TW201542014A - Method for obtaining network information by a client terminal configured for receiving a multimedia content divided into segments - Google Patents

Abstract

Client terminal (CT) configured for receiving a multimedia content divided into segments and provided by at least one remote server (SE), each segment being available in one or more representations, comprising a communication module (2) configured for receiving a network information comprising an ordered list of caches (DANE) along the path between the server (SE) and the client terminal (CT).

Description

The user terminal is configured to receive the multi-segment split multimedia content to obtain the network Information method

The present invention relates generally to, but not limited to, adaptive streaming techniques over HTTP (Hypertext Transfer Protocol), and more particularly to a method in which a user terminal is configured to receive multi-segment split multimedia content for network information.

This paragraph is intended to introduce a variety of different aspects of the subject matter of the present invention, which may be related to the various aspects of the invention described below and/or claimed. Therefore, it should be understood that such statements should be read from this point of view and are not intended to be recognized as prior art.

Adaptive streaming over HTTP is quickly becoming the primary technology for multimedia content distribution. Among the HTTP adaptive streaming protocols that have been used, the most famous one is from Apple's HTTP Instant Streaming (HLS) source. Silver Smooth Streaming (SSS) from Microsoft Corporation, Adobe Dynamic Streaming (ADS) from Adobe, and 3GPP (3rd Generation Mobile Communications Partner Program) developed by the SA4 Working Group Dynamic Adaptive Streaming (DASH) over HTTP.

When a user terminal wants to adaptively stream an audiovisual content (or A/V content), it must first obtain a file describing how the A/V content can be obtained. This step is usually done by an HTTP protocol from a URL (a global information addresser, ie a web address), but by other means (eg broadcast, Email, SMS (short message service), etc. to achieve. The tool property specification basically lists the available representations of the A/V content (in terms of bit rate, resolution, and other characteristics) that are pre-generated and transmitted to the remote server, for example, by a remote server. User terminal.

In fact, the corresponding A/V content can be obtained on an HTTP server. The different quality streams, the highest quality is associated with a high bit rate, and the lowest quality is associated with a low bit rate, which allows for the allocation of many different terminals that may be subject to highly diverse network conditions.

The entire data stream is divided into several segments, so that multiple segments are made so that one user terminal can smoothly switch from one quality level to another quality level between the two segments. As a result, the video quality will be different during playback, but few The interruption (also known as freezing) is damaged.

Depending on the agreement, the nature of the tool can be presented in a variety of different formats. The HLS protocol for Apple Computer is an M3U8 playlist, called the "main playlist". Each element of this playlist column is another playlist column, one for each representation. According to other agreements (such as DASH), the tool property specification file consists of one or more XML files, describing all representations one after the other. In any case, generating a tool property profile is as simple as generating a text file and writing the text based on a deterministic grammar.

It is well known that a user terminal selects the best representation at a known point in time based on its available bandwidth, so that the network variation is optimally balanced between quality (such as video quality) and stability. The available bandwidth is dynamically determined at each receive segment. In practice, the round-trip time defined between issuing a HTTP request for a known segment and receiving a corresponding HTTP response (hereinafter referred to as HTTP RTT) is typically measured. And used to estimate the available bandwidth along the transmission path.

This is often the case when a cache memory is set along a transmission path between a user terminal and a remote server, if another user has previously requested the same segment with the same representation, or if a content is delivered The network (CDN) has provided the segment in the cache memory, and the segment can already be stored in the cache memory.

Therefore, the response to an HTTP request for the known segment is faster than if the segment was from a remote server. Because the transmission path is short, the HTTP RTT of the HTTP request is compared between the user terminal and the cache. The user terminal is much smaller than the remote server.

In addition, if there is a cache memory along the transmission path (the requested segment is stored in the cache memory), the highest rate is better, especially when the cache memory is connected to the remote server. When congestion occurs on the transmission path.

Since a user terminal usually does not recognize the answer sent by a remote server or an intermediate cache, it is in fact observed that the transmission path is switched from the "user terminal to the server" path to the "user terminal to the fast". When taking the memory path, it will change the bandwidth Think of end-to-end network conditions.

Therefore, the bandwidth estimate performed by the user terminal is overestimated and does not accurately reflect the end-to-end transmission path characteristics as expected.

This overestimation often results in a bad experience for the end user. In fact, if the estimated bandwidth is higher than expected, the adaptive streaming user terminal typically requests a segment from a higher quality representation (e.g., a higher bit rate). Therefore, when the representation is changed, the requested fragment has a lower probability of being in a cache memory (by assuming that the cache memory system is played by the previous user terminal at the constant bit rate to play the same multimedia content). full). The download time associated with the requested segment should be longer than expected, causing the requested segment to arrive too late, and the user terminal will switch back to the lower quality representation that may be found again in the cache.

As a result, the user terminal switches back and forth between high quality segments and low quality segments - constantly interrupted by cache misses - which completely compromises the advantage of cache.

In addition, since the HAS user terminal does not perceive the contents of the cache memory, it misses the acceleration capability of the cache memory and the advantage of reducing the network load.

The present invention seeks to remedy at least some of the above disadvantages to enhance the end user's experience quality.

