FIELD OF THE INVENTION

The invention relates to a method of making a plurality of digital content titles available to a plurality of stations through a network. The invention further relates to a system for making a plurality of digital content titles available to a plurality of stations through a network.

BACKGROUND OF THE INVENTION

Traditionally, content such as audio and video was broadcast to receiving stations using broadcast systems, such as terrestrial broadcasting, cable broadcasting or satellite broadcasting systems. Increasingly, digital content is also distributed via wide area networks, in particular the Internet. The distribution via a network may, for example, involve a video-on-demand service or audio/video (A/V) such as for example provided by the RealOne Player and the Windows Media player. The network may be a centralized distribution network, a CDN-based distribution system, or a peer-to-peer (P2P) network. In the centralized distribution approach, a centralized server is deployed to support client stations to access digital content, such as multimedia content. The client retrieves the content from the server through the network. The data may pass through several routers and through a server of an access provider. The routers/servers may perform caching of data to provide quicker access to content that has already been retrieved earlier from the centralized server. A cache system includes a cache memory that is usually faster than the background storage of the server. For example, the cache may be formed by RAM and the background storage by a hard disk array or tape array. Since cache memory is more expensive, the cache only includes a subset of the set of titles that can be retrieved by stations. A cache controller controls which titles are in the cache. To this end a so-called caching strategy is employed, such as "Least Recently Used", "First IN, First Out", etc., as described in more detail in, e.g., US pat. app. ser. no. 09/374,694 (attorney docket PHA 23,737), herein incorporated by reference, filed for Chanda Dharap for SEMANTIC CACHING and published under PCT as WO 01/13265. This document relates to caching resources based on the semantic type of the resource. The cache management strategy is customized for each semantic type, using different caching policies for different semantic

types. Semantic types that can be expected to contain dynamic information, such as news and weather, employ an active caching policy wherein the resource in the cache memory is chosen for replacement based on the duration of time that the resource has been in cache memory. Conversely, semantic types that can be expected to contain static resources, such as encyclopedic information, employ a more conservative caching strategy, such as LRU (Last Recently Used) and LFU (Least Frequently Used) that is substantially independent of the time duration that the resource remains in cache memory. Additionally, some semantic types, such as communicated e-mail messages, newsgroup messages, and so on, may employ a caching policy that is a combination of multiple strategies, wherein the resource progresses from an active cache with a dynamic caching policy to a more static caches with increasing less dynamic caching policies. The relationship between semantic content type and caching policy to be associated with the type can be determined in advance, or may be determined directly by the user, or could be based, at least partly, on user-history and profiling of user- interaction with the resources. In the Content Distribution Network (CDN) approach, such as for example deployed by Akamai, a large number of servers are deployed at the edge of the network. A centralized server distributes part of the content to the edge servers that can provide this content quickly to the client stations. To this end, the edge servers are located near the end users. Thus part of the content is replicated on the edge servers. The replication is performed under control of a replication strategy that determines which title should be replicated on which edge servers. Criteria can be load and network proximity. Distribution of emails through a CDN is described in, e.g., US pat. app. ser. no. 09/900,561 (attorney docket US 000171), incorporated herein by reference, filed for Peter Kaars for EMAIL DISTRIBUTION ON THE EDGE and published under PCT as WO 03/032599. This patent document relates to a service provider who enables to distribute electronic content via a data network to a plurality of addressees. In order to do this, respective servers are identified, via which to deliver the content to respective groups among the plurality of the addressees. A copy of the content is then sent to the respective server. The respective server is enabled to send individual copies of the content to individual ones of the addressees in the respective group, e.g., based on the relevant e-mail addresses of the addressees in the respective group. Accordingly, this describes distribution of digital content through a CDN by executing a replication strategy that is driven by delivery requests of user that have been issued by the user before the content is actually available.

In a peer-to-peer (P2P) distribution approach, each peer station in the P2P network contributes its resources, such as storage, to the network. A peer station is usually also a client station that has once obtained content titles (e.g., through the network) with the intention of rendering it. Examples of P2P distribution systems are Napster, CentreSpan, Vtrails and Kazaa. A P2P network allows sharing of content by a large community at very low cost, as dedicate servers are not needed and the maintenance cost of the network is quite low. To improve the performance of the known P2P networks replication of titles is essential. A replication strategy can determine how many replicas of a title are needed and where to place those titles. Replication can accelerate delivery of content and improve robustness. Most replication and caching strategies favor titles that are in high demand. A certain latency in adapting to the actual demand for a title is then inevitable as long as the title has not been distributed long enough by the network to have gained sufficient knowledge of the demand for the title. The approach of WO 03/032599 is suitable for titles where the predicted demand based on user requests before the title is actually available indeed represents the actual demand during the distribution.

