WO2016008913A1

WO2016008913A1 - Method of changing channel and gateway for digital tv streams

Info

Publication number: WO2016008913A1
Application number: PCT/EP2015/066137
Authority: WO
Inventors: Jan Rosseel
Original assignee: Softathome
Priority date: 2014-07-15
Filing date: 2015-07-15
Publication date: 2016-01-21
Also published as: FR3024005A1; FR3024005B1

Abstract

The present invention relates to a method of changing channel in a gateway connected to a plurality of audio-video stream consumer modules and managing a plurality of channels delivered by a platform, this method comprising: - a step of subscribing to at least one channel to be displayed and to n additional channels placed in circular buffer memories, - a step of changing in which the gateway: - returns to the subscribing step if the next channel is not one of the channels placed in memory; - commences a continuous broadcasting of the next channel placed in memory if this next channel is one of the channels placed in memory; the continuous broadcasting being carried out on the basis of the corresponding circular buffer memory either commencing with an optimal point, or by reading in an accelerated manner for a predetermined duration, selects another set of n channels to be prestored in buffer memory based on said next channel.

Description

Channel change method and gateway for digital TV streams

The present invention relates to a channel change method and a digital TV stream system that reduces the time of channel change in a context of parallel reception of audio-video streams by multiple receiving stations within a digital home.

In particular, the invention relates to the context of viewing digital TV channels or other forms of live or retransmitted audiovisual content. In analog TV channels, viewing a channel involves tuning to an analog channel, and then displaying the images is instantaneous.

With the advent of TV and digital media, this has changed. While digital TV has many advantages, one of the major drawbacks is the increased time it typically takes before displaying the first image of a channel that has been selected. Multiple factors contribute to this prolonged time, but the main factor is the time it takes to receive the reference image.

In digital TV streams, there is a sequence of digital images. These images can not be transmitted as raw pixel data, as this would require too much bandwidth. For this reason, these images are compressed. However, compressing individual images does not lead to significant compression. In digital video streams, the property that successive images are most similar is used. Backgrounds typically remain rather static while a foreground object moves in rather regular ways. This has given rise to compression techniques that deduce most of the images from adjacent images (previous or later) instead of transmitting only individually compressed images. To obtain sufficient quality, however, at certain intervals, a complete image is transmitted for reference. Such an image is called a frame I called "I-Frame" in English for "Intra-coded Frame". When a module needs to display such a compressed video stream, it must wait for such an I-Frame to arrive so that it has a reference picture from which other frames that the module receives can be rebuilt. This implies that in the worst case, one must wait for the interval between two I-Frame before being able to restart the reconstruction of a video stream. For highly compressed streams, this duration can be a few seconds. In digital cable TV, some offers address the above problem in the consumer device (Set-Top-Box). These devices use multiple tuners to adjust to different strings (typically the current string, and the previous one and the next one). All these channels are then analyzed by video playback equipment in parallel. When a user then zaps for the neighboring channel, the video displayed on the screen only needs to switch from a stream being interpreted to another stream already being processed.

US 2010 328 527 discloses a method for outputting a first digital video stream. The method includes: i) receiving a first transport stream comprising a first plurality of time multiplexed video channels; ii) storing the first transport stream to obtain a first transport stream stored in memory; iii) selecting a first specific channel of the first plurality of time multiplexed video channels to obtain a first selected video channel; iv) searching for the first transport stream stored for a first start of a first group of images of the first selected video channel, and v) outputting the first transport stream stored in memory starting with the first beginning. of the first group of images to obtain the first digital video stream. In this document, the beginning of the group of images of the selected channel is available in the buffer memory almost instantaneously.

EP 1 775 953 discloses a method of manipulating a plurality of digital video streams, each of the digital video streams being encoded into frame sequences. Each of the frame sequences comprises a key frame, at least one reference frame and additional frames, said keyframe and at least one reference frame suitable for deducing frame information for said additional frames. The method disclosed in EP 1 775 953 comprises displaying at least one first digital video stream to be displayed and buffering at least one of said key frame or said at least one frame of reference for a mobile window of at least a second digital video stream for extracting moving images from the keyframe or buffered reference frame (s) when the second digital video stream is selected to be displayed .