The invention relates to a method for obtaining network information by a user terminal, the user terminal is configured to receive a plurality of segments and multimedia content provided by at least one remote server, and each segment can be used for one or more representations. The method is characterized in that the network information includes an ordered list listing the cache memory along the path between the server and the user terminal.

In an embodiment, the network information may further include a representation list of the segments stored in the cache memory for at least some of the cache memories in the ordered list.

In an embodiment, the user terminal transmits a request to a target cache that belongs to the ordered list, and receives a representation list of the segments stored in the target cache. The representation list is complete. A list of differences between the list or previously stored clips and the currently stored clip.

In an embodiment, the user terminal is stored according to the target cache memory. A representation list of the segments is stored to update the representation list of the segments stored by the cache.

In another embodiment, the network information is provided by a tool property profile received by the user terminal from a server, the tool property specification listing the available representation of the multimedia content at the server.

In addition, the tool property specification file can include an ordered list of cache memory along the path between the server and the user terminal.

In addition, the tool property specification file identifies the representation of the segments stored by the cache memory for each cache memory of the ordered list. The cache access representation list is added to the tool property specification file by the cache memory encountered along the path between the server and the user terminal, and the list can be incrementally constructed.

In addition, the tool property specification can be modified by adding the own identifier to each cache memory encountered along the path between the server and the user terminal to establish the ordered list.

In addition, the cache memory can modify the tool property specification by adding a connection information.

In another aspect, the user terminal can query at least some of the cached memory in the ordered list by using an auxiliary communication path, different from a data path for transmitting data, in order to determine the cached memory to be queried. The representation of the stored fragments of the body.

In another aspect, the network information can be appended to a message received by the user terminal from the server, including an extension header for allowing the cache memory receiving the message to report its presence in the message.

In particular, an arrayed table can be created in the extension header to identify each of the cache memories encountered.

In addition, a connection information can be associated with each cache memory of the ordered list.

In one embodiment, the method includes downloading a plurality of segments from at least one cache memory belonging to the ordered cache memory list.

In one embodiment, the receiving network information is through a network interface similar to the interface used to receive the segment or different from the interface used to receive the segment.

According to various embodiments, the adjustment is applied to receive network information (such as an ordered list) The network interface is Wifi wireless, ADSL (asymmetric digital subscriber line), cable, mobile and/or broadcast (such as DVB (European wireless digital TV transmission standard), ATSC (American wireless digital TV transmission standard) interface.

According to various embodiments, the network interface adapted to receive segments is a Wifi, ADSL, cable, mobile, and/or broadcast (eg, DVB, ATSC) interface.

In an embodiment, the user terminal uses a protocol to receive network information, similar to the protocol used to receive the segment (eg, http, Flute), or each has a different (segment: http, flute; network information: TR69 broadcast forum Or a broadcast agreement, xmpp IETF (instant messaging protocol), DDS OMG (object management group).

In one embodiment, the network information is stored in a random access memory (RAM).

In one embodiment, the segments are stored on a local memory (hard disk, flash memory) and/or decoded by a decoder and/or displayed on a display.

Moreover, the present invention includes a user terminal configured to receive a plurality of segments and multimedia content provided by a remote server, each segment being usable in one or more representations. According to the present invention, the user terminal includes a communication module configured to receive a network message, including an ordered list of cache memories along a path between the server and the user terminal.

In addition, the network information may further include a further representation list of the segments stored in the cache memory for at least some of the cache memories in the ordered list.

In addition, the communication module can be configured to query at least some cache memories in the ordered list by using an auxiliary communication path, different from a data path for transmitting data, in order to determine each of the queried queries. The representation of the fragment stored in the cache.

Moreover, the present invention relates to a method for transmitting network information from a cache memory, the cache memory system configured to deliver a multimedia content segment provided by at least one remote server to a user terminal, each segment being usable One or more representations.

In accordance with the present invention, the network information includes an ordered list of cache memories along a path between a server and the user terminal.

In an embodiment, the cache memory receives an ordered list and updates it to be the cache memory closest to the user terminal, and the closest cache memory system is the closest. Conducive to access speed and the most provincial network resources (such as saving the network resources of uploading multiple pieces of content on the user terminal, and / or the fastest access to multiple pieces of content). Next, the cache memory forwards the updated ordered list down to the user; accordingly, the ordered list is incrementally constructed.

In one embodiment, the cache memory adds a list of region cache representations associated with the cache memory to a received representation list, the received representation list being associated with the received order. The memory list is cached; then, the cache memory forwards the updated representation list to the user; accordingly, the representation list is incrementally constructed.

Moreover, the present invention relates to a cache memory adapted to transmit a network information, the cache memory system configured to deliver at least one multimedia content provided by a remote server to a user terminal, each fragment system Can be used in one or more representations. In accordance with the present invention, the network information includes an ordered list of cache memories along a path between a server and the user terminal.

In an embodiment, the cache memory system (also referred to as network cache memory) is included in a set, including: a home gateway, a public gateway, and a corporate gateway. , hotspots, mobile phones, cars, Internet Service Provider (ISP) network components, and corporate agent hosts.

The invention further relates to a computer program product that can be downloaded and recorded from a communication network on a computer readable medium and/or can be executed by a processor, the computer program product including code instructions for implementing the steps of the above method.

Furthermore, the present invention also relates to a non-transitory computer readable medium, comprising a computer program product recorded thereon and executable by a processor, the computer program product comprising program code instructions for performing the steps of the method.