For completeness, also see US ser. pat. app. no. 09/971,474 (attorney docket US 018166) filed for Eugene Shteyn and Jean Moonen for DIGITAL CONTENT CATERING SERVICE, herein incorporated by reference and published as US20030069964. This document relates to catering of electronic content information via a network. A consumer is enabled to request delivery via the network of a specific piece of content before a deadline selectable by the consumer. The catering service then completes the delivery before the deadline under control of a bandwidth profile of data traffic on at least a segment of the network. The scheduling of delivery deadline enables the service to optimize usage of the network's bandwidth while maintaining a quality of service.

SUMMARY OF THE INVENTION

It is one of the objects of the invention to provide and improved method and system of the kind set forth wherein the latency in making a title available sufficiently is reduced. The invention provides a method of making a plurality of (preferably digital) content titles (hereinafter referred to as "network titles") available to a plurality of stations through a network. The method includes obtaining for each network title a corresponding measure of an audience share during a broadcast of a corresponding title (hereinafter referred

to as "broadcast title"); and controlling the availability of the network titles in dependence on the respective measures.

The inventor has realized that many multimedia titles have already been broadcast in one way or another before being made available through the network. For broadcasts typically an audience share is measured. This audience share is, for example, used to determine an advertisement rate for advertisement broadcast via the same channel before, after or during the broadcast of the title. Using this measure to pro-actively control the availability of the title in the network instead of adapting the availability as the demand becomes clearer can significantly reduce the latency in adapting to the right demand level. A title that has proven to be very popular during a broadcast may be given a high availability in the network.

TV audience shares are usually measured by national broadcasters associations or other organizations using a number of sample households. Increasingly, the shares are measured in real-time (e.g., every minute) and available 'instantly' at the measuring organization. Such a measuring organization is, for example, the Broadcasters Audience

Research Board (BARB) in the UK. According to the measure of the dependent claim 2, the audience share measures are received through a communication system (e.g., the Internet) from a measuring system that measures for each title broadcast through a broadcasting system a corresponding audience share. In this way, the measured share can quickly be used to control the availability of the title in the network. This is particularly useful for titles that are first broadcast and soon after the broadcast are made available for downloading. Examples of such titles are news broadcasts. By using the actually measured share instead of a predicted share, the availability can 'instantly' be adjusted to an unexpected situation. For example, if an important news event has occurred the news broadcast will most likely be highly popular for downloading.

According to the measure of the dependent claim 3, an association is established between a network title and a broadcast title. This association may be that the titles are identical (e.g., in the case of a news broadcast that is also available for downloading). For movies, the titles need not be fully identical. For example, the broadcast title may be shorter than the network title. The content does not even need to be the same but may be related. For example, if in a broadcast programme a movie is discussed and a preview shown and this programme gets a relatively high share, this can be used as an indication that also the actual movie will be popular for downloading.

According to the measure of the dependent claim 4, the audience share measure influences the caching strategy in the server, for example a high share means: keep copy of the title in the cache, a low share means: not in the cache.

According to the measure of the dependent claims 5 and 6, the measure influences the replication strategy. For example, a high share results in a high degree of replication, a low share means a low degree of replication.

According to the measure of the dependent claim 8, priority is also given to recently broadcast titles. In particular, broadcast titles with a high share caused by a special event (e.g., exceptional sporting results, accidents, etc.) will be particularly popular for downloading in a relatively short period after the broadcast.

According to the measure of the dependent claim 9, a peer station in a P2P network can restrict access to network titles stored in the peer station that have a corresponding measure above a predetermined threshold. In this way, a peer station can control that it is not flooded by demands for downloading to the extent that the performance for local task is adversely affected.

To meet an object of the invention, a system for making a plurality of digital content titles (hereinafter referred to as "network titles") available to a plurality of stations through a network includes a distribution controller for obtaining for each network title a corresponding measure of an audience share during a broadcast of a corresponding title (hereinafter referred to as "broadcast title"); and for controlling the availability of the network titles in dependence on the respective measures.

To meet an object of the invention, a server for use in the system includes a background storage for storing the network titles; a cache for providing faster access to a subset of the stored network titles; and a cache controller for executing a caching strategy in dependence on the respective measures of the network titles stored in the storage.

