FR2855695A1 - Multimedia programs radio broadcasting method for e.g. PDA, involves adding header to multimedia programs before broadcasting each program, where header indicates type of client device for which program is destined - Google Patents

Abstract

The method involves adding a header to multimedia programs before broadcasting each program. The header indicates the type of a client device (2) for which a program is destined, to multiple client devices. The order and/or the frequency of program formats are modified in broadcasting cycle according to static or dynamic rules on the types of the client devices. Independent claims are also included for the following: (a) a client device functioning with a server which broadcasts multiplexed programs; (b) multiplexed programs broadcasting server adapted to function with a client device.

Description

METHOD AND DEVICE FOR RADIO CYCLIC BROADCASTING TO CLIENTS

DIFFERENT.

GENERAL TECHNICAL AREA.

  The invention relates to the general field of telecommunications.

  More specifically, the invention relates to the field of cyclic radio broadcasts of multiplexed programs to users.

STATE OF THE ART.

  Some public or private sites welcome a large number of visitors or 10 users. Such sites are for example large companies, museums, hotels, airports and / or stations, etc. Visitors may wish to have access to information when they are on the site, in particular thanks to a portable device in their possession, and this without a cable type connection. The users' portable device can be of the personal digital assistant or Personal Digital Assistant (PDA) type according to the English terminology generally used by those skilled in the art, laptop or mobile phone, etc. The information may be information on the site or on a company, or may be more general information, such as for example news, advertisements or a consultation of a page or of an Internet type site for example.

  There are known servers for cyclic radio broadcasting of programs containing information to portable client devices. In the remainder of the description, "program" is understood to mean information content in the broad sense, such as for example a page or a website, video or audio files, etc. Several methods are possible for the transmission and broadcasting of programs.

  One of the methods used is the cyclic broadcasting of multiplexed programs - or cyclic "multicast" broadcasting according to the English terminology generally used by a person skilled in the art. Such a broadcasting mode allows the radio transmission of a stream of programs dedicated to clients who have declared themselves to the broadcasting server. It also saves bandwidth on the server side. We can thus deliver information to a large number of customers simultaneously.

  Such a mode of dissemination, however, has a number of drawbacks. Generally, multicast cyclic broadcasts are only suitable for one type of client device. Programs are only provided in one format, which is not necessarily suitable for all client devices connected to the server. The diversity of client devices makes these broadcasting systems very limited. Indeed, each type of client device has a screen size, a number of displayable colors and specific processing capacities. It is difficult to remedy this drawback, since the multicast stream is unidirectional from the server to the client. It is therefore impossible to declare to the server the type of the client device during the broadcast so that the server broadcasts an adequate format to the client device. In addition, multicast broadcasting is ill-suited to viewing multimedia programs comprising an aggregation of graphic and textual content, such as, for example, programs such as the Internet page, video or audio files. Finally, one of the strong constraints of a multicast type radio link comes from the losses caused by the non-homogeneity of the radio environment. Thus, when a transmission problem occurs in a data exchange, it is the set to which these data correspond that must be fully retransmitted. The access time to information is therefore considerably increased for the user. 25 PRESENTATION OF THE INVENTION.

  The invention aims to overcome these drawbacks.

  One of the aims of the invention is to propose a method and a device making it possible to broadcast a program for a large number of different types of client devices.

  3 o One of the other aims of the invention is to propose a method and a device making it possible to adapt the broadcasting of program formats to the population and to the types of client devices which is connected to the server.

  Another object of the invention is to propose a method and a device allowing the declaration of the type of client devices to the server.

  One of the aims of the invention is to propose a method and a device making it possible to optimize access to the programs broadcast, in particular by better management of the broadcast bandwidth.

  One of the aims of the invention is in particular to allow faster access to the programs broadcast.

  One of the other aims of the invention is to propose a method and a device more suited to access to multimedia programs comprising an aggregation of graphic and textual content, in particular programs of the Internet page type.

  Another object of the invention is to provide a method and a device that does not require the installation of another browser or other rendering software on the client device, other than that already installed as standard on the device.