In the following, several specific aspects are set forth which are to be construed as the scope of the invention. In fact, the invention may encompass various aspects that may not be presented below.

1‧‧‧Connection interface

2‧‧‧Communication module

3‧‧‧Adaptive Streaming Module

4‧‧‧Video Player

5‧‧‧ Processor

6‧‧‧Storage components

7‧‧‧Calculator

8‧‧‧Selection module

9‧‧‧Evaluation module

B‧‧‧Internal busbar

CT‧‧‧user terminal

DANE‧‧‧Smart Cache Memory

M‧‧‧ method

N‧‧‧ network architecture

N1‧‧‧ network

RNE‧‧‧Old cache memory

SE‧‧‧ server

S1‧‧‧ calculation steps

S2‧‧‧Selection steps

S3‧‧‧Download steps

S20‧‧‧ preliminary calculation steps

S21‧‧‧The most frequently cached representation selection step

S22‧‧‧ Target combination download time calculation step

S23‧‧‧Comparative steps

S24‧‧‧ Determination step

S25‧‧‧Alternative notation sequence table creation steps

S26‧‧‧Selection, determination, calculation steps

S27‧‧‧Rejection steps

S28‧‧‧Maintenance steps

S29‧‧‧Additional table creation steps

S4‧‧‧ indicates that there is a sequence listing procedure

S41‧‧‧Select the highest i step

S42‧‧‧ Calculate U(k) and T(k) steps

S241‧‧‧Increase utility function steps

T(k)‧‧‧ download time

U(k)‧‧‧ utility function

The present invention will be described below by way of embodiments and implementation examples (not to be limited), and the review panel can further understand the technical features of the present invention. 1 is a schematic diagram depicting a client-server network architecture in which an embodiment of the present invention may be implemented; FIG. 2 is a block diagram illustrating a user terminal example in accordance with an embodiment of the present invention; A method for downloading a predetermined sequence of multimedia content segments implemented by the user terminal of FIG. 2 is depicted in a flowchart.

In Figures 1 and 2, the blocks shown are purely functional entities that do not necessarily correspond to physically separate entities, that is, they can be developed in software, hardware, or in one or more integrated circuits. , including one or more processors.

Whenever possible, the same reference numerals will be used throughout the drawings to refer to the same or similar parts.

It will be appreciated that the drawings and description of the present invention have been simplified for a clear understanding of the present invention, only the related elements are illustrated, and the typical digital content delivery methods and other elements found in the system are eliminated for clarity, however, Class elements are well known to the art and a detailed discussion of such elements is not provided herein.

According to an embodiment, the present invention is illustrated in relation to an HTTP Adaptive Streaming Protocol (or HAS), and the present invention is of course not limited to this particular environment, and other adaptive streaming protocols may of course be considered and implemented.

As shown in FIG. 1 , a client-server network architecture supported by one or several network N (only one network is shown), wherein the present invention can be implemented, including one or several users. The terminal machine CT, one or more HTTP server SEs, a plurality of smart cache memories DANE, and one or more old cache memories RNE. According to DASH, such servers are also referred to as media origins, which for example produce a media presentation description (or MPD), a so-called tool property manifest. This is the source of content distribution: multimedia content can come from some external entities and be converted to HAS format at the origin of the media.

A smart cache memory DANE is a cache component in network N, configured to understand the delivered HAS content, using MPEG-DASH terminology, and considers the smart cache memory system DASH to detect network components (DANE) .

An old cache memory RNE is a cache component in the network N that does not recognize the type of data transferred through it, or at least it does not understand the HAS aspect. In MPEG-DASH terminology, it is considered to have an old cache. Memory system regular network element (RNE).

The user terminal machine CT desires to obtain a multimedia content from one of the HTP servers SE, the multimedia content being divided into a plurality of segments (also referred to as data blocks), assuming that the multimedia content is available in the server SE differently Notation. Once requested by the user, the HTTP server SE can stream the segments to the user terminal CT using one or more TCP/IP (Transmission Control Protocol/Internet Protocol) connections using an HTTP adaptive streaming protocol.

According to the embodiment shown in FIG. 2, the user terminal CT includes at least: - a connection interface 1 (wired and / or wireless, such as Wi-Fi, Ethernet, etc.), connected to the first network N1; - a communication module 2, comprising a protocol stack for communication to the HTTP server SE, in particular the communication module 2 comprises a TCP/IP stack as is well known in the art, and of course any other type that enables the user terminal CT to communicate to the HTTP server SE. Network and/or communication component; - an adaptive streaming module 3 that receives HTTP streaming multimedia content from an HTTP server SE, the module continuously continually matching the network limit with its own limited bits The frequency rate selects the segment; - a video player 4 adapted to decode and generate the multimedia content; - one or more processors 5 for performing storage in a non-transitory memory of the user terminal CT An application and program; - a storage component 6, such as a transient memory, for buffering the segments received from the HTTP server SE before being transmitted to the video player 4; - an internal bus B for connection These different modules and those skilled in the art are familiar with performing the general All family members terminal functions.

