FR2787662A1 - METHOD FOR INCREASING THE AUTONOMY OF DIGITAL INFORMATION RECEIVERS AND CORRESPONDING INFORMATION TRANSMISSION SYSTEM - Google Patents

Abstract

Successive frames are materialized by grouping packets and inserting at the start of each packet a rank identification block of the packet in the frame. Time-spaced network signaling blocks with a short maximum duration, less than that of transmission of a packet, are inserted into the successive frames. Each receiver tests (700, 701) the presence or not of a network signaling block corresponding to its own network, then synchronizes (71) on the successive occurrences of the packet identification blocks corresponding to its group.

Description

I Method for increasing the autonomy of information receivers

  The invention relates to the transmission of digital information between at least one transmitting station and a plurality of receivers, using sub-carriers of radio frequency channels as well as an asynchronous packet transmission protocol containing each a predetermined number of elementary blocks of data. Those skilled in the art are already familiar with such a transmission protocol, in particular by document pr ETS 300 751 (May 1997 version), entitled "Radio broadcast systems; System for Wireless Infotainment Forwarding and Teledistribution (SWIFT)", available from ETSI Secretariat in Sophia Antipolis (France), and

  hereinafter referred to as "SWIFT document".

  Currently, this type of asynchronous transmission protocol is used for transferring files intended for a group of users, and in particular for radiofrequency broadcasting of

written journals.

  As opposed to a synchronous transmission protocol, such as that provided for by the RDS (Radio Data System) transmission standard, and in which the various transmission frames are synchronized on a perfectly defined time clock, an asynchronous transmission protocol does not East,

  by definition, not synchronized on a predetermined time basis.

  Consequently, the information receivers operating on this type of asynchronous protocol must remain permanently on so as to be able to acquire, at any time, which by definition is not known in advance, digital information which are for them. This poses a significant problem of autonomy for receivers operating with their own power supply, for example portable microcomputers operating on batteries or on accumulator. The invention aims to provide a solution to this problem and to

  increase the flow of useful data to receivers.

  The invention therefore provides a method for increasing the autonomy of digital information receivers, in which the information is transmitted from at least one fixed station by sub-carriers of radio frequency channels using an asynchronous packet transmission protocol each containing a number

  predetermined elementary blocks.

  According to a general characteristic of the invention, at least one specific channel is selected from the radiofrequency channels so as to materialize at least one specific identifiable transmission network. A transmission frame comprising a predetermined number of successive packets is defined within the transmission protocol carried by this network. A specific elementary package identification block is inserted into each packet of the frame, at a predetermined location, containing an identifier of the rank of the packet in the frame and a useful part, the useful part comprising a binary word making it possible to designate, within the group of receivers, families of receivers, the number of families of receivers being greater than the number of bits of the binary word, the logical value of each bit of an identification message indicating whether the receivers of the family corresponding are

  likely to receive a message.

  A group indication corresponding to a packet rank in said frame and an indication is stored in each receiver.

  of family corresponding to a family rank in a package.

  We analyze the identification message to determine the type of compression algorithm to use to code the message.

  of identification on the number of bits of the binary word.

  We compress the identification message.

  The receiver being tuned to said specific transmission channel, analyzes the rank identifier contained in the next block

package identification.

  The method furthermore comprises a synchronization step in which, the receiver being locked onto said specific transmission channel, and moreover having an idle state and an active state, the setting in the active state of each receiver on time is synchronized successive occurrences of packet identification blocks whose rank identifiers correspond to

  the group indication stored in the receiver.

  Each receiver being able to decompress the binary word to find the identification message, the receivers not belonging to the families identified in the idle state are returned and the receivers belonging to the families identified in the active state are maintained. This reduces the duration of activity of the receptors and increases their autonomy, in particular in the case, conventionally the

  more unfavorable, numerous and short messages.

  The identification message can be compressed by truncating at least one bit or by coding the lengths of the series of successive bits having the same value or comprising only "0" except the last bit to "1". It is also possible to perform compression by segmentation, identification of the segments comprising only bits of the same value, for example "0", coding of the location of said segments and coding of the other segments. These last two variants are interesting when the number of bits at "1" is small compared to the total, in other words if the number of messages to be transmitted during a given time interval is small compared to the number of families. We can also divide the identification message into a number k of parts, k being greater than or equal to 2, we apply an operator "or logic" between the parts, then we carry out the coding on the sequence thus obtained whose length is equal to that of the identification message divided by the number k. You can transform the identification message by permutation before division

  and insert the number k in the first bits of the binary word.

  Advantageously, an identifier of the type of compression chosen is inserted into a header of the packet identification block. The number of bits at 1 in the identification message can be analyzed in order to determine the type of compression algorithm to be used. to encode the identification message on the number of bits

of the word binary.