  To this end, the invention proposes a method of cyclic broadcasting of multiplexed programs broadcast by a server, each program 20 being in several different formats and intended for several client devices of different types, characterized in that it comprises the step whereby before broadcasting each program, a header is added to it indicating to client devices the type of client device for which the program is intended.

  The invention is advantageously supplemented by the following characteristics, taken alone or in any one of their technically possible combination: - each format is published at each broadcast cycle, using a multi-publication server; - each format is published thanks to a specific server by program format; - the order and / or the frequency of the program formats in the broadcast cycle is modified according to static or dynamic rules on the types of client devices; - the type of the client device is declared to the server when it is connected to the server; - the declaration is made using a pseudo user agent; - the order and / or the frequency of the program formats of the broadcast cycle are modified according to the types of client devices declared connected to the server; 0 - when a type of client device is declared and that no broadcast program format corresponds to it, a program format adapted to this type of client device is then broadcast; - at the request of a user, the client device selects the required programs whose format corresponds to its type and gives access to them; - For each accessed program, a pre-reading of the programs linked to the accessed program is carried out so that access to said linked programs is optimized upon a request from the user; the prefetching is carried out by the client device and comprises on the one hand a step of syntactic analysis of the accessed program and links between the accessed program and linked programs and on the other hand a step of storing the linked programs in means client device memory; - part of the analysis is performed on the server side; - a header is added at the server level to the broadcast programs indicating to the client device the links of the programs to each other, thus reducing the analysis performed by the client device to the analysis of the header; The header is added once and for all to the programs during an initialization phase before their cyclic broadcasting; - the header is added in real time to each program at each cycle of the cyclic broadcast; - a prefecture of certain programs linked, at least at a lower level, to the programs themselves directly linked to the accessed program is carried out; - the number of lower program levels subject to pre-reading is configurable; - the client device removes the header from the program before giving access to the user.

  The invention also relates to the client device and the server for implementing the method.

PRESENTATION OF THE FIGURES.

  Other characteristics, objects and advantages of the invention will emerge from the description which follows, which is purely illustrative and not limiting, and which must be read with reference to the appended drawings in which: - Figures 1A and 1B schematically represent a first possible embodiment of an assembly comprising a server and a client device according to the invention; - Figure 2 schematically shows a sequence diagram of a first set of operations performed in the server / client device assembly of Figures 1A and 1B; - Figure 3 schematically shows a sequence diagram of a variant of the process of Figure 2; - Figures 4A and 4B schematically show a variant of the embodiment of Figure 1; - Figure 5 schematically shows a sequence diagram 25 implementing a set shown in Figures 4A and 4B; - Figure 6 shows a first example of a program broadcast cycle; and - Figure 7 shows a second example of a program broadcast cycle.

  DETAILED DESCRIPTION OF THE INVENTION.

  FIG. 1A schematically represents a first possible assembly according to the invention.

  The set includes a server I for cyclic broadcasting of 5 multiplexed programs adapted to operate with a client device 2.

  The server 1 and the device 2 are connected to each other by a radio connection 3 for cyclic broadcasting of multiplexed programs - or multicast connection according to the English terminology generally used by those skilled in the art.

  The device 2 is portable and can be of the personal digital assistant or Personal Digital Assistant (PDA) type according to the English terminology generally used by those skilled in the art, laptop or mobile phone, etc. The broadcasting server 1 comprises a server 11 of the Web type 15 containing for example programs comprising for example textual and graphic content, and in particular pages and sites of the Internet type.

  The server 1 also comprises a multicast cyclic broadcasting server 12 connected to the Web server by a link using a transmission control protocol - or "transmission control protocol" (TCP) according to the English terminology generally used by the person skilled in the art. job.

  The server 1 also includes a server 13 comprising a syntax analyzer 131 - or "parser" according to the English terminology generally used by a person skilled in the art - and advantageously a generator 132 of headers.

  For its part, the client device 2 comprises a block forming a proxy server 21 - or "proxy" according to the English terminology generally used by those skilled in the art. In the embodiments of FIGS. 1A and 1B, the server 21 is advantageously a cache proxy server 21. The device 22 advantageously includes a browser 22 capable of requesting access to a program broadcast by the server 1.