In this embodiment, the user terminal CT is a portable media component, a mobile phone, a tablet or laptop, a television, a set-top box, a game device or an integrated circuit. In fact, the user terminal CT may not Including a complete video player, but only some sub-components such as de-multiplexing and decoding the media content, and relying on an external component to display the decoded content to the end user. In this example, the user terminal CT is an HTTP Adaptive Streaming (HAS) capable video decoder, such as a set-top box.

According to the present invention, each user terminal machine CT is configured to implement a method M for downloading a predetermined sequence of multimedia content segments stored in a remote server SE, and selecting the sequence so that its features best match some quality criteria. ,described as follows.

To implement method M, the user terminal CT preferably has knowledge of the network architecture N.

For this purpose, once the user terminal CT requests a multimedia content, a server SE transmits the tool property specification file to the user terminal through an HTTP response, and the work property specification file includes the multimedia content at the server SE. A list of available representations may also include an intelligent cache memory DANE with an ordered list of bits between the server SE and the user terminal CT. In this example, the server SE is aware of the network architecture (as in a managed network) and can pre-establish this ordered list. The server SE can provide an ordered list that depends on the requesting user terminal CT.

In the ordered list, each of the smart cache memories DANE encountered can be identified by: - a unique identifier; - a connection information, such as one that allows access to the smart cache memory Service access point (as explained below).

When the user terminal machine CT receives the tool property specification file, the ordered list of the smart cache memory DANE (completed with additional connection information as needed) provides an extension of the nature of the tool.

In the ordered list, the smart cache memory DANE is preferably listed by considering its proximity to the user terminal CT: the first element in the ordered list corresponds to the wisdom closest to the user terminal CT Type cache memory.

In a variation, when the server SE is unaware of the network architecture, the ordered list may not be pre-established. In this example, each smart cache DANE checks (due to a dedicated check module) the content type header of the HTTP response of the server SE to identify that the response contains a tool property specification file, and then For example, the corresponding service access point is added to the ordered list by its own service access point to update the tool property specification file.

The two methods for establishing the ordered list can exist at the same time, so that some smart cache memory DANE that the server SE does not know receives the HTTP of the server SE. When responding, you can add yourself to an existing ordered list. Obviously, in this example, the smart cache memory preferably checks its existence in the ordered list before adding some data.

The following is an illustrative (but not limited) example of an ordered list of intelligent cache memory DANE in an MPEG-DASH tool nature specification (truncated):

(The rest of the MPD content is truncated )

In addition to the intelligent cache memory class information of the ordered list, the tool property specification file may indicate the representation of the multimedia content segment stored in each intelligent cache memory DANE of the ordered list, for example, by The representation of each stored multimedia content segment (or a portion thereof) is added to an additional list associated with the corresponding smart cache memory identifier. This additional list may be established by the server SE (if it is a managed network) or by the various smart cache DANEs encountered.

In another variation, to determine the network architecture and to specifically determine the intelligent cache memory hierarchy of the ordered list, the server SE transmits an HTTP response to the user terminal CT, the response including an extension header, The smart cache memory DANE receiving the response can report its presence to the user terminal CT and report to the next cache memory between the cache memory DANE and the user terminal CT. The HTTP extension header can be defined as follows:

This header extension can be provided to the server SE and/or smart cache memory The DANE is a means for indicating the presence of the downstream smart cache DANE and/or to the user terminal CT, such means being attached to any response (especially including a response with the nature of the tool).

The server SE can be attached to an HTTP response using the syntax "X-Smart Cache Memory: Necessary", which is used to force the downstream smart cache DANE to attach its own identifier due to a dedicated modification module. And attach the connection information as needed. Upon receipt of this header, each of the smart cached DANEs pre-considers the ordered list with its identifier and connection information. Advantageously, the first smart cache that receives the HTTP response discards the "necessary" token.

An advantage of this change is that the HTTP extension header can be attached to other response messages (such as further content data transmission). Once the HTTP response including a tool-specific profile is intercepted, the extension header can also be used, but not modified. The nature of the tool is defined.

As an illustrative (but not limiting) example, an HTTP response header can be used to indicate the smart cache memory hierarchy of the ordered cache memory list: X-SmartCache-List: 157.254.235.97:12345, 74.125.226.230:12345 (X-Smart Cache Memory - List: 157.254.235.97:12345, 74.125.226.230:12345)

In still another aspect of the embodiment, the signaling message for querying the smart cache memory DANE can be exchanged through an auxiliary communication channel (or path) that is separate from the primary channel that transmits the data. This auxiliary channel is not related to the main channel of the data. This signaling mechanism is called out-of-band signaling.

A separate mechanism for out-of-band signaling can be developed for the first mechanism, a separate protocol is implemented to target a particular smart cache memory DANE, and for the second mechanism, the HTTP protocol utilizes a second TCP (transport) Control protocol) connection for signaling, thanks to the natural upward selection of HTTP requests (a request sent by the user terminal CT towards the server SE naturally spans the encountered smart cache DANE and The old cache memory RNE can respond at the appropriate time). In any case, these mechanisms do not affect the old cache memory RNE.

The signaling overlap consists of a smart cache memory hierarchy of the ordered list and its identifier, according to a first mechanism that uses a particular protocol. Once the call overlap is obtained, different network devices can communicate with each other. You can use the smart cache interface to exchange cache management jobs. The smart cache interface can be accessed through a smart cache access point. The fetch point is provided as part of the tool's property specification or included in the HTTP response header. Allow DASH/HAS users to motivate the smart cache interface. By constructing a message containing the job identifier and its parameters and transmitting it to the smart cache interface, each job can be motivated.