  It is also possible to insert, in successive frames, elementary network signaling blocks containing at least one network identifier, and mutually separated by a variable number of elementary blocks less than the number of blocks in a packet. We

  then stores a network indication in each receiver.

  In other words, if the number of blocks in a packet leads to a transmission duration of said packet of approximately 5 seconds, we will choose to space the network signaling blocks by a number of blocks corresponding to a maximum duration. 0.5 seconds for example. This being, within the meaning of the present invention, the "variable" nature of the number of elementary blocks separating two successive network signaling blocks means that this number is not necessarily fixed and predetermined in advance (although it may be) but that it may possibly vary during the transmission, provided, of course, that the corresponding duration of spacing remains less than a maximum duration fixed in advance which must be in all

  event less than the transmission time of a packet.

  In a so-called acquisition step, each receiver detects the specific transmission channel by testing the presence of a network signaling block containing a network identifier corresponding to said network indication stored in the receiver. The

  receiver thus locks onto the specific transmission channel.

  The invention therefore applies in a particularly advantageous manner to paging and makes it possible, from an asynchronous transmission protocol, to synchronize the alarms of paging receivers at successive instants temporally spaced by a

same time interval.

  In addition, the fact of inserting in the successive frames, separate network signaling blocks with a spacing duration less than or equal to a predetermined maximum duration (for example 0.5 seconds), itself less than the transmission time of a packet (typically 5 seconds) makes it possible to accelerate the acquisition of the identifier of the transmission network by the receivers, and consequently the synchronization of their sleep and wake-up cycles, which considerably increases the autonomy of these receivers. Although it is theoretically possible to insert each packet identification block at any predetermined place in the packet, in particular if the type C packets described in the SWIFT document are used, it is particularly advantageous, in particular in order to minimize the risk of loss of messages intended for certain receivers, to insert each packet identification block at the start of

each package.

  According to an embodiment of the method, in the acquisition step, the receiver is wedged on one of the transmission networks and the test is carried out, at most during a test duration equal to said predetermined maximum duration (0, 5 seconds for example), the presence, on this channel, of a network signaling block containing a network identifier consistent with said network indication and, if the test result is negative at the expiration of said test duration , the receiver searches for another transmission channel to carry out a new test. It is also particularly advantageous to insert in the network signaling block an indication of the rank of this block in said packet. Thus, in the acquisition step, after the receiver has analyzed the content of the network signaling block, the receiver is put in its idle state until the next packet identification block is received. This allows to further increase

the autonomy of the receivers.

  However, as a variant, the receiver can be kept in its active state after analysis of the content of the network signaling block until the reception of the next packet identification block. The subdivision of a group of receivers into several families makes it possible to put the receivers whose families are not designated at the start of each packet, in particular because no message is intended for them in their idle state which allows

  further increase the autonomy of these receivers.

  Several specific identifiable transmission networks can be materialized. Transmission networks can be changed during the actual transmission. Also, in the event of a network modification during transmission, the new network identifier is inserted into the network signaling blocks and advantageously is inserted into the packet identification blocks specific network modification information. For example, a specific bit is set to 1. Thus, in the synchronization step, all the receivers of the group associated with the corresponding packet identification block are forced to remain in their active state after analysis of the content of the identification block. at least until the reception and analysis of the content of the next network signaling block, so that they can determine the new transmission network

and stall there.

  According to one embodiment of the method, an elementary message signaling block containing the address of a receiver intended to receive a message is inserted into a transmission packet, after the packet identification block. 'a time indication relative to the message start time. After analysis of the content of the message signaling block, the receiver can possibly be placed in an idle state until the message is received. As an indication, the time indication may specify that the message itself will be located later in the same packet, or else in a next packet. In fact, in a paging application, due to the random distribution of the addresses of the receivers intended to receive a message, a greater number of messages than a packet can contain, may have to be sent. This difficulty is resolved by the use of the indirect mechanism constituted by the insertion of message signaling blocks. In other words, instead of inserting the message directly into the designated packet, the message signaling block is sent to designate the useful data of the message and also includes a particular bit (flag) indicating whether the designated message is delayed or not. Thus, in the case of a delayed message, that is to say transmitted in a subsequent packet, the receiver can go into its idle state after analysis of the content of the message signaling block and then wake up each time reception d a packet identification block, even if this packet identification block does not correspond to its group, to check whether this packet identification block contains a family identifier corresponding to its family. If this is the case, the receiver will then analyze the new message signaling block indicating the place of this message proper in this

current package.