  The block 21 forming the cache proxy server comprises a proxy server 211 connected, via the multicast connection 3 to the cyclic multicast server 12 of the server 1. The block 21 also includes a syntax analyzer 212, as well as means forming the cache memory 5 213. The means 211, 212 and 213 are firstly interconnected and firstly connected to the browser 22.

  FIGS. 1A and 1B only show one type of client device 2. As we have seen in the present description, several types of client devices 2 can be connected to the server 1. The invention allows also to broadcast several program formats corresponding to different types of devices. Such an embodiment of the invention is described in the following description.

  Referring again to FIG. 1A, it is noted that the fact of embedding on the client device 2 a cache proxy block 21 allows the client device 2 to perform operations for processing the broadcast programs and to optimize access. program information. It may in particular be a question of better management of the broadcasting bandwidth. The embarkation of the proxy block 21 notably makes it possible to speed up access to program information. For example, a certain amount of information on the programs broadcast by the server 1 can be placed in advance on the side of the client device 2.

  The means 21 in fact allow the device 2 to perform a pre-reading of the programs linked to the programs accessed during a request. Such a pre-reading allows the access to said linked programs to be accelerated upon a request from the user.

  Such a prefecture operation - also called prefetching according to the Anglo-Saxon terminology generally used by those skilled in the art is possible as long as the capacity of the cache memory 213 is sufficient.

  3 o We understand that because of the prefecture, the consultation of the programs linked to the accessed program, at the request of the user, is faster. Indeed, the requests listed below, usually made in series, are made in parallel thanks to the prefecture.

  - Consultation of a program, - Request from another program to the cache proxy block 21, - Request, from the cache proxy block, from the other program to the server 1, and - Return of information from the other program to cache proxy 21.

  According to a first method, the prefetching of the programs is carried out by the client device 2. Such a prefecture comprises a first step of analysis of the accessed program in order to determine the links between the accessed program and the programs which are linked to it. Such a step is carried out by the syntactic analyzer means 212. A second step in the prefecture is the storage of the linked programs in the memory means 213 of the client device 2.

  The prefecture can be carried out on o several lower levels of programs linked to each other, depending on the constraints of use. These usage constraints can for example be the capacity of the memory 213 of the client device 2.

  To limit the size of the cache means 213, the prefetch level can for example be limited to 1. Thus, a program A1 can be linked to three programs B1, B2, B3 by three links. It is level 1. Programs C1 and C2 for example can be linked to program B1, then we are at level 2 compared to program A1.

  Thus, a request to read the program A1 leads to the automatic caching of the programs B1, B2, B3. Thus, a request on B2 will instantly restore the contents of the B2 programs on the client device 2. The prefecture will be performed on the C1 and C2 programs linked to the B2 program, by caching which is done instantly. It is then understood that the process takes place step by step in the same way on all the programs accessed.

  Figure 2 illustrates this example.

  Thus, in FIG. 2, the server I can be distinguished in communication with the cache proxy 21, itself connected to the browser 22. Thus, during a step referenced by 300 in FIG. 2, the cyclic broadcasting of the set of programs, for example referenced by A1, 5 B1, B2, B3, Cl, C2. The programs are broadcast by the means 12 of the server 1. During step 301, the information is read cyclically by the cache proxy 21. During a step 302, a user requests access to the program A1. This access request is transmitted to the browser 22 by the user. The browser 22 itself transmits this request during a step 303 to the cache proxy 21. During step 304 the cache proxy 21 searches for the program requested in the cache memory 213. It is assumed that, during a first phase (when the client device 2 has just connected to the server 1 for example), the program A1 is not present in the cache memory 213. In this case, during step 305 the cache proxy 21 reads of the stream broadcast cyclically by the server 1. It then finds the program A1 in the cycle, and puts it in the cache memory 213 during a step 306.