In an illustrative (but non-limiting) example, the messaging message can be created as a JSON string, which is easily parsed in many server-side manuscript languages. Once the messaging message is received, the smart cache is available. The memory DANE can execute the requested job.

In particular, the messaging message between the user terminal CT and the smart cache memory DANE can be transmitted by a number of suitable protocols.

In a first illustrative example, a WebSocket protocol is used that allows bidirectional peer-to-peer communication between a user terminal CT and a known smart cache memory DANE. In a second illustrative example, XMPP (Instant Messaging Protocol) may be used to allow for point-to-multipoint messaging (eg, cache memory update notifications for all user terminals of a smart cache), Of course, other agreements can be used.

Due to the first mechanism, the user terminal CT can exchange messages with a known smart cache memory DANE. To discover the segments stored in the smart cache DANE and their representations, the following message can be used. : get/sCached([segment/ds: < list > ])

( Get /s cached ([fragment / ds: < list > ]) )

Upon receiving the message, the smart cache DANE can reply to an update table containing the identifiers of the cached segments. The user terminal can transmit this message to any smart cache memory DANE upstream thereof for an accurate overview of the individual content of the individual smart cache.

In addition, a smart cache memory DANE can use a callback to a registered user terminal to notify the user of changes to its cached content, for example, the user terminal has preliminarily specified the segments they want to monitor. Currently, such callback mechanisms are easily implemented using HTML5 websocket technology and server-side events, but other techniques can be used to achieve similar results.

cacheUpdate([segment/ds: < list > ])

( Cache update ([fragment / ds: < list > ]) )

According to the second mechanism, the user terminal CT can also use HTTP as a signaling means to query the cache memory chain (smart cache memory DANE and old cache) Recalling the body RNE) to discover the contents of the cache. These outgoing HTTP messages are transmitted to the same server address in a TCP (Transmission Control Protocol) communication period and follow the same path of the primary data delivery communication period. Clearly establish a send HTTP request to avoid triggering data transfer.

For this purpose, the HTTP HEAD method or a minimum byte range GET (such as requesting a single byte of data) can be used to transfer multiple HEADs (or 1 byte range) for different representations of the intended segment. The GET) request, and the extension of the recommendation, the user terminal CT can establish a location map of the cache representation.

In particular, to query the contents of a cache memory, the user terminal CT can use the "Cache-control" HTTP header with the "only-if-cached" command. To request the content that is closest to the cache. However, compared to the first machine production, the HTTP-based mechanism implicitly allows only the first cache memory on the path towards the server SE to be queried.

As an extension to alleviate the shortcomings of querying only the first cache memory, the "only if cached" guide of the call request may further include a depth indication of the cache memory, which should be considered as: Cache-Control :only-if-cached,depth = 3

( Cache - Control: only if cached, depth = 3 )

The depth value is interpreted by the smart cache memory DANE, so that when the representation of the requested segment is not quickly fetched into a smart cache memory DANE, if the depth value allows, the smart cache memory Then forward the request to the mail. Each time a smart cache memory DANE does not store the requested representation, the depth value is decremented and the HTTP header is modified accordingly before forwarding the request. When a smart cache memory DANE receives a send request having a depth value equal to 1, if the cache miss is not present, the send request is not forwarded.

At a subtle point, the smart cache DANE storing the request representation can include a supplemental guide in its response indicating that the depth is reached, such as: X-SmartCache-Info:depth = 2

( X-Smart Cache Memory - Information: Depth = 2 )

Then, the user terminal machine CT uses the value received from the smart cache memory DANE, and can learn the smart cache memory considered by subtracting the return value from the initial depth value of the call request of the user terminal machine CT. The distance of the body DANE (converted with cache memory).

Alternatively, the user terminal CT may limit the depth of the query to the known cache memory by identifying a known cache memory in its signaling request. When the signaling request reaches the identified smart cache DANE, and the smart cache DANE does not store the requested segment representation, the identified smart cache DANE does not The request is forwarded to the server SE and replied negatively. For this purpose, a specific extension of the target smart cache identifier is added to the cache control of the HTTP send request "Only if cached" guide, such as: Cache-Control: only-if-cached , until =< smartcache_id >

( Cache - Control: Only - if - cached until ~ < Smart Cache Memory_Identifier >)

The smart cache memory identifier used in the signaling request may be any known identifier of the smart SAP cached DANE or any unique identifier of the smart cache memory DANE.

When the HAS user terminal machine CT perceives the smart cache memory DANE set on the path of the server SE, and perceives the cache representation list held by each smart cache memory DANE, the user terminal machine CT can be implemented. Method M.

Assuming that the contents of the smart cache DANE between the user terminal CT and the server SE are known for the predetermined k-segment (defining the k-segment sequence), the method M can determine the k of the segments of the sequence to be downloaded. Representation. The method M is preferably performed periodically after downloading only m segments (1 <= m <= k) or upon receiving a cache memory update message. Such updates allow for further download optimization of segments from m+1 to m+m (using predictions up to m+k).