  The invention also relates to a digital information transmission system, comprising at least one fixed station transmitting information by sub-carriers of radio frequency channels using an asynchronous packet transmission protocol each containing a predetermined number of elementary blocks , and several receivers comprising means for radio frequency reception and means for processing the data received. According to a general characteristic of the invention, the station includes generation means capable of generating, within the transmission protocol carried by at least one specific identifiable transmission network, materialized by at least one specific transmission channel selected from said channels radio frequency, successive transmission frames each comprising a predetermined number of successive packets. The station also includes insertion means for inserting into each packet of the frame, at a predetermined location, a specific elementary packet identification block containing an identifier of the rank of the packet in the frame and a useful part, the useful part comprising a binary word making it possible to designate, within the group of receivers, families of receivers, each group of receivers being subdivided into several families, and the number of families of receivers being greater than the number of bits of the word binary, the logical value of each bit of an identification message indicating whether the receivers of the corresponding family are capable of receiving a message. The station includes means for analyzing the identification message in order to determine the type of compression algorithm to be used to code the identification message on the number of bits of the binary word, and means for compressing the identification message . Each receiver includes a memory containing a group indication corresponding to a packet rank in said frame, and an indication of the family to which the receiver belongs, processing means having an active state and a rest state, and test means. able to analyze the rank identifier contained in the next packet identification block and decompress the binary word for

  find the identification message.

  The receiver processing means furthermore comprise control means able to synchronize temporally the setting in the active state of each receiver set on a specific transmission channel, on the successive occurrences of packet identification blocks identifying packets whose ranks correspond to the group indication stored in the receiver, to maintain in the active state the receivers whose family indications correspond to the identification messages and to set

  the resting state of the other receptors.

  According to one embodiment of the invention, the station comprises means for inserting into successive frames elementary network signaling blocks containing at least one network identifier, these network signaling blocks being mutually separated by a variable number of elementary blocks less than the number of blocks in a packet. The receiver processing means are capable of analyzing the content of the packet identification blocks corresponding to their group and, in the presence in said identification blocks of specific network modification information, the means for controlling all the Group receivers, associated with the corresponding packet identification block, force these receivers to remain in their active state at least until the reception and analysis of the next network signaling block. The memory of each receiver has a network indication. The receiver processing means include means for testing, for each transmission channel received by the receiving means, the presence of a network signaling block containing an identifier

  network consistent with said network indication.

  Other advantages and characteristics of the invention

  will appear on examination of the detailed description of a mode of

  realization and implementation, in no way limitative, and attached drawings in which: - Figures 1 to 4 very schematically represent a structure of elementary blocks usable according to the invention compatible with a SWIFT transmission protocol, - Figure 5 illustrates very schematically the internal architecture of a receiver according to the invention, - Figure 6 very schematically illustrates the internal architecture of a station according to the invention, - Figures 7 and 8 are schematic flowcharts of a mode of setting implementation of a method according to the invention and, - Figures 9 and 10 illustrate time diagrams relating to the operation of a receiver in two particular cases of transmission. Although the invention is not limited thereto, we will now rely in the example which will be described, on an asynchronous transmission protocol of the type described in the SWIFT document. Indeed, such a transmission protocol allows a high net speed of information transmission, typically between 6 kilobits / s and 10 kilobits / s, which is particularly useful for

transmission of voice messages.

  Those skilled in the art know the characteristics of such an asynchronous packet transmission protocol each containing a predetermined number of elementary blocks. We recall here the essential characteristics relating in particular to the structure of these elementary blocks. Those skilled in the art may possibly, for more details, refer to the aforementioned SWIFT document, the

  content is for all intents and purposes incorporated into this description.

  According to this transmission protocol, each elementary block BLC comprises (FIG. 1) a BIC identifier followed by an AND header indicating the type of the block. The header is followed by a useful part DU containing the data proper of the BLC block. This useful part

  DU is followed by CRC check bits and PRY parity bits.

  Each of these blocks comprises a total of 288 bits. These blocks are combined in packets of 272 blocks and these packets are transmitted one after the other asynchronously, that is to say without it being possible to determine the times of transmission temporally and in advance. of these

  packages and corresponding contents.

  The transmission time of a packet is typically of

around 5 seconds.

  This asynchronous transmission protocol is conveyed by a subcarrier (centered around 76 kHz) of radio frequency channels,

  typically located in the FM band.

  To convey the asynchronous transmission protocol according to the invention, at least one specific identifiable transmission network is materialized by selecting from said radiofrequency channels at least one specific channel. Of course, a transmission network can use several radio frequency channels. Of

  even, several identifiable transmission networks can be defined.

  A transmission frame comprising a predetermined number of successive packets, for example sixteen packets, is then defined for this transmission protocol according to the invention. We then insert into each packet of the frame, preferably at the start of each packet, a specific BLTS packet identification block (Figure 3) identified by a specific digital word ET2 (for example on five bits) of the head AND. This BLTS identification block contains, in the DU part, an IR identifier making it possible to identify the rank of this packet, in other words its position, in the frame. Thus, this IR rank identifier will make it possible to designate a group of receptors. Similarly, provision is advantageously made for the useful part DU to include a binary word MSF making it possible to designate, inside It

  of the receptor group, families of receptors.