  Step 307 corresponds to the sending of the information of the requested program A1 to the browser 22, so that A1 can be consulted by the user. Step 308 consists of a syntactic analysis by the analyzer 212 of the program A1. In step 309, the programs linked to the program A1, namely the programs B1, B2, B3, are cached. Thus, when the user sends a request for access to the program B2, and the browser 22 transmits this request to the cache proxy 25 during step 311, step 312 allows a quick search of the program B2 in the cache memory 213. The step 313 of returning the content of the program B2 to the browser 22 is thus faster. During step 314, a syntax analysis program B2 is carried out. We then find the links from B2 to other programs (for example C1 and C2). Step 315 is a step of reading the multicast stream (step included in step 309 above also mentioned).

  Step 316 is a caching of the programs C1 and C2 linked to the programs B2.

  It can be understood from this diagram that the time saved when requesting the program B2 is equal to the sum of the time required to read the multicast stream and the time to cache the program B2 before it is sent to the browser. 22.

  We have just described a first embodiment according to which the syntactic analysis of the programs is carried out by the cache proxy 21. It corresponds to the embodiment represented in FIG. 1A. 10 The program prefecture consumes two main machine resources. First, the analysis of programs in order to find existing links requires a syntactic analyzer - or "parser" according to the terminology generally used by those skilled in the art. Second, caching of related programs requires cache memory.

  We do not want to reduce the volume of the cache memory 213 too much, so that the client device 2 can receive several program formats, possibly relatively large in size.

  On the other hand, according to a variant of the embodiment shown in FIG. 1B, it is possible to limit the size of the analyzer 212 by adding on the server side a block 13 comprising a parser 131 and a generator of headers 132 , the block 13 being connected to the diffusion block 12. The other elements of the system are identical and bear similar reference numerals. The syntactic analysis of the programs broadcast is thus performed on the side of the server 1, which makes it possible to limit the use of the machine resources of the client device 2.

  The syntax analysis module is split in two, in order to save the client device 2 the burden of analyzing the content of the 30 programs. The analyzer 131 analyzes the programs to be broadcast syntactically. The header generator 132 enriches the content of the programs to be broadcast by adding a header of the eXtendible Markup Language (XML) type according to the Anglo-Saxon terminology generally used by a person skilled in the art, grouping the links present on the program, and allowing to indicate to the client device 2 said links.

  The header of page A1 can be: <xml> <a URL name = "B1"> Bl.html </uneURL> <a URL name = "B2"> B3.html </uneURL> <a URL name = "B3 "> B3.html </uneURL> </xml> <html> ...

  The operations taking place on the side of the client device 2 are therefore now the only analysis of the header added by the module 132.

  The analyzer 212 is therefore limited to a function of extracting the content of the header, and no longer to an analysis of the entire program as in the example of FIG. 1A. The storage of the linked programs is also carried out on the side of the client device 2.

  FIG. 3 schematically represents the sequences taking place during the use of an assembly according to FIG. 1 B. During a first step 400, the multicast server 12 cyclically broadcasts the programs A1, B1, B2 ... to block 13 through a link 4 using a TCP transmission control protocol. Block 13 performs, during a step 401, the analysis of the programs broadcast cyclically thanks to block 131. There is also the creation of headers grouping the information on the links thanks to block 132. Steps 400 25 and 401 are carried out either once and for all on all the programs during an initialization phase before their cyclic broadcasting to a client device 2. This situation is represented in FIG. 3. According to another solution, the header is added in real time for each program at each cycle of the cyclic broadcast. Of course, the creation of the headers 30 and the enrichment of the programs by the headers can be carried out upstream of the multicast server. These steps are not necessarily performed by the block 13 in the server 1. During a step 402, the enriched programs are broadcast cyclically thanks to the connection 3 to the client device 2, and in particular its cache proxy server 21. During a step 403, the cache proxy server performs a cyclic reading of the information. During a step 404, a user of the client device 25 wants to consult the program A1 and for this sends a request to the browser 22 which transmits during a step 405 said request to the block 21. It is assumed as before that the request A1 is made during a first connection of the client device 2, and therefore that the cache memory 213 is empty. Therefore, during step 406, the device 2 searches for the program A1 in the cache memory 213, does not find it, and reads, during step 407, the enriched multicast stream. In step 408, the cache proxy 21 caches the information of the program A1 and sends in step 409 the content of the program A1 to the browser 22 so that the user can access it. Step 410 consists of an analysis of the header of the program A1 by the analyzer 212 of the device 2. It will be recalled that this analysis consists of extracting and processing the content of the header. During step 411, a caching 213 of the programs linked to the program A1, B1, B2, B3 is carried out.