To meet an object of the invention, a station for use in the system, wherein the network includes a peer-to-peer distribution system with a plurality of stations, includes a storage for storing a plurality of network titles and a controller for restricting access to network titles stored in the storage that have a corresponding measure above a predetermined threshold.

The approach of the invention, the availability based on audience share, can be combined with the approach of US20030069964 mentioned above in order to allocate bandwidth under combined control of measured audience share and deadlines requested by customers. For example, a high audience share for a particular piece of content could be used

to offer a scheme of deadlines to choose from so as to further optimize bandwidth in the system of US20030069964.

BREIF DESCRIPTION OF THE DRAWING These and other aspects of the invention are apparent from and will be elucidated, by way of example and with reference to the accompanying drawing wherein:

Fig. 1 is a block diagram of a centralized distribution system according to the invention;

Fig. 2 is a block diagram of a CDN according to the invention; and Fig. 3 is a block diagram of a P2P distribution system according to the invention.

DETAILED EMBODIMENTS

Fig. 1 is a block diagram of a system 100 according to the invention and its relationship to other systems. The system 100 makes a plurality of digital content titles (hereinafter referred to as "network titles") available to a plurality of stations (shown are stations 170, 172, and 174) through a network 160. The network is preferably a wide area network, in particular Internet. The system 100 includes a distribution controller 102. The distribution controller 102 obtains for each network title a corresponding measure of an audience share during a broadcast of a corresponding title (hereinafter referred to as

"broadcast title"). In an embodiment of the invention, the distribution controller 102 obtains the audience share from a share measuring system 150. Such a measuring system 150 in itself is already known for measuring TV audience shares. These shares are usually measured by national broadcasters associations or other organizations using a number of sample households that have received the broadcast. In the example of Fig. 1, a title is broadcast by a broadcasting system to a plurality of broadcast receivers, shown are receivers 130, 132 and 134. The broadcast may occur using any suitable broadcasting technique, including terrestrial broadcast, satellite broadcast, broadcasting via cable systems, or web-casting through the Internet. A receiver can typically select from receive many simultaneous broadcast. The measuring system 150 measures which title being broadcast has been selected and/or is being rendered by the receivers 130, 132, 134. Based on this data, the measuring system 150 determines the audience share. Such a calculation is well-known and will not be described in detail here. Usually, the share is only measured for a subset of the total numbers of receivers 130, 132, 134. The broadcast receivers may provide the required information to the

measuring system 150 in any suitable way. Traditionally, telephone links have been used for this. More recently, systems use fully automatic, regular updating of the information to the measuring system 150 via a wide area network 152, in particular though the Internet. From the data supplied by the receivers 130, 132, 134, the measuring system 150 compiles the shares of the titles being broadcasts. In a preferred embodiment, the measuring system communicates the share to a distribution controller 102 in the system 100 via a communication system (e.g., Internet or a direct connection). Alternatively, the system 100 may collect the share data directly from the receivers 130, 132, 134 and as such perform the role of the measuring system 150. The share data may provide only the relative popularity of the programs (e.g., a percentage of the potential number of viewers that watched the program). Usually, the share data also includes the absolute number of viewers of each broadcast program. Both types of information can in principle be used as input for the system according to the invention, where the absolute number of viewers provides most information. It will be appreciated that more than one measuring system 150 may be used, e.g., one for each country or one for each broadcast provider (e.g., BBC, Sky, etc.). If multiple measuring systems are used, preferably, the audience share represents the absolute number of viewers of a program in order to be able to compare data of the different measuring systems.

In this description a program broadcast by, for example, the broadcasting system 120 is referred to as a broadcast title. It may be any type of title, e.g., a live broadcast of a news program, a pre-recorded TV show, or a movie. The content may also be of any type, e.g., audio, audio and video, or other forms of multimedia. A title may be identified in any suitable way, e.g., using a name and/or identifying number. For a movie, the title identifier may also include information like the distributor, director, etc. For a live broadcast, the title identifier may also include a channel name, like BBCl, CNN, MTV, FoxKids, etc. The broadcast title may at the moment of broadcasting already have been finished and stored in a storage system 122. It may also be a live broadcast and recorded in a storage system, like storage system 122. A title that is associated with a broadcast title is made available to the stations 170, 172, and 174 through the network 160. Titles that can be received (e.g., downloaded) through the network 160 are referred to as network titles.