  The number n of receiver families is greater than the number of bits of the binary word MSF. MI denotes an identification message on n bits whose logical value of each bit indicates whether the receivers of the corresponding family are capable of receiving a message. The higher the number n of bits of the identification message MI, the fewer there are receivers per family and therefore the receivers of a given family are put in their active state less often and this setting in the active state concerns less receivers. It thus succeeds in reducing the active periods of the receptors. However, if the MI identification message were transmitted directly, a high number of bits would be used which would be harmful to the transmission of the useful data. Now, we have noticed that the number of bits at "1" was small, less than 10% o if n = 1024. We perform a compression of the identification message MI to code it on the

number of bits of the MSF binary word.

  If the MI identification message is 1024 bits, it can be coded without compression on a number of bits which depends on the number of bits at "1" of said MI message. The bit of rank 1023 of the MI message requires 10 bits for its coding, 1023 = 1 1 1 1 1 1 1 1 1 1. With 128 bits, an MI message can therefore be transmitted, without losses

maximum 12 bits at "1".

  To reduce the number of bits to be transmitted, compression will be carried out to transmit the content of the identification message MI on 128 bits. To indicate the type of compression carried out, we insert in the header ET2 of the BLTS identification block,

  an identifier IC for example on two bits.

  BLNS network signaling blocks (FIG. 2) are also inserted in the successive frames, the type of which is identified by a specific word ET1 in the header ET. This BLNS block contains at least one NWD network identifier defining the characteristics of the transmission network, and in particular the frequency of the transmission channel. As will be seen in detail below, these BLNS blocks are inserted by the transmitting station in the successive transmission frames so that they are mutually spaced from a maximum number of elementary blocks of any type, whether they are BLC type blocks, BLTS blocks or other types of blocks. This maximum number corresponding to a predetermined maximum spacing duration, typically 0.5 seconds, less than the duration of transmission of a packet. The spacing time between two consecutive BLNS blocks may vary provided that it remains less than the maximum duration. It is in fact the means of insertion of the station which decide on the insertion of these BLTS blocks in the successive frames. It should therefore be noted here that the insertion of BLNS blocks takes place in a totally asynchronous manner with respect to

when inserting BLTS blocks.

  The invention also provides for using another specific BLMS elementary block which is in fact a message signaling block (FIG. 4) and whose type ET3 appears in the AND header. This BLMS block contains in the useful part DU the address ADR of the receiver for which a message is intended, as well as a time indication DFS relative to the instant of start of message. If the message is transmitted in the same packet as that containing the BLMS block, this DFS indication defines for example the rank of the first block of the message proper, intended for the receiver, relative to the start of the packet. This DFS time indication may also indicate that the actual message to the receiver will be transmitted not

  in the package containing this BLMS block, but in a subsequent package.

  If we now refer more particularly to the figure which represents the internal architecture of an RC receiver according to the invention, it can be seen that this receiver, for example an autonomous portable paging receiver, is equipped with an antenna reception receiving the SRF radio frequency signal. This antenna is connected to reception means 1 comprising, at the head, a high frequency stage 10, followed by a specific filtering circuit 11 making it possible to extract the SWIFT subcarrier. This filtering stage is itself connected to a decoder, for example that marketed by the Japanese company OKI under the reference MSM 9553 and capable of delivering as an output, after error correction and parity check,

  the elementary blocks cleared in particular of the CRC and PRY bits.

  The reception means also comprise a specific circuit, not shown here for the purposes of simplification, research and automatic control of the frequency of the carrier signal, so as to be able to lock onto one of the transmission channels. The output of the reception means is connected to processing means 2 incorporating a microprocessor 20, for example 8 bits, associated, via a communication bus, with a

  RAM 21 and ROM 22.

  Control means 3 are capable of momentarily activating the reception means and the processing means by delivering respectively to these two means corresponding control pulses. These control means can be incorporated in a conventional manner within the microprocessor 20, or

  well can be achieved by a conventional specific external circuit.

  These control means are therefore able to put the receiver, either in an active state in which it is capable of receiving and processing the data contained in the carrier signal, or in a

resting state.

  The entire receiver is supplied by power supply means 4 comprising a cell element associated with a DC-DC converter used to raise the voltage of this cell element to that necessary for the operation of the

microprocessor.

  An IDRR network indication, an IDG group indication corresponding to a packet rank in said frame, an IDSG family indication corresponding to the family to which the receiver belongs, the ADD address proper of the receiver, is stored in the read-only memory of the receiver. receiver, and decompression data allowing the processing means 2 to

  decompress the binary word MSF according to the identifier IC.

  If we now refer more particularly to FIG. 6, we see that the transmission station ST essentially comprises analysis means MA and insertion means MIS. The analysis means MA are provided for studying the number of bits at "1" of the identification message MI and their distribution. Different types of compressions are carried out to determine which one will cause the least loss, the losses resulting in families of receivers remaining unnecessarily in the active state. This determination is carried out by comparison between the identification message MI and the said message after compression and decompression. The result of the

  comparison determines the value of the bits of the identifier IC.