An object of the invention is to provide a method and a system for fast zapping between TV channels which solve the disadvantages of the solutions of the prior art.

The purpose of the present invention is to switch from an audio-video stream, such as a TV channel, to another audio-video stream with reduced latency to improve the quality of delivery of end users.

According to the present invention, the above-defined object is achieved with a chain-changing method in a gateway connected to a plurality of audio-video stream consumer modules and managing a plurality of channels provided by a platform, this method comprising:

a subscription step during which the gateway subscribes to at least one display channel and to n additional strings, said additional strings being chosen from predetermined parameters and respectively placed in memory in circular buffers,

a step of change in response to a next chain change instruction, in which step the gateway:

- returns to the subscription step if the next string is not one of the strings placed in memory;

- starts a streaming broadcast of the next string stored in memory if this next string is one of the strings placed in memory; the streaming being effected from the corresponding circular buffer either by starting at an optimum point or by reading in an accelerated manner for a predetermined time,

the gateway selects another set of n strings to be pre-stored in buffer memory on the basis of said next string. The setpoint comes from one of the consumer modules connected to the gateway.

The gateway according to the invention advantageously uses a circular buffer, one for each string to be buffered, so that data can be immediately reached and read. According to the invention, a circular buffer memory is a buffer memory for storing data in the first-in, first-out manner (FIFO for "first-in, first-out"). The buffer is limited in size, and when the end of writing the data is reached, the following data is stored at the beginning of the buffer, and writing continues from there.

The present invention makes it possible in particular to optimize the zapping time of multicast distributed streams by eliminating the time required for the subscription to the selected stream. Indeed, when a client requests access to a stream that is already stored by the gateway, the multicast subscription is no longer necessary (it has already been realized). Thus, from the point of view of the user, this subscription time has been won.

The gateway according to the invention may be an intermediate functional module receiving signals from a platform such as a module generating and supplying simultaneously or not more consumer modules such as a video reading module or any streaming reader. audio-video. The present invention makes it possible to optimize not only the time of zapping but also the load of the home network. To do this, we use a single gateway per household instead of having one for each consumer module.

"N" can be fixed or variable. The method according to the invention allows a reduction of zapping time thanks to the gateway which buffers the data of the channels to which the video playback module is likely to switch. Thus, the gateway can start broadcasting continuously from an optimal point, or can play the stream stored in accelerated memory, allowing the video player to find an optimal point faster than it could in the live stream from the platform.

The gateway is composed of hardware and software configured to implement the method according to the invention. The hardware includes a microprocessor capable of immediately identifying an optimal point in a buffer memory, and accelerating video in the buffer if the optimal point is not identified (recognized). In this case, it may be encrypted strings, for example, for which the optimal point is not identifiable: the method according to the invention can therefore operate without encrypted string decryption means.

Advantageously, the n additional chains are chosen from a plurality of chains. If the next string is one of these n additional strings, then the zapping time will be less than the typical zapping time. According to the invention, the n additional chains can be chosen by providing the following chains that an end user can choose, by presetting these chains and buffering the stream of the chain. When an end user actually adjusts to one of the predicted strings, the gateway can play the stream from the buffer, either by starting at an optimal point or by sending the buffered stream accelerated to the consumer module (client module). The present invention thus uses different techniques:

- techniques for predicting the following strings that could be selected,

techniques for recognizing optimal points or for reading buffered streams in an accelerated manner.

Preferably, the optimal point may be an I-frame or group of pictures, the streaming starting at the optimum point when such an optimal point is identifiable in the buffer memory. The so-called I-frame and "image group" (GOP) form reference points in a stream from which a reading can immediately begin. This means that there is no timeout for an I frame to appear in the stream of a string before playback begins. This reading method is also ideal in that it requires no change in the clients reading the media.

It is understood that an I-frame or I-frame is a reference frame in a compressed video stream that can be decoded without the need for information from other adjacent images in the stream. The group of images (GOP) is a group of successive images in a coded video stream beginning with an I frame.