According to this embodiment, and as shown in FIG. 3, the method for downloading a predetermined k-segment multimedia content sequence at the user terminal machine CT comprises the following steps, the sequence being stored in a server SE and usable in different representations :- Calculate (step S1) the following items for some or all of the available representation combinations of the segments stored (or not stored) in the identified cache memory DANE (as described above), the cache memory system Set between the user terminal CT and the server SE: ■ A value of the quality utility function U(k) is used for each combination of k representations; ■ a prediction time T(k) is used for downloading The combination step S1 is performed by a calculator 7 of the user terminal machine CT shown in FIG. 2, and in a variation, the calculator 7 can be integrally formed with or as part of the processor module 5; - among the decision values of the utility function, selecting (step S2) the highest utility function value having a download time that is inferior to a time threshold (e.g., corresponding to the play time of the sequence of segments), the selection step S2 being A selection module 8 of the user terminal CT is implemented; - at least one initial representation sequence associated with the selected combination is downloaded (step S3) at the user terminal machine CT, upon receipt of information about the selected combination, ie by communication The module 2 and/or the adaptive streaming module 3 manages the download of the selected combination.

At least some of the steps S1 to S3 may of course not be performed in the user terminal machine CT, but in an external network device (such as a server, gateway, proxy host, etc.).

In particular, for a known combination, the utility function U(k) depends on: - the overall quality of the representation of the combination ( k ) ;- a variability σ of the representation of the combination; - the cache of the representation is not the cost M(k).

If the overall quality observed by the end user on the user terminal is proportional to the representation quality in the combination, since the higher bit rate is used to provide higher quality, the sum of the bit rates of the representation can be used, Take ( k ) indicates, and its average is an estimate of this overall quality.

This variability can be expressed, for example, by the number of variations in the value represented by the combination.

The cost of cache miss is the bandwidth cost of impacting network resources, and the cost of transmitting data in the server resources, both of which are proportional to the bit rate of the segment downloaded from the server. Therefore, this cost For example, it can be represented by a sum of bit rates having a segment representation of cache misses.

In an illustrative (but non-limiting) example, a known combined utility function U(k) can be derived from the following formula: among them:- ( k ) is an average bit rate of the representation combination, k is the number of fragments of the sequence; - σ is the number of variations of the representation combination (and thus the instability of the sequence); - α is the variation A weighting parameter; - M(k) is the average cost of the cache miss for the representation combination; - the beta system is a weighted parameter for the average cost of the cache miss.

In particular, the average bit rate of a known combination of representations can be determined by the following formula: The bit rate of a known representation of the Ri-series fragment i, which belongs to the combination.

Furthermore, the variation σ of the known representation combination can be obtained by the following formula: The number of variances σ is increased by both the number of changes between the representations and the size of the change from the average representation.

Further, for example, the cache miss average cost M(k) of the known representation combination is described by the following formula: Wherein when the representation of a segment i is retrieved from the server SE, Si=1, in other cases (the representation is cached in one of the identified smart caches DANE), Si=0 .

By maximizing the utility function U(k) for the sequence, the high bit rate with low variance and very few cache misses will have priority, and the weighting parameters α and/or β can be adjusted to define the variation. Number and / or cache miss tolerance. Obviously, by setting the value of α and/or β to zero, the number of variations and/or cache misses can be ignored.

The utility function U(k) is calculated for each candidate combination considered, in order to determine a combination that meets the following criteria: - high average bit rate (and implies high quality); - stable bit with small variability Meta-rate, which means consistent video quality; - Retrieves the maximum number of segments from the cache to minimize the load on the server SE and network N.

In addition, a combined estimated download time T(k) can be calculated due to the following formula: Where: - the BW _server is the download bandwidth observed when downloading a known representation of the combination from the server; - BW _ci is from the closest intelligent cache memory DANE holding the known representation The download bandwidth observed when downloading a known representation of the segment i of the combination, Ci corresponds to an element of a vector C by taking the index of the closest cache memory holding the representation The various representations used for this combination are established.

In the denominator, since Si is 0 or 1, only one of the items Si or (1-Si) is not equal to zero, so the denominator is linked from the bandwidth BW _{server of the} server SE or from the considered smart type. Take the bandwidth BW _{ci of the} memory DANE. The time required to download each representation is obtained by dividing the number of bit rates for each selected segment representation by the bandwidth BW _server .

A combination of this download time T(k) should preferably be inferior to the playback time of the sequence, otherwise the playback will be interrupted due to buffer bleed, thus: T ( k ) < k . Data block_ period

According to this embodiment, the user terminal machine CT calculates (step S1) the utility function U(k) and the download time T(k) for each available representation combination of the k-segment, r represents (from the server SE) U(k) and T(k) are calculated r ^k times, respectively, using the number of representations.

Thus, by performing step S2, the user terminal machine CT can select the best sequence of k representations to avoid buffer drain.

The following table shows the contents of a group of smart cache memory DANE for illustrative (but not limited) examples for an 8-segment sequence ranging from 1 to 8 (1 is the first segment of the sequence to be downloaded) The value in the grid corresponds to the index of the smart cache memory DANE holding the considered fragment/representation. (kbps=thousands per second)

By considering the following bit rate vector R = {R1 = 2500 kbps, R2 = 2500 kbps, R3 = 2500 kbps, R4 = 2500 kbps, R5 = 2500 kbps, R6 = 2500 kbps, R7 = 2500 kbps, R8 = 2500 kbps}, which corresponds to a constant representation Method 2500 kbps k-segment sequence. The C vector is established by taking an index of the closest smart cache memory DANE holding the representation 2500 kbps for each segment. Using the sample content of the smart cache memory provided in the above table, C={1,1,2,1,0,0,0,1}, 0 means in the identified smart cache memory The representation is not fast enough to represent 2500 kbps.