  For compression, one can use for example a coding method called RLC (Runlength coding), see figure 7. The code obtained is a succession of values representing the lengths of the series of: "1" and "0". These values are called code words. As an example, the following uncompressed sequence

  000000111001111111 will have the code 6327.

  For sequences that are poor in "1", it is preferable to code series of "0" ending in a "1". The code is given by the length of these series as in the following example: uncompressed sequence 00000100110000001-code 6317. The advantage is that, if the bits at "1" are isolated, which is quite likely due to the small number of families having to simultaneously receive messages, for each bit at "1" there is only one code word while in coding

classic there are two.

  When using this modified coding, the most unfavorable case is to have a series of length 1024, that is to say a single bit at "1" which is placed on the last position. It is therefore necessary to reserve for each 10-bit code word. With 128 bits, you can code losslessly,

  sequences with at most 12 bits at "1".

  To improve this performance, the coding can be modified by the following preprocessing: - the original sequence is divided into 32 equal segments, each having 32 bits, - the segments which do not contain a bit at "1" are eliminated, in this way , we keep all the bits at "1" in a shorter sequence in which the maximum number of bits at "0" separating two bits at "1" is equal to 62 because any sequence of bits at "0" of greater length has been reduced by 32 bits or a multiple of 32 bits. The maximum length of the series can now be coded with only 6 bits, - on the other hand, to recover the sequence when decompressing, it is necessary to memorize the positions of the segments which have been removed; for this purpose, a 32-bit card is used; each bit corresponds to a segment; if the segment is removed, the corresponding bit is "0", otherwise it is "1". With this variant, sequences comprising at most 16 bits at "1" can be losslessly coded. Thus, the binary word MSF will have the following structure: 32 bits of coding of the segments comprising either only "0", or at least one "1"; sixteen 6-bit series

or 128 bits in total.

  As an example, we could have Segment 1 Segment 2

  00000000000000000000000000000000 0001000000000000000000000000000000

  32-bit card 01 ... Identifying and extracting segments of a particular type, for example comprising only "0" bits, makes it possible to reduce the length of the chain to be coded and to increase the number of bits to "1"

  that can be transmitted without loss.

  Preferably, both modified RLC encoding (series of "0" ending in "1") and the preprocessing of 32-bit segment elimination all at "0" are used to losslessly encode sequences comprising a number bits to "1" higher and reduce

  thus the number of receivers unnecessarily returned to the active state.

  If losses are accepted, performance can be pushed further. By losses is meant the fact that certain bits at "0" of the original sequence are found at "1" in the sequence after decoding by the receiver, knowing that any bit at "1" at the origin is retained after decoding . This results in the fact that any decoder which must be awakened will do so, but that some decoders which must remain at rest will nevertheless be awakened. We can for example fold the original sequence in half and perform the RLC coding on this reduced sequence. The sequence folds making a "or logic" between these two halves. This gives a sequence twice as short but which keeps all the bits at "1". If there are no bits at "1" superimposed, the folded sequence will have the same number of bits at "1"

than the original sequence.

  The folded sequence can then be divided into 32 segments which are shorter, only 16 bits. In this case, the maximum length of a series to be coded is 32 and the code word must have only 5 bits. We manage to encode a sequence comprising up to 19 bits at

IF 1 ".

  When we do the decompression, when unfolding, we obtain a sequence which has, at most, a double number of bits at "1", ie a

  100% error. The unfolded sequence will have two identical halves.

  If we fold the original sequence in 4, we obtain a sequence of 256 bits, which keep all bits at "1". In this case, the code word must have only 4 bits which give the possibility of coding 24 bits after having kept 32 bits for the card. When unfolding, we will have, at most, four times the number of bits at "1" which corresponds to a

300% error.

  If there were bits superimposed on the fold, then the error is smaller but the probability of having them is low because of the number

reduced bits to "1".

  To identify the length of the code word which can be 6, 5 or 4 bits, two bits are kept in the coded sequence and a

  code word in the lossless version with 300% loss.

  The sequence coded at 128 will include the 32-bit card which identifies the series comprising only bits at "0" and those

comprising at least one bit at "1".

  In the case of lossless coding, the coded sequence then comprises an identification zone for the length of the code word comprising two bits at "0" then fifteen code words at 6 bits, the

  bits 125 to 128 are not used.

  In the case of coding with 100% loss, the sequence includes an identification zone comprising a bit at "1", then 19

5-bit code words.

  In the case of coding with 300% loss, the sequence includes an identification zone for the length of the code word over 2 bits having the values "0" and "1" respectively, then 23 code words

  at 4 bits, the last two bits not being used.