According to one embodiment of the invention, the streaming is performed in an accelerated manner when the optimal point is not identifiable. Preferably, in this case, the size of the corresponding buffer is smaller than the size of a jitter buffer contained in a consumer module connected to the gateway. Indeed, in some cases when the content is encrypted, the gateway may not be able to determine the location of a frame I or a beginning of GOP. In these cases, the gateway has no alternative to start reading the stream from the starting point in the circular buffer. Doing this at the rate at which the packets were received would only lead to a delayed read of the stream, but not to an accelerated zapping.

To counteract this, one can use a technique in which the buffer is played in an accelerated manner according to the invention. For example, if the buffer holds m seconds of data, it can read in m / 1 seconds (with 1 "1). Thus, the first frame I in the buffer memory will be seen 1 time faster than in a scenario without pre-memorization. To perform this work, the client modules must have a jitter buffer of at least m seconds, and the jitter buffer should be regulated to not discard packets until the jitter buffer contents significantly exceed equivalent of m seconds of data. If this is not done, there may be macroblocks during the read when the jitter buffer is overflowing, or try to reduce the length of the stored data.

According to an advantageous embodiment of the present invention, the n additional chains can be determined from a chain numbering plan. Advantageously, the n additional chains may be neighboring chains.

According to a preferred embodiment of the invention, the n additional chains can be determined by statistical analysis of selected strings in the history. Statistical analysis allows a prediction of a set of strings that have a high level of probability to be selected for the next channel change. This is a clever way to anticipate the chain change. The gateway could learn user preferences by memorizing which strings users typically select after another string. Each time a user zaps, a counter for the newly selected channel is incremented in a "next string" list linked to the previous channel.

When a user then selects a channel, the gateway can pre-select the n highest count entries in the "Next Channels" list of the newly selected channel.

Additional fine-tuning is possible in this approach, for example, you could remove older entries from the following channel lists. It would be necessary to remember time stamps of the moment when the zapping occurred. This makes it possible to adapt user behaviors that can change over time.

According to an advantageous embodiment of the present invention, the n additional chains can be determined from a statistical analysis of selected channels per consumer module in the history. Thus, we can keep the following lists of channels per client module. This can be useful in the case where a certain module is used predominantly by a certain member of a family while another module is mainly used by another member.

According to another advantageous embodiment of the present invention, the n additional chains can be determined by statistical analysis of chains selected by user in the history. For example, one can keep "next string" lists per user. This allows for the most advanced behaviors based on the preferences / behaviors of individuals rather than on overall behavior influenced by the whole family. To do this, we can consider a mechanism that can identify which user controls a given consumer module at a certain time. This can be accomplished in a number of ways including, but not limited to, logon to the gateway and time lists of who connects to a certain module at a given time. Each user can be identified by login access to the gateway.

The method according to the invention is particularly remarkable, but not only, in that the n additional channels are judiciously chosen taking into account the presence of several consumer modules.

Advantageously, in the case of several consumer modules, for example connected at the same time to the gateway, the n additional channels can be determined from priority definitions between different consumer modules and / or end users. For example, if each consumer module has a list of following strings, the list that will be considered is the list of the consumer module with the highest priority level.

According to a preferred embodiment of the invention, the n additional strings can be determined such that the total bit rates of strings to be buffered are below a predetermined threshold. Indeed, as the chains can be more or less encrypted or can provide a different quality level (high or low definition), the number of additional channels to buffer is limited by the total bit rates to allow smooth processing of the channels by the gateway.

According to the invention, the n additional channels can also be determined taking into account the content when this content meets certain predetermined criteria relating to the user and / or the consumer module. The content may be a program or an upcoming program identified for example by access to electronic guides of local or remote programs. For example, the criteria may be a set of elements to conclude that the content corresponds to the focus of the profile of the user and / or the consumer module.

According to the invention, the n additional channels can also be determined taking into account the current audience of the channels. Indeed, a channel with a large overall audience is more likely to be retained by the user. According to the invention, the n additional channels can also be determined by taking into account the elapsed time for a program being broadcast in relation to the total duration of this program. Indeed, the probability that a program that started long enough is retained by the user decreases. It may be useful to replace this flow with another, more likely.