  Thus, when the user sends a request for access to the program B2 to the browser 22, the latter passes on a request to the cache proxy 21 during step 413. Step 414 of searching for the program B2 in the cache 213 is therefore performed more quickly.

  Sending the content of the B2 program during step 415 is therefore also rapid. Step 416 consists of an analysis of the header of the program B2 B2 by the device 2. Step 417 corresponds to a reading of the multicast stream (step which is understood during step 411 cited above), in order to to provide in 418 a caching 213 of the programs linked to B2, for example C1 and C2.

  The client device 2 also preferably has means which remove the header of the accessed program before giving access to the user.

  In all of the above developments, only one type of client device 2 has been considered. The method may thus be intended for only one type of client device. It is recalled, however, that the device 2 can be of several types, for example mobile phone, laptop or personal digital assistant.

  FIGS. 4A and 4B represent variants of the embodiments of FIGS. 1A and 1B making it possible to consider and broadcast to several types of client devices 2.

  In FIGS. 4A and 4B, the elements identical to FIGS. 1A and 1010 B bear similar reference numerals.

  Several program formats are broadcast.

  In FIG. 4A, the broadcasting of the different formats is made possible by the presence of a specific Web server by broadcasting format. Thus, for the same program, there is a Web server 111 for the format compatible with personal digital assistants, a Web server 112 for the format compatible with mobile phones, and a Web server 113 for the format compatible with portable computers.

for example.

  In FIG. 4B, the broadcasting of the different formats is made possible by the presence of a multi-publication web server 11, capable of generating the different broadcasting formats.

  In both cases, the different formats for the programs are edited once and for all during an initialization phase or for each program for each broadcast cycle.

  In the two embodiments of FIGS. 4A and 4B, the server 1 includes an analyzer 131 making it possible to analyze the formats of each program which is going to be broadcast. The server 1 comprises a header generator 132 which will generate a header, the header being added to each program before it is broadcast. The header indicates to client device 2 the type 30 of client device for which the broadcast program is intended.

  FIG. 5 schematically represents the different steps implementing the assembly according to FIGS. 4A or 4B.

  Thus, during a step 500, the programs of the various servers 111, 112, and 113 or of the multi-publication server 11 are enriched by block 13, the headers indicating the type of program format.

  Thus, each header can be: <xml> <typeDeClient> PDA </typeDeClient> </xml> <html> ... or <typeDeClient> GSM </typeDeClient> 10 <WML> ...

  During a step 502, the multicast server 12 performs a prioritization algorithm making it possible to place the different program formats in the cycle chain to broadcast. Several prioritization algorithms are possible and are described in the rest of this description. During a step 503, the multicast server 12 distributes the enriched programs cyclically thanks to the link 3 with the proxy block 21. One of the differences with the embodiments of FIGS. 1A and 1B is that the proxy block does not necessarily include a cache memory, as indicated by the dotted lines in FIG. 4A. In other words, placing a format header at the client devices can be independent of the prefecture method described above. As part of the method for generating format headers whose steps 500 to 503 have just been described, it is not necessary to perform a preread of the programs. Thus, the client devices can be satisfied with recognizing and giving access, at the request of the user, to the program whose format is adapted to their type, without performing a pre-reading of the programs linked to the accessed programs.

  It is understood that the proxy 21 may however include a cache memory in order to be able to perform a prefetching of the programs 30 in addition to recognition of the formats of the programs.

  Referring again to FIG. 5, during a step 504, a user of a client device 2 (for example PDA) makes a request on the program A1. This request is transmitted by the browser 22 to the proxy 21 during a step 505. During a step 506, the proxy 21 performs a multicast stream reading and displays in step 507 the format of the program A1 which corresponds to the device used by the user, namely the format corresponding to the PDA. In the embodiments shown in FIGS. 4A and 4B, the analyzer 212 only makes a selection of the programs for which a request has been launched, and the format of which corresponds to its type. This is an analysis of the header of the enriched program, but as to its format. This is not a syntactic analysis of a header at the top of the program to determine links between the programs.