  If the number of bits at "1" of the original sequence, that is

  say of the identification message MI, is greater than 23, a so-called permutation method is used which is lossy compression. For the permutation, the new position of a bit is calculated according to the formula: Pn = Pa * k modulo N; Pn is the new position and Pa is

the old position.

  The new position is obtained as the remainder

  modulo N of the product between the anterior position Pa and a parameter k.

  If k and N are prime to each other and if N is greater than 1024 (N must be greater than the length then the transform is reversible, which means that we can find the original sequence. For example, for N = 1031 and a k which is not a multiple of 1031, we obtain a sequence of length 1030 which has the same number of bits at

"1" as the original sequence.

  The reverse operation consists in solving, in integer values the following equation: Pa * k = x * N + Pn We know k, N and Pn, we look for an integer x for which we obtain a value of whole Pa. The value of the parameter k depends on the strategy that is used to reduce the dimension of the

  sequence and therefore compression itself.

  Compression is achieved by folding the permuted sequence. The folding is done, as in the RLC method, by dividing the sequence into several equal segments which are joined together afterwards, by a "or logical" operation. This operation guarantees the

preservation of bits at "1".

  The worst case occurs when there are no "1" bits superimposed in the folded sequence. The folded sequence will then have the same number of bits at "1" as the original sequence. Consequently, on unfolding, there will be "n" times the number bits at "1" of the original sequence. For a compression ratio of 8, this means an error of 700%. So you have to look for a permutation that

  promotes overlapping and therefore minimizes error.

  The compression algorithm is as follows: we choose N = 1031 (1031 is the first prime number greater than 1024)

  we calculate 1024 different permutations, for k = 1 ...

  1024. Each permuted sequence will have 1030 bits.

  we complete each permuted sequence up to the length of 1062 bits, by attaching zeros to the left we divide each extended sequence into 9 segments at 118 bits and we do the folding. One obtains a sequence reduced to 118 bits one calculates the error with the unfolding and one chooses the permutation which minimizes this error one keeps the corresponding value of k in the 10 bits which remain free from the coded sequence of length 128 bits for the

transmit to receivers.

  For decompression, the extended sequence is constructed by attaching 9 reduced sequences of 118 identical bits. We obtain a sequence of 1062 bits of which we retain the first 1030 bits. We do the

  reverse permutation and we keep only the first 1024 bits.

  The coded sequence then has a structure where the parameter k occupies the first ten bits and / or the sequence reduced by

  folding occupies the next 118 bits.

  The insertion means MIS of the transmitting station ST are produced for example by software within a PC type microcomputer and receiving, for example for a designated receiver on the one hand, its ADD address, its indication IDG group, IDSG family and IDRR network, and on the other hand, the data relating to the actual MMS message intended for it. These MIS insertion means then generate the headers and the useful parts DU of the various elementary blocks which will constitute the packets, and in particular those relating to the packet identification blocks, including the identifiers IC, of network signaling, and of message signaling. All these elements are then transmitted to a COD coder, for example that marketed by the Swedish company SECTRA under the reference TSE 760 which completes the formatting of these blocks, by adding in particular the error correction blocks CRC and parity. PRY. Then, all of these BL blocks, gathered in packets, are

transmitted within the SRF signal.

  In general, the functioning of the

  transmission according to the invention is briefly illustrated on

the flowchart in Figure 8.

  After switching on the receiver, it will be placed in an acquisition phase 70 preceding a time synchronization phase 71. In phase 70, the reception means of the RC receiver will first of all be fixed on the one of the transmission channels and the processing means will test on the one hand the presence on this transmission channel of a subcarrier carrying the specific transmission protocol, that is to say in this case the SWIFT protocol , and, on the other hand, the presence in this protocol of a network signaling block containing a network identifier NWD consistent with the network indication IDRR stored in the memory

of the receiver (step 700 and 701).

  If this test is negative after a test duration Tm equal to the predetermined maximum duration of separation of two consecutive network signaling blocks, in this case 0.5 seconds (step 702), the means of reception of the receiver will then search for another transmission channel (step 703) to carry out a new test until a positive result is obtained, except in the case of the use of an A frame (see SWIFT document) during the transmission of the vertical error correction blocks which are

  without network signaling blocks.

  In this regard, it can be foreseen that, if after a predetermined period of time, for example one minute, the receiver still does not find suitable transmission channels, it will start

automatically in resting state.

  In the event of a positive test, the receiver processing means will then analyze the rank identifier IR contained in the next BLTS packet identification block. In this regard, the receiver control means can maintain the latter in its active state after the analysis of the content of the acquired BLNS network signaling block until this next block is received.

BLTS packet identification.

  As a variant, it is possible to insert into each BLNS network signaling block an indication BN of the rank of this block in the packet which contains it (FIG. 2). Also, in order to further increase the autonomy of the receiver, the control means of the latter can put the receiver in its resting state after analysis of the content of this BLNS network signaling block,

  until the next BLTS packet identification block is received.