In a second aspect, the present invention relates to a gateway comprising a storage medium non-transient computer readable encoded with computer executable instructions which, when executed, effectively perform a chain change method as described above.

In order to illustrate the invention, a preferred form is shown in the figures; however, the present invention is not limited to precise arrangements and components.

Figure 1 is a schematic view illustrating an arrangement according to the present invention wherein an intermediate functional module is disposed between a producer module and a consumer module;

Figure 2 is a schematic view illustrating a set-top box as an intermediate functional module connected to a television as a consumer module;

FIG. 3 is a schematic view illustrating a subscription step of an intermediate functional module according to the invention;

Figures 4a and 4b are simplified views illustrating buffer memories within an intermediate functional module during a chain change; and

FIG. 5 is a schematic view illustrating an example of processing of several consumer modules by an intermediate functional module. While the invention is subject to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will be described here in detail. It should be understood, however, that the drawings and the detailed description are not intended to limit the invention to the particular form disclosed, but on the contrary, the invention must cover all modifications, equivalents and alternatives within the scope of the present invention as defined by the appended claims.

One of the problems solved in the present invention is well known and described for example in US 2010 328 527 A1.

A well-known disadvantage of digital television receivers compared to analogue television receivers is that the time it takes to switch from one channel to another is much longer in a digital television receiver. It is an object of the present invention to provide a method and system for decreasing the time of channel change in digital television receivers by means of transport stream buffering. In systems where the transport stream originates from a tuner / demodulator, at least one buffer per tuner / demodulator is required. A tuner / demodulator in a digital television receiver produces a transport stream that contains one or more television channels or services. A service is selected and then decoded. Decoding the video stream of a particular service can begin at the beginning of a group of images. The waiting time for the start of a group of images is an important part of the time it takes to go from one channel to another, for example the DVB / MPEG2 broadcast television in Europe typically uses a group length images of 12 frames, equivalent to 12/25 seconds. For ATSC / MPEG2 television broadcasting in the United States, the image group length is typically 15 frames, equivalent to 15/30 seconds, so that the group start delay time of images can be worth up to half a second. If the chain hop is towards a new service in the same transport stream or to a service in another transport stream that is currently being received by another tuner, then it is still necessary to wait for the start of a group of images before decoding can begin.

The solution disclosed in US 2010 328 527 A1 is to buffer the entire transport stream that includes a plurality of time-multiplexed video channels.

The method according to the present invention aims to reduce the "zapping" time by means of optimization means deployed in a functional module located between the consumer module the video streams (the client) and the producer module of the TV streams (" the platform ") . This is achieved by using methods to provide a set of following strings that an end user could choose by presetting these strings and buffering the stream. When an end user actually sets one of the expected strings, the gateway can read the stream from the buffer, either by starting at an I frame or by sending the buffered stream to the client module in an accelerated manner. .

According to a preferred embodiment, the gateway according to the invention is an intermediate functional module generally designated 1 in FIG. 1. The intermediate functional module 1 is a hardware device or a software method / function or a mixture of the two that found between a platform that is a module 2 producing / streaming media content and another module 3 consuming such content. FIG. 2 illustrates an advantageous embodiment of the invention in which the intermediate functional module 1 is a set-top box equipped with hardware and software means for implementing the method according to the invention. It is connected to a TV 3 which is capable of reproducing an audio-video TV transmission in the form of television signals. The user controls the set-top box 1 via an infrared remote control 4. The channels on the television 3 pass through the set-top box 1 and can come from any type of transmission medium such as IPTV (ADSL for Asymmetric Digital Subscriber Line "via telephone line or fiber optics), satellite, cable or digital terrestrial TV by wireless communication.