  However, as has already been explained, the analyzer 212 of FIGS. 4A and 4B is advantageously capable of carrying out such a syntactic analysis of the headers in order to determine the links between the programs and to perform a prefetching of said programs.

  It is recalled that several prioritization algorithms are possible during step 502.

  We can imagine a very simple broadcasting cycle where the different formats for each program are repeated one after the other, as shown in FIG. 6. The formats generated by the servers 111, 112 and 113 are thus repeated successively for the different pages A1, B1, B2, B3 for example.

  One can imagine a prioritization process where the order and / or frequency of program formats in the broadcast cycle can be modified according to static or dynamic rules.

  Static rules include, for example, decisions by the system administrator or statistical or marketing data on the types of client devices usually connected to the server. For example, if we note that statistically the population of 30 client devices 2 is divided into 70% mobile phones, 20% laptops, and 10% PDAs, the formats for mobile phones will be repeated in the proportions 7 by compared to the formats of laptops and PDAs which will be in the proportions 2 and 1 respectively.

  The dynamic rules take into account in real time the types of client devices connected to the server. It is thus also possible, according to an advantageous variant, to declare to the server the type of client device 2 when it is connected to the server. Such a declaration can be made using a pseudo user agent - or pseudo user agent according to the English terminology generally used by those skilled in the art - compatible with the unidirectional multicast flow and comprising for example a string 10 of characters specifying the type of client device (personal digital assistant, mobile phone, laptop ...). In this case, the server 1 has knowledge of the number of clients and of the type of client devices (for example by means of a counter system). This increases the frequency and / or the order of formats corresponding to the most numerous type of client device. FIG. 7 represents a broadcast cycle in which the number of clients using a laptop computer is greater than the number of clients using a PDA, itself greater than the number of clients using a mobile telephone. The format generated by the server 113 is more often repeated for each program than the format generated by the server 112 dedicated to portable telephones.

  In the latter case also, the server 1 can advantageously include means making it possible to transmit in the broadcast cycle a format corresponding to a new type of client device if the format did not exist before this new declaration in the broadcast cycle.

  Another criterion for prioritizing certain formats over others would be, for example, the bandwidth occupied by broadcasting for a given type of client device. One can thus imagine that the programs for portable computers are sent less frequently, because they consume more bandwidth than the programs for PDAs for example.

  The method of placing a format header on the programs before their broadcast can be carried out without the prefetching process.

  However, as will be understood, the method of placing a format header on the programs before their broadcasting is advantageously combined with the method of prefecture of the programs linked to a program which is accessed by a user.

  Conversely, a method for prefecting programs linked to an accessed program can be carried out alone, but is advantageously combined with a method for indicating the format of the pages broadcast.

Claims (31)