In a variation of this embodiment, the user terminal machine CT does not calculate the utility function U(k) and the download time T(k) for each available representation combination of the k-segment, but instead initially via an evaluation module 9- The smart cache memory from the user terminal CT and the server SE is used to reserve the content knowledge of the desired k-segment-calculation (in one step S20) to the number of segments to be used for each possible representation, In a further step S21, the most frequently retrieved representation, also called the target representation, is selected.

Based on the above example table, k=8, the selected notation has a representation with a bit rate equal to 2500 kbps, which has 5 cached segments. The bit rate vector R corresponding to the target combination has the same value equal to the target representation R={R1=2500 kbps, R2=2500 kbps, R3=2500 kbps, R4=2500 kbps, R5=2500 kbps, R6=2500 kbps, R7=2500 kbps, R8=2500 kbps The bit rate of }.

In a further step S22, the user terminal machine CT calculates the download time T(k) of the target combination due to its calculator 7.

In a further step S23, the user terminal machine CT compares the calculated download time T(k) (also referred to as the initial T(k)) with the play time.

When the initial T(k) is at least equal to the play time ( T ( k ) k . In the data block_period, for a segment of the smart cache memory in which the target representation is not stored (the fifth, sixth and seventh segments in the example above), the user terminal machine determines (in step S24) An alternative cache representation of the fragments.

For this purpose, the user terminal machine CT establishes an alternative ordered list of alternative representations in step S25, in which the cached higher quality alternative representation is listed in ascending order, followed by a list of caches in descending order A lower quality alternative representation.

In a step S26, the user terminal machine CT: - Select the first representation in the alternative ordered list; - Determine a new combination, similar to the target combination, except that the alternative representation chosen replaces the target representation for the corresponding fragment (according to the example in the table above, the new combination) Vector R system {2500 kbps, 2500 kbps, 2500 kbps, 2500 kbps, 2500 kbps, 4500 kbps, 2500 kbps, 2500 kbps, }); and - Calculate the download time for this new representation combination.

If the T(k) of the new combination is increased by T(k) of the target combination, the alternative representation of the selection is rejected and the new combination is rejected in step S27, and the next representation in the alternative ordered list is repeated. S26, in order to modify the target combination by using the next representation to establish another new combination.

If the T(k) of the new combination is lower than the play time, the user terminal machine CT starts downloading the new combination (step S3).

If the T(k) of the newly combined T(k) related target combination is reduced but still at least equal to the play time, the alternative representation selected in the alternative list remains unchanged (step S28), and the target is replaced by considering the new combination. In combination, step S26 is repeated.

The above example table is known, and this table will be continuously tested with R6=4500 kbps, R5=6500 kbps, R7=6500 kbps, and R5=1200 kbps, R7=1200 kbps, unless an intermediate solution is found.

When all the representations in the alternative ordered list have been tried and if the T(k) for calculating the latest combination is not inferior to the play time, the user terminal machine CT establishes an additional substitute list in step S29. Includes a lower quality representation than the target representation, arranged in descending order. Steps 26, 27 and 28 are applied to the additional ordered list.

At the end of this algorithm, T(k) satisfies this limit, or no lower possible value is used for T(k).

When the initial T(k) is lower than the play time, the user terminal CT tries (in one step S241) to increase the utility function U(k) by using a cached representation with a higher rate, In a first step S4, the user terminal machine CT establishes a representational ordered list greater than the initial target representation in an ascending order.

Then for each representation of this list, the user terminal CT:

- selecting (step S41) the highest i, so that the corresponding segment for considering the representation can be in a cache Obtained in memory.

- Calculate (step S42) U(k) and T(k) by using the considered representation to change the previous Ri for the obtained representation combination.

Then, if T(k) still satisfies the limit and the new value of U(k) is greater than a previous value, the user terminal machine CT keeps the new combination unchanged and repeats step S41 with the next (lower) i. If T(k) no longer satisfies the limit or U(k) is reduced, the user terminal machine CT rejects the new value of Ri to recover the previous combination.

The test of the second highest notation is performed when all indices i have been considered.

Obviously, without departing from this embodiment, other heuristics can be used to satisfy the limit on T(k) while boosting U(k).

In fact, at least some of the steps S20 to S29 may not be performed in the user terminal machine CT, but may be performed in an external network device (such as a server, a gateway, a proxy host, etc.).

Once the desired combination (segment sequence) has been determined, the user terminal machine CT downloads the subsequent segments from the sequence without delay until the m segment is reached (1) m k), or receive an update message from a smart cache, or receive a rate change. In any of the foregoing events, the user terminal CT repeats the method M.

The network information includes an ordered list listing the cache memory DANE along the path between the server SE and the user terminal CT, and the network information includes the cache memory DANE as needed. A representation of the stored segments for at least some of the cached memory in the hierarchy.