The audience shares are in principle available for the corresponding respective broadcast titles. In a preferred embodiment, the system according to the invention can establish an association between a broadcast title and a network title. In the situation where a title is first broadcast and then in an unmodified form is made available through the network

160, this association is trivial: the titles are identical. In particular, for movies, the titles need not be fully identical. For example, the broadcast title may be shorter than the network title. In this case, preferably automatically an association is established by comparing identifying information of the broadcast title and the network title (e.g., the movie name, director, etc. will be the same, even if the content is not exactly the same). The content of the broadcast title and the network title may still be related, even if one of the titles is not a full subset of the other one. For example, in a broadcast programme a movie may be discussed and a preview shown. In such a situation, a human operator may be used to establish a link between the broadcast title and a network title. To this end, the system 100 may have a user interface (not shown) for enabling a human operator to establish the link. Preferably, automatic techniques, like fingerprinting are used to establish the link. Such techniques are known, for example for detecting for a small broadcast audio and/or video segment from which music track or movie the segment comes from.

Once the distribution controller 102 has information for a plurality of network titles on the respective associated audience shares, the distribution controller 102 uses this information for controlling the availability of the network titles in the network 160. The availability may be controlled in any suitable way. Much effort on developing mechanisms for controlling availability of video titles through a network has taken place for video-on- demand and near video-on-demand systems. Such techniques may be applied in the system according to the invention. In particular, caching and/or replications strategies in the network may be controlled by the distribution controller 102.

Fig. 1 shows one embodiment of the distribution of the network titles through the network. In this so-called centralized distribution approach, a centralized server is deployed to support the client stations 170, 172, 174 to access the digital content. In this case, preferably the distribution controller and the server are located in the same apparatus 100 (or directly connected). The clients 170, 172, 174 retrieve the content from the server 100 through the network 160. The data may pass through several routers and through a server of an access provider (not shown). The routers and/or access provider servers in the network may also perform caching of data to provide quicker access to content that has already been retrieved earlier from the centralized server. This is a type of re-active caching resulting in higher availability for titles that have already proven to be popular in the network (i.e. have already been downloaded frequently). Such re-active caching is not part if the invention, but may be used in combination with the invention. In the system according to the invention, the distribution controller 102 uses the audience share information to pro-actively control the

availability of the network titles in the network 160. In the centralized approach, the server 100 includes a background storage 104 for storing the network titles. The background storage 104 may be of any suitable type. In particular storage as is known from video-on-demand systems may be used, such as disk array systems. The availability of the network titles may be controlled by effecting the bandwidth from the background storage 104 to the network 160. Many techniques are known for this, such as storing the most popular title on the outermost sectors of a hard disk, using a higher degree of striping for most popular titles, etc. Any such techniques may be used. In a preferred embodiment, the server 100 includes a cache 106 for providing faster access to a subset of the stored network titles. A cache controller 108 of the server 100 executes a caching strategy in dependence on the respective measures of the network titles stored in the storage. For example, if the cache 106 has a capacity for storing one hundred network titles, the cache controller 108 may control that one hundred network titles stored in storage 104 that have the highest estimated popularity, based on the audience shares of the associated broadcast titles, are copied into the cache 106. Each time a new network title is stored in the storage 104, the cache controller 108 checks whether the title's popularity is in the top one hundred. If so, the least popular title in the cache 106 is removed and the new title is copied to the cache 106. This caching strategy may be combined with other suitable caching strategies. In particular, priority may be given to new network titles or titles recently broadcast. Also, titles that had a high audience share but are relatively 'old' in the system and no longer popular may be removed from the cache (or more general: the availability is reduced). It will be appreciated that the system according to the invention aims to achieve an initial optimal availability of a new title and any suitable strategy may be used to adjust the availability later on.

Fig. 2 shows an alternative distribution network, wherein a plurality of servers is used. In particular, edge servers may be used as known from the Content Distribution

Network (CDN) approach. A centralized server 200 distributes part of the content to the edge servers 210, 212, 214 that can provide this content quickly to the client stations. The edge servers are preferably located near the client stations. As described above, part of the content is replicated on the edge servers 210, 212, 214. The replication is performed under control of a replication strategy that determines which title should be replicated on which edge server(s). Criteria can be load and network proximity. The centralized server 200 may mostly be the same as described for server 100 of Fig. 1. Same numbers are used to indicate unchanged elements. In this embodiment, the distribution controller 202 controls the replication over the edge servers 210, 212, 214 in dependence on the audience shares. For