CLAIMS.   1. A method of cyclic broadcasting of multiplexed programs broadcast by a server (1), each program being in several different formats and intended for several client devices (2) of different types, characterized in that it comprises the step according to which, before broadcasting each program, is added a header indicating to the client devices (2) the type of client device for which the program is intended.   2. Method according to claim 1, characterized in that each format is published in each broadcast cycle, thanks to a server (11) for multipublication.   3. Method according to claim 1, characterized in that each format is published by means of a server (111, 112, 113) specific by program format.   4. Method according to one of claims 1 to 3, characterized in that the order and / or frequency of the program formats in the broadcasting cycle is modified according to static or dynamic rules on the types of devices. clients.   5. Method according to one of claims 1 to 4, characterized in that the type of the client device is declared to the server (1) when it is connected to the server by means of a pseudo user agent.   6. Method according to claim 5, characterized in that the order and / or the frequency of the program formats of the broadcast cycle is modified according to the types of client devices (2) declared connected to the server.   7. Method according to one of claims 5 to 6, characterized in that, when a type of client device (2) is declared and that no broadcast program format corresponds to it, a suitable program format is then broadcast to this type of client device.   8. Method according to one of claims 1 to 7, characterized in that the client device (2) selects the programs whose format corresponds to its type and gives access to them on request by a user.   9. Method according to one of claims 1 to 8, characterized in that for each accessed program, a prefecture of the programs linked to the accessed program is carried out so that access to said linked programs is optimized upon request from the of the user.   10. The method as claimed in claim 9, characterized in that the prefetching is carried out by the client device (2) and comprises on the one hand a step of syntactic analysis of the accessed program and links between the accessed program and related programs and on the other hand a step of storing the linked programs in memory means (213) of the client device.   11. Process according to one of claims 9 to 10, characterized in that part of the analysis is carried out on the server side (1).   12. The method of claim 11, characterized in that one adds, at the server level, a header to the broadcast programs indicating to the client device the links of the programs to each other, thereby reducing the analysis performed by the client device to the header analysis.   13. The method of claim 12, characterized in that the header is added once and for all on the programs during an initialization phase before their cyclic broadcast.   14. Method according to claim 12, characterized in that the header is added in real time to each program at each cycle of the cyclic broadcasting.   15. Method according to one of claims 9 to 14, characterized in that a pre-reading of certain programs linked, at least at a lower level, is carried out to the programs themselves directly linked to the accessed program.   16. Method according to claim 15, characterized in that the number of lower levels of programs subjected to a pre-reading is configurable.   17. Method according to one of claims 9 to 16, characterized in that the client device removes the header of the program before giving access to the user.   18. Client device adapted to operate with a server (1) for cyclic broadcasting of multiplexed programs, each broadcast program being in several different formats corresponding to 2 5 different types of client devices, characterized in that it comprises means (21) able to select the programs whose format corresponds to its type thanks to a header added to the broadcast programs indicating to the client device (2) the type of client device for which the program is intended. 30 19. Device according to claim 18, characterized in that it is of the mobile telephone, laptop computer or personal digital assistant type.   20. Device according to one of claims 18 to 19, characterized in that it comprises means forming browser capable of requesting access to a broadcast program, and in that it comprises means suitable for, for each program accessed , perform a prefecture of certain programs linked to the program accessed by means of a header 10 added to the broadcast programs indicating to the client device the links of the programs to each other.   21. Device according to claim 20, characterized in that it comprises means forming a server (21) cache proxy capable of carrying out the prereading of the programs.   22. Device according to claim 21, characterized in that the cache proxy server comprises means (212) able to analyze syntactically links between the accessed program and linked programs and means (213) able to store linked programs .   23. Device according to one of claims 20 to 22, characterized in that it comprises means (21) capable of carrying out a prefecture of the 25 programs linked to at least one level lower than the programs themselves directly linked to the program accessed, this level being configurable.   24. Device according to one of claims 20 to 23, characterized in that it comprises means capable of removing a at the head of the program indicating the links of the programs between them before giving access to the user .   25.Server (1) for cyclic broadcasting of multiplexed programs adapted to operate with a client device (2), characterized in that it comprises means (13) suitable for, before the broadcasting of each 5 program, add a header indicating to client devices the type of client device for which the program is intended.   26.Server according to claim 25, characterized in that it comprises a multi-publication server (11) able to publish each format at 1 0 each broadcast cycle.   27. Server according to claim 25, characterized in that it comprises a server (111, 112, 113) specific by program format.   28. Server according to one of claims 25 to 27, characterized in that it comprises means capable of declaring the type of the client device with the server when it is connected to the system.   29. Server according to one of claims 25 to 28, characterized in that it comprises means capable of modifying the order and / or the frequency of the program formats of the broadcast cycle according to static rules on the types client devices in the system or dynamic on the types of client devices declared in the system.   30.Server according to one of claims 25 to 29, characterized in that it comprises means capable of being able to broadcast in the cycle a program format adapted to a type of client device having just declared and not declared until so.   31.Server according to one of claims 25 to 30, characterized in that it comprises means forming browser (22) capable of requesting access to a broadcast program, and in that the server comprises means (13) able to add a header to the broadcast programs indicating to the client device (2) the links between the programs, the header then allowing the client device 5 to perform a pre-reading of certain programs related to the accessed program.