Please note that in accordance with various variations of the present invention, receiving the network information is via a network interface, similar to the interface used to receive the segment, or different than the interface used to receive the segment.

According to a particular embodiment, the network interface of the user terminal is adapted to receive network information from a Wifi, ADSL, cable, mobile and/or broadcast (eg, DVB, ATSC) interface.

According to various embodiments, the network interface of the user terminal is adapted to receive segments from a Wifi, ADSL, cable, mobile, and/or broadcast (e.g., DVB, ATSC) interface.

According to various embodiments, the user terminal uses a protocol to receive network information, similar to a protocol (eg, http, Flute) for receiving segments. According to various embodiments, the user terminal uses a different protocol to receive network information: - Fragment transmission and reception can use http or flute protocol; - Network information transmission and reception can use TR69 broadband forum agreement or a broadcast agreement (such as xmpp IETF, DDS OMG (object management group)).

According to various embodiments, the network information is stored in a random access memory (RAM).

According to various embodiments, the segments are stored on a local memory (such as a hard disk or a flash memory) and/or decoded by a decoder and/or displayed on a display.

According to various embodiments, the user terminal belongs to a set comprising: - a portable media device; - a mobile phone; - a gaming device; - a set-top box; - a television set; - a tablet; a laptop; and - an integrated circuit.

The flowchart and/or block diagrams in the figures illustrate the configuration, operation, and functionality of systems, methods, and computer program products that may be implemented in accordance with various embodiments of the present invention. Accordingly, blocks of the flowcharts or block diagrams can represent a module, segment, or portion code that includes one or more executable instructions for performing the specified logical functions. It should also be noted that in some alternative implementations, the functions indicated in the blocks may not occur in the order indicated in the figures. For example, depending on the function involved, the display of consecutive two squares may in fact be performed almost simultaneously, or The blocks may sometimes be executed in the reverse order, or the blocks may be executed in an alternate order. It should also be noted that blocks of the block diagrams and/or flowcharts, and combinations of blocks in the block diagrams and/or flowcharts can be implemented by a special purpose hard disk system that performs the specified function or behavior, or by a special purpose hard disk. Implemented in combination with computer instructions. Although not explicitly illustrated, embodiments of the invention may be utilized in any combination or sub-combination.

As will be appreciated by those skilled in the art, several aspects of the present invention can be embodied as a system, method, or computer readable medium. Thus, several aspects of the present invention can take the form of a complete hardware embodiment, a complete Software embodiment (including firmware, resident software, microcode, etc.), or knot An embodiment of the software and hardware aspects is generally referred to herein as a "circuit," "module," or "system." Moreover, aspects of the invention may take the form of a computer readable storage medium, any combination of one or more computer readable storage media.

A computer readable storage medium can take the form of a computer readable program product embodied in one or more computer readable media, and having computer readable code embodied thereon can be executed by a computer. A computer readable storage medium as used herein is considered to be a non-transitory storage medium that provides an inherent ability to store information therein, as well as an inherent ability to provide information retrieval therefrom. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or component, or any suitable combination of the above. It should be understood that although a more specific example of a computer readable storage medium to which the present invention is applicable is provided below, it is immediately apparent to those skilled in the art that it is merely an example rather than a comprehensive list: portable computer disk; hard disk Random Access Memory (RAM); Read Only Memory (ROM); Erasable Programmable Read Only Memory (EPROM or Flash Memory); Portable Disc Read Only Memory (CD-ROM) An optical storage device; a magnetic storage device; or any suitable combination of the above.

1‧‧‧Connection interface

2‧‧‧Communication module

3‧‧‧Adaptive Streaming Module

4‧‧‧Video Player

5‧‧‧ Processor

6‧‧‧Storage components

7‧‧‧Calculator

8‧‧‧Selection module

9‧‧‧Evaluation module

B‧‧‧Internal busbar

CT‧‧‧user terminal

Claims (6)

A method for obtaining network information by a user terminal (CT) configured to receive a plurality of segments and multimedia content provided by at least one remote server (SE), each segment being usable for Or a plurality of representations, wherein the network information includes an ordered list listing cached memory (DANE) along a path between a server (SE) and the user terminal (CT). A user terminal configured to receive a plurality of segments and multimedia content provided by at least one remote server (SE), each segment being usable in one or more representations, the user terminal comprising a communication module (2) configured to receive network information, including an ordered list, listing cached data (DANE) along a path between the server (SE) and the user terminal (CT). A method for transmitting network information by a cache memory system, the cache memory system configured to transmit a multimedia content segment provided by at least one remote server (SE) to a user terminal, each segment being usable for one or more In the representation, the network information includes an ordered list listing the cache memory (DANE) along the path between a server (SE) and the user terminal (CT). A cache memory adapted to transmit network information, the cache memory system configured to deliver a multimedia content segment provided by at least one remote server (SE) to a user terminal, each segment being usable for one or more In the representation, the network information includes an ordered list listing the cache memory (DANE) along the path between a server (SE) and the user terminal (CT). A computer program product downloadable from a communication network and/or recorded on a computer readable medium and/or executable by a processor, the computer program product including code instructions for implementation as claimed in the patent application Method of 1 or 3 items. A non-transitory computer readable medium, comprising a computer program product recorded thereon and executable by a processor, the computer program product comprising code instructions for implementing the method of claim 1 or 3.