  When this BLTS block is acquired (step 704), the processing means analyze the content of this BLTS block, and in particular

  analyze the IR rank identifier it contains.

  If this IR rank identifier corresponds to the IDG group indication stored in the receiver, the receiver control means will then automatically time synchronize the passages of the receiver in its active state at successive instants separated from the initial instant of BLTS block acquisition of a multiple

integer of the frame duration.

  If the IR rank identifier does not correspond to the IDG group indication, the receiver control means will then put the latter in its idle state and then determine the occurrence of the BLTS identification block whose rank corresponds to the receiver group and then automatically synchronize the receiver alarm clocks to

  from this moment, taking into account the duration of the frames.

  The time synchronization of the successive awakenings of the

  receiver is then performed (step 71).

  In other words, all the receivers of the same group will therefore wake up, that is to say go into their active state, at each instant of reception of the packet comprising at the head the BLTS packet identification block. whose IR rank indication

corresponds to their group.

  All the receptors in the group will also analyze the

  BL2 packet identification block ET2 head, in particular

the IC identifier.

  All the receivers in the group will analyze the other bits of the identifier IC, deduce therefrom the type of decompression algorithm which must be used to retrieve the content of the identification message MI and perform decompression to know the

  families likely to receive a message.

  After analysis of the content of this MSF word, the control means place in their resting state, all the receivers whose IDSG family indications do not match the families

designated in the word MSF.

  In general, all the receivers of a family, which woke up at the start of the corresponding packet, can go into their rest state before the end of the transmission of said packet if all the messages have been acquired. This being the case, it is possible that at least some of the receivers remain active even after the end of transmission.

of their package.

  In the synchronization phase, when the BLTS block includes an indication of modification of the IM network (FIG. 3) the control means of all the receivers of the group force them to remain awake until the reception of the next signaling block of BLTS network, including those who were not designated in the MSF family word. After analysis of the BLNS network signaling block comprising the new characteristics of the transmission network, the reception means can for example be stalled

on another channel.

  We will now describe, with particular reference to FIG. 9 and FIG. 10, a particular example of addressing a receiver. It is assumed in these figures that each frame, having a duration TT, consists of three packets each having the duration TP. The packets

  are numbered from 0 to 2. There are therefore three groups of receivers.

  Each group of receptors is assumed to be subdivided into two

families numbered 0 and 1.

  Finally, it is assumed that the RC 1 1 receiver has been programmed in

  factory so as to belong to family 1 of group n 1.

  In FIGS. 9 and 10, the reference BLTS0 designates the identification block of packets of rank 0. The reference BLTS2 designates the identification blocks of packets of rank 2. The reference BLTS10 designates a identification block of packets of rank 1 designating for example furthermore only the family 0 while the reference BLTS 11l designates a block of identification of packets of rank 1 further designating the family 1. It is therefore seen in FIG. 9 that the receiver RC1 1 goes into its state active during the occurrence of the BLTS10 block, which designates group 1 to which the receiver belongs but which returns to its resting state after analysis of the content of this block because family 1

is not designated in it.

  The receiver RC1 1 returns to its active state upon reception of the BLTS block 11 and remains awake after analysis of this block since family 1 is designated. The receiver remains awake until reception of the BLMS message signaling block, in which, in this case, the time indication DFS indicates that a BLDD message is intended for the receiver RC11 and that this message intervenes in the current packet, a well-defined moment. In this case, the receiver RC1 1 returns to the idle state until the occurrence of the first block

  elementary data, of the BLC type, of the BLDD message.

  In FIG. 10, the reference BLTS20 designates the packet identification block of rank 2 further designating the family 0. The reference BLTS01 designates the packet identification block

  corresponding to group 0 and family 1.

  In this FIG. 10, it is assumed that the time indication DFS of the BLMS block indicates that the message intended for the RC receiver 11 is not transmitted in the same packet as this BLMS block but is transmitted in a subsequent packet. Under these conditions, the receiver RC 1 returns to its rest state after the analysis of the BLMS block and will then wake up at each subsequent occurrence of a packet identification block even if it does not correspond to its group. . The receiver, after analysis of this packet identification block, will remain awake if this block also contains an indication in the word MSF designating its family. On the other hand, the receiver will go back to sleep if this identification block does not include a family designation as is the case for the BLTS20 block (group 2 but

family 0).

  Also, in the present case, the receiver RC1 1 wakes up and analyzes the BLTSO1 identification block (group 0 family 1) and remains awake until the analysis of the BLMS message signaling block which indicates to it that its message properly says BLDD will be passed in this packet. The RC11 receiver goes back to sleep until reception

BLDD message.

