used in practical applications. For example, first, it is checked whether the predicted media content has been stored in the cache. If there is no storage, the upper-level server is queried whether the media content is stored, and if it is already stored, it is obtained from the upper-level server; if the predicted media content is not in the upper-level server, For storage, the query caches whether there is suitable media content for transcoding, and whether the transcoder has the ability to perform transcoding, and if the conditions are met, the transcoding is performed.

It should be noted that, for the storage of media content in the cache, the cache management system may determine whether to store the transcoded media content or the media content obtained from the superior content server in the cache according to a certain cache management policy. The specific cache management strategy is not the focus of the present invention, and therefore will not be described again. Provided is a media content online adaptation device, see FIG. 3, the device comprises:

The predicting unit 301 is configured to predict a media content required by the user before the non-first user request arrives;

The obtaining unit 302 is configured to obtain the predicted media content required by the user;

The response unit 303 is configured to: when the user request arrives, return the obtained media content to the user if the media content actually requested by the user is the same as the media content required by the predicted user.

The prediction unit 301 is specifically configured to: predict the media content required by the user according to the network bandwidth condition and/or the virtual buffer status of the user player.

For specific implementation, referring to FIG. 4, the obtaining unit 302 may include:

The query subunit 3021 is configured to query, from the cache, the media content that matches the predicted media content required by the user. The requesting sub-unit 3022 is configured to: when the media content that matches the media content required by the predicted user does not exist in the cache, obtain the media content required by the predicted user by using a manner requested by the upper server;

Or,

The transcoding sub-unit 3023 is configured to: when the media content that matches the media content required by the predicted user does not exist in the cache, use the method of transcoding the media content existing in the cache to obtain the predicted The user needs the media content.

In an actual application, the obtaining unit 302 may further include:

The selecting subunit 3024 is configured to determine, according to whether the same content in the cache but the different bit rate, the fragmentation delay, the request delay requested by the upper server, the transcoding load condition, and/or the user level, determine the acquisition of the prediction. The way the user needs the media content.

Referring to Figure 5, the selection subunit 3024 can include:

The first determining sub-unit 30241 is configured to determine whether there is media content in the cache that is higher than a bit rate of the media content required by the predicted user; if not, the method may be used to obtain the predicted content in a manner requested by the upper server. Media content required by the user;

a second determining sub-unit 30242, configured to determine, if the determination result of the first determining sub-unit is yes, whether the level of the user can be transcoded; if not, the method can be used to obtain the predicted result by using the request of the upper-level server Media content required by the user;

a third determining sub-unit 30243, configured to: if the determination result of the second determining sub-unit is yes, determine that the high-bit rate media content existing in the cache is transcoded into the predicted media required by the user Whether the transcoding delay of the content is less than the delay of the request from the upper server; if not less than, the media content required by the predicted user may be obtained by using the request of the upper server;

The transcoding determination sub-unit 30244 is configured to: if the determination result of the third determining sub-unit is YES, obtain the predicted user by using the media content that is stored in the cache to be transcoded The media content needed.

With the embodiment of the present invention, the media content required by the user can be predicted before the non-first-time user request arrives, and if the prediction is accurate, the response to the user request can be directly returned to the user. Obviously, the response delay of the user request can be greatly reduced as compared to the manner in which the corresponding media content is prepared after receiving the user request.

A person skilled in the art can understand that all or part of the steps of implementing the above embodiments can be completed by a program to instruct related hardware, and the program can be stored in a computer readable storage medium, and the program is executed. The method includes the following steps: predicting media content required by the user before the non-first user request arrives; obtaining the predicted media content required by the user; and when the user request arrives, if the user actually requests the media content and the location If the predicted media content required by the user is the same, the obtained media content is returned to the user. The storage medium is as follows:

ROM/RAM, disk, CD, etc.

The above description is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present invention. It should be considered as the scope of protection of the present invention.

Claims

Rights request

A method for online adaptation of media content, comprising:

Obtaining the predicted media content required by the user;

2. The method according to claim 1, wherein the predicting media content required by the user comprises:

The media content desired by the user is predicted based on network bandwidth conditions and/or virtual buffer conditions of the user player.

The method according to claim 1, wherein the obtaining the media content required by the predicted user comprises:

The media content matching the predicted media content required by the user is queried from the cache.

The method according to claim 3, wherein if the media content that matches the predicted media content required by the user does not exist in the cache, the obtaining the media required by the predicted user The content also includes:

Obtaining, by means of a request from a superior server, the media content required by the predicted user; or

The media content required by the predicted user is obtained by transcoding the media content existing in the cache.

The method according to claim 4, wherein: the request delay, the transcoding load condition and/or the user level requested by the level server determine a manner of obtaining the predicted media content required by the user.

The method according to claim 5, wherein the determining the manner of obtaining the predicted media content required by the user comprises:

Determining whether there is media content in the cache that is higher than a bit rate of the predicted media content required by the user;

If yes, determine whether the user's rating can be transcoded; If so, determining whether the transcoding delay from the high bit rate media content present in the cache to the predicted media content of the user is less than the delay requested by the upper server; if less than, And the method of transcoding the media content existing in the cache to obtain the predicted media content required by the user.

7. A media content online adaptation device, comprising:

The device according to claim 7, wherein the predicting unit is specifically configured to: predict media content required by the user according to a network bandwidth condition and/or a virtual buffer condition of the user player.

The device according to claim 7, wherein the obtaining unit comprises: a query subunit, configured to query, from the cache, the media content that matches the predicted media content required by the user.

The device according to claim 9, wherein the obtaining unit comprises: a requesting subunit, configured to: when there is no media content in the cache that matches the media content required by the predicted user, Obtaining the predicted media content required by the user by using a request from the upper server;

Or,

a transcoding sub-unit, configured to: when the media content that matches the media content required by the predicted user does not exist in the cache, use the method of transcoding the media content existing in the cache to obtain the prediction The media content that the user needs.

The device according to claim 10, wherein the obtaining unit further comprises: a slice, a transcoding delay, a request delay requested by the upper server, a transcoding load condition, and/or a user level, determining to obtain the The way in which the user's desired media content is predicted.

12. The apparatus according to claim 11, wherein the selecting subunit comprises: a first determining subunit, configured to determine whether there is media content in the cache that is higher than a bit rate of the media content required by the predicted user;

a second determining subunit, configured to determine, if the determination result of the first determining subunit is YES, whether the level of the user can be transcoded;

a third determining subunit, configured to: if the determination result of the second determining subunit is YES, determine to transcode the high bit rate media content existing in the cache into the predicted media content required by the user Whether the transcoding delay is less than the delay requested by the upper server;

a transcoding determining subunit, configured to: if the judgment result of the third judging subunit is YES, use the method of transcoding the media content existing in the cache to obtain the predicted user requirement Media content.