The method according to the invention can be implemented not in a set-top box but in an integrated manner within a TV or a player such as a DVD recorder connected to a TV. In a digitally encoded video stream, the successive images contain an I frame that is an entire still image that does not require any additional information to be reconstructed. In a conventional module, a digitally encoded video stream can only be read starting from an I frame. When "settling" such a live stream, the module must wait until to encounter such a frame I before you can display something on the screen. This increases the time between the moment the user presses the button to reach the stream, and the flow display. Depending on the types of video encoding used, this time delta may be substantial, and this degrades the end-user Quality of Experience (QOE). According to the invention, in FIG. 3, a client subscribes to the intermediate functional module 1 (gateway) to a live media stream. The intermediate functional module 1 for its part subscribes to the corresponding flow 5 on the producer module 2 (platform), the stream 5 then being sent to the consumer module 3. According to the invention, the intermediate functional module subscribes to only to stream 5 on producer module 2, but must also subscribe to up to n additional streams 6 on the platform. Said additional flows 6 are processed inside the intermediate functional module 1 while waiting for a change of channels. These n chains are chosen or planned taking into account fixed and / or variable parameters. Once determined, the n additional channels are buffered in small circular buffers on the intermediate functional module 1.

When the customer subsequently intends to reach another stream, two cases are possible:

the stream is not one of the pre-stored stream.

In this case, the intermediate functional module 1 goes back to a subscription step directly to the producer module 2;

- This is one of the stored streams. In this case, the intermediate functional module 1 will perform the following steps:

the intermediate functional module 1 begins the streaming of the pre-stored stream towards the client, while it continues to receive the flow of the producer module 2, and continues to buffer this stream in the circular buffer. Continuous streaming from the circular buffer starts at an optimum point, or will be read in an accelerated manner for a period of time; the intermediate functional module 1 selects another set of n streams for pre-storage based on the selected channels.

FIGS. 4a and 4b illustrate a change between the chain 4 and the chain 3. The intermediate functional module 1 receives the stream currently viewed from the chain 4 in conjunction with three other chains: the chain 7, the chain 3 and the chain 6. Thus , four streams are respectively buffered within four buffers bl to b4. The channel 4 is processed by the buffer memory b3 which supplies the consumer module 3. When the user uses his remote control to adjust to the channel 3 pre-stored in the buffer memory b2, the intermediate functional module 1 goes to the buffer memory b2, reaches a frame I in order to immediately supply the consumer module 3 with the string 3 as described in FIG. 4b. Thus, the channel currently watched is processed by the buffer memory b2. Another set of n additional strings is determined and respectively applied to the buffers b1 (string 8), b3 (string 4) and b4 (string 5).

Various methods can be used to detect which streams need to be buffered, and how many can be pre-stored. In general, one will pre-memorize flows whose use / visualization is most likely in a certain situation. The probability of a certain string being used can be calculated in different ways by taking into account several factors. Some mechanisms and factors that can be used are described below. Pre-selection based on a service plan. In the case where a numbered service plan is available, it will be possible to select neighboring flows. Namely: if the channel 5 is selected, one can for example pre-select the channels 4 and 6. These channels are typically chosen if a user zaps through a service offering using the P- / P + zapping buttons. from the remote control.

Preselection based on the past behavior of the user.

The intermediate functional module 1 can learn the preferences of a user by memorizing which strings the user typically selects after another string. Each time a user zaps, a counter is incremented for the new channel selected in a "next string" list related to the previous channel.

When a user then selects a channel, the intermediate functional module 1 can preselect the n highest count entries in the "next chain" list of the newly selected channel.

Additional fine adjustments are possible in this approach:

· Older entries can be removed from the following channel lists. This will require memorizing timestamps of the moment when the zapping occurred. This allows adaptation to user behaviors that may change over time.

· The following lists of channels can be maintained per client module. This can be useful in the case where a certain module is used predominantly by a certain member of a family while another module is used mainly by another member. Indeed, the intermediate functional module 1 can feed multiple consumer modules as described in Figure 5, as well as multiple producer modules.

• You can keep the following lists of channels per user. This allows for the most advanced behaviors based on the preferences / behaviors of individuals rather than the more general behavior influenced by the entire family. Advantageously, it is possible to envisage a mechanism that can identify which user controls a given consumer module at a certain time. This can be accomplished in several ways, including, but not limited to, logging into the gateway and time lists of who connects to a certain module at a given time. The limits of the preselection.