example, a network title with a high share is pro-actively distributed over a high percentage of the edge servers (i.e. very near to the stations 170 to 178), whereas a title with a lower popularity is replicated over less edge servers (possibly more remote from the stations 170 to 178). The edge servers may be integrated with access provider servers, as is illustrated for the edger server 210 and 212. For these edge servers the stations 170, 172 and 174 have direct access to the content without having to go through the network 160. An edge server may also be located nearer to the station (e.g., being accessible through less routers) without being an access provider server. In the embodiment of Fig. 2, the edge server 214 is such a server. This server 214 is then accessible by the stations through a wide area network 220, such as the Internet. In principle, any station on the network 220 might be able to access the edge server 214, but it is intended that only stations in its locality (e.g., stations 176 and 178) use the server 214. In the embodiment of Fig. 2, the distribution controller 202 ensures that the most popular network titles are pro-actively replicated from the storage 104 of the source server 200 on the respective storages (not shown) of the edge servers 210, 212, 214. Preferably, the distribution controller 202 pro-actively pushes the popular content to the edge servers 210, 212, 214. The edge servers 210, 212, 214 will typically provide the content to the respective stations 170, 172, 174, 176, 178 on demand of the individual stations. It will be appreciated that this supply to the stations 170, 172, 174, 176, 178 may be automated as well (e.g., on a subscription basis, possible using profiles to filter, etc.). It will also be appreciated that the edge servers 210, 212, 214 may also use the audience share information, for example for improving the availability of its locally stored content to the respective stations. The edge servers 210, 212, 214 may do this in the same way as described for the central server 100 of Fig. 1, for example by using a cache and a cache controller (not shown in Fig. 2).

Fig. 3 shows another alternative distribution network, using the so-called peer- to-peer (P2P) distribution approach. Shown are three peer stations 370, 380 and 390. Each peer station 370, 380 and 390 in the P2P network contributes its resources, such as the local storage 372, 382, 292, respectively. According to the invention, availability of the network titles in the P2P network is controlled in dependence on the respective audience shares. This may be achieved in many ways. Fig. 3 shows an embodiment wherein each peer station 370, 380 and 390 includes a respective distribution controller 374, 384, and 394. The distribution controller 374, 384, 394 may ensure that a network title in its local storage 372, 382, 392, respectively, may be pro-actively pushed to other peer stations in order to better meet expected demand and to increase robustness in case a peer station fails or is switched-off by a user. A peer station is usually also a client station that has once obtained content titles (e.g.,

through the network) with the intention of rendering it. Thus, the distribution controller 374, 384, 394 executes a replication strategy influenced by the audience share. The peer stations may also perform caching, in a manner similar to as described for server 100 of Fig. 1. Instead of or in addition to replication, a local distribution controller 374, 384, 394 may also use the audience share to ensure that the local peer station is not overloaded by downloads of a too popular title. To this end, it may use the audience share to proactively check the popularity of titles. If the audience share of a title is above a threshold, the local distribution controller 374, 384, 394 may restrict the number of downloads for that title (or all titles in its local storage). It will be appreciated that as time passes the restrictions may be lifted depending on actual demand. In the P2P network of Fig. 3, the main source server 300 may still have a role in controlling the availability. For example, the main server 300 may actively push a popular title to some peer stations to kick-start the replication. The server 300 will typically also distribute the audience share information in combination with the network title. Persons skilled in the art will appreciate that any suitable combination or mix of the three main embodiments of Figs. 1 to 3 is possible. Furthermore, in many practical implementations the method will be executed using one or more suitably programmed processors. Typically, a peer station will be conventionally personal computer equipped with suitable software. The servers (central server and edge servers) will typically be implemented on Internet server platforms, using suitable software. It will be appreciated that the invention also extends to computer programs, particularly computer programs on or in a carrier, adapted for putting the invention into practice. The program may be in the form of source code, object code, a code intermediate source and object code such as partially compiled form, or in any other form suitable for use in the implementation of the method according to the invention. The carrier be any entity or device capable of carrying the program. For example, the carrier may include a storage medium, such as a ROM, for example a CD ROM or a semiconductor ROM, or a magnetic recording medium, for example a floppy disc or hard disk. Further the carrier may be a transmissible carrier such as an electrical or optical signal which may be conveyed via electrical or optical cable or by radio or other means. When the program is embodied in such a signal, the carrier may be constituted by such cable or other device or means. Alternatively, the carrier may be an integrated circuit in which the program is embedded, the integrated circuit being adapted for performing, or for use in the performance of, the relevant method.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design many alternative

embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. Use of the verb "comprise" and its conjugations does not exclude the presence of elements or steps other than those stated in a claim. The article "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the device claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.