Claims (10)

  1. Method for increasing the autonomy of digital information receivers, in which the information is   transmitted from at least one fixed station (ST) by sub-   carriers of radiofrequency channels using an asynchronous packet transmission protocol each containing a predetermined number of elementary blocks, characterized in that at least one specific channel is selected from said radiofrequency channels so as to materialize at least one specific transmission network identifiable, a transmission frame comprising a predetermined number of successive packets is defined within the transmission protocol conveyed by this network, a specific elementary packet identification block is inserted into each packet of the frame ( BLTS) containing an identifier (TR) of the rank of the packet in the frame and a useful part (DU), the useful part (DU) comprising a binary word (MSF) making it possible to designate, within the group of receivers, families of receivers, the number of families of receivers being greater than the number of bits of the binary word (M SF), the logical value of each bit of an identification message (MI) indicating whether the receivers of the corresponding family are capable of receiving a message, a group indication corresponding to a packet rank is stored in each receiver said frame and a family indication corresponding to a family rank in a packet, the identification message (MI) is analyzed in order to determine the type of compression algorithm to be used to encode the identification message (MDI) on the number of bits of the binary word (MSF),   the identification message (MDI) is compressed, then, the receiver being calibrated on said specific transmission channel, the rank identifier (TR) contained in the next packet identification block is analyzed, the receiver being calibrated on said specific transmission channel and having an idle state and an active state, the setting in the active state of each receiver is temporally synchronized with the successive occurrences of packet identification blocks (BLTS) including the rank identifiers (IR ) correspond to the group indication (IDG) stored in the receiver (RC), each receiver being able to decompress the binary word (MSF) to find the identification message (MI), the receivers that do not belong to the families identified in the state of rest and the receptors belonging to the families are maintained identified in the active state.   2. Method according to claim 1, characterized in that the identification message (MI) is compressed by truncation of at minus a bit.   3. Method according to claim 1, characterized in that the identification message (MI) is compressed by coding the   lengths of the series of successive bits having the same value.   4. Method according to claim 1, characterized in that the identification message (MI) is compressed by coding the   lengths of series of bits at "0" terminated by a bit at "1".   5. Method according to claim 3 or 4, characterized in that the identification message (MI) is compressed by segmentation, identification of the segments comprising only bits of the same value, coding of the location of said segments and coding others segments.   6. Method according to claim 3 or 4, characterized in that the identification message (MI) is divided into a number k of parts, k being greater than or equal to 2, an operator or logic is applied between the parts , then coding is performed on the sequence thus obtained, the length of which is equal to that of the message   identification (MI) divided by the number k.   7. Method according to claim 6, characterized in that the identification message (MI) is transformed by permutation prior to division and that the number k is inserted into the word binary (MSF).   8. Method according to one of the preceding claims,   characterized by the fact that an identifier (IC) of compression or not and of the type of compression chosen is inserted in a header (AND) of the block packet identification (BLTS).   9. Method according to one of the preceding claims,   characterized by the fact that the number of bits at 1 in the identification message (MI) is analyzed in order to determine the type of compression algorithm to be used to encode the identification message (MI)   on the number of bits of the binary word (MSF).   10. Digital information transmission system, comprising at least one fixed station (ST) transmitting the information by sub-carriers of radio frequency channels using an asynchronous packet transmission protocol each containing a predetermined number of elementary blocks, and several receivers (RC) comprising radio frequency reception means (1) and processing means (2) of the received data, characterized in that the station comprises generation means (MIS) capable of generating, within the transmission protocol carried by at least one specific identifiable transmission network materialized by at least one specific transmission channel selected from said radiofrequency channels, successive transmission frames each comprising a predetermined number of successive packets, insertion means (MIS) for inserting into each packet of the frame, at a predetermined location, a blo c specific packet identification (BLTS) containing an identifier (IR) of the rank of the packet in the frame, and a useful part (DU), the useful part (DU) comprising a binary word (MSF) making it possible to designate, inside the group of receivers, families of receivers, the number of families of receivers being greater than the number of bits of the binary word (MSF), the logical value of each bit of an identification message (MI) indicating whether the receivers of the corresponding family are capable of receiving a message, the station comprising means for analyzing the identification message (MI) in order to determine the type of compression algorithm to be used to encode the identification message (MI) on the number of bits of the binary word (MSF), and means for compressing the identification message (MI), and by the fact that each receiver (RC) comprises a memory containing a network indication (IDRR), an indication corresponding group (IDG) a packet rank in said frame, and an indication of the family (IDSG) to which the receiver belongs, processing means (2) having an active state and a rest state, test means capable of analyzing the identifier of rank contained in the next packet identification block and to decompress the binary word (MSF) to find the identification message (MI), and control means (3) capable of temporally synchronizing the bringing into the state active of each receiver set on said specific transmission channel, on successive occurrences of packet identification blocks (BLTS) identifying packets whose ranks correspond to the group indication stored in the receiver, to be maintained in the state activates the receivers whose family indications correspond to the identification messages (MI) and to put in the state of rest the other receptors.