The amount of strings that can be stored at a certain time can typically be limited by physical constraints. The selection mechanism may need to take this into account. When a user selects a new channel, a new list of channels to pre-memorize needs to be created. Some elements that can be considered to create this new list may for example be:

• the number of streams that are being pre-memorized,

• the bit rates of these flows,

• the number of streams that one wishes to pre-memorize following the selection of new flows,

• technical limitations (number of flows, total bit rate that is received / stored, ...)

• availability of multiple bit rate versions of a given content stream (based on commonly known technologies such as adaptive bit rate streaming), The number of tuners available if the content stream is received from a producer module that is a tuner (satellite, TNT, cable, etc.),

• priority of users viewing channels through the intermediate module. Some client modules may have higher precedence / precedence than others. Similarly, some users who are logged in may have higher priority than others. Higher priority modules or users would have more pre-stored streams than other modules / users to increase their probability of occurrence of a pre-stored stream when they have selected a new stream to watch. The present invention therefore proposes a new method which allows video reading modules to zap faster between frequently-viewed channels, by the use of an intermediate module which pre-stores these channels in memory and allows them to be broadcast continuously. ("Streaming") optimally towards said video playback modules.

Many variations and modifications will occur to those skilled in the art once the above disclosure has been fully appreciated. It is understood that the following claims are interpreted to encompass all such variations and modifications.

Claims

A method of changing a channel in a gateway connected to a plurality of audio-video stream consumer modules and managing a plurality of channels provided by a platform, the method comprising:

a subscription step during which the gateway subscribes to at least one string to be displayed in response to a user instruction and to n additional strings, the said n additional strings being chosen from predetermined parameters and respectively being placed in memory in circular buffers,

a step of changing in response to a change instruction to another channel designated as the next string, in which step the gateway:

- starts a streaming broadcast of the next string stored in memory if this next string is one of the strings placed in memory; the streaming being performed from the corresponding circular buffer or starting at an optimum point when such an optimal point is identifiable in the buffer, this optimum point being a so-called I-Frame or a group GOP ("Group Of Pictures"), or by accelerated reading of the stream stored when the optimal point is not identifiable,

the gateway selects another set of n strings to be pre-stored in buffer memory on the basis of said next string.

2. Method according to claim 1, characterized in that the size of the corresponding buffer is smaller than the size of a jitter buffer contained in a consumer module connected to the gateway.

3. Method according to any one of the preceding claims, characterized in that the n additional chains are determined from a chain numbering plan.

4. Method according to claim 3, characterized in that the n additional chains are neighboring chains.

5. Method according to any one of claims 1 to 2, characterized in that the n additional chains are determined from a statistical analysis of strings selected in the history.

6. Method according to claim 5, characterized in that the n additional chains are determined from a statistical analysis of selected channels per consumer module in the history.

7. The method of claim 5 or 6, characterized in that the n additional channels are determined from a statistical analysis of channels selected by user in the history.

8. Method according to claim 7, characterized in that each user is identified by a login access to the gateway.

9. Method according to any one of the preceding claims, characterized in that the n additional chains are determined from priority definitions between different consumer modules and / or end users.

10. Method according to any one of the preceding claims, characterized in that the n additional strings are determined such that the total bit rates of strings to be placed in memory are below a predetermined threshold.

11. Method according to any one of the preceding claims, characterized in that the n additional channels are determined taking into account the content when this content meets certain predetermined criteria relating to the user and / or the consumer module.

12. Method according to any one of the preceding claims, characterized in that the n additional channels are determined taking into account the current audience of the channels.

13. A method according to any one of the preceding claims, characterized in that the n additional channels are determined by taking into account the time elapsed for a program being broadcast in relation to the total duration of this issue.

A gateway connected to a plurality of audio-video stream consumer modules and managing a plurality of channels provided by a platform, said gateway comprising a non-transient computer readable storage medium encoded with instructions computer executable which, when executed, efficiently performs a chain change process comprising:

a change stage in response to a change instruction to another channel designated as the next string, in which step the gateway: