Abstract: The present invention concerns channel codes particularly well adapted to transmission in channels in which errors tend to occur in bursts. Moreover, the codes according to one embodiment of the invention using an algebraic geometric curve are easy to decode and have a relatively high minimum distance. The invention also relates to the corresponding encoding and decoding methods, as well as the devices and apparatuses adapted to implement those methods. Application is in particular to mass storage, and to systems of communication by OFDM.

Abstract: In order to evaluate the noise related to data issuing from a turbo-encoder, estimates of the noise related to its systematic output, to the data issuing from its first elementary encoder, and to the data issuing from its second elementary encoder are determined. At least two of the estimates of the noise related to the systematic output, to the data issuing from the first encoder, and to the data issuing from the second encoder, are added. The results of these additions are divided by the number of augends added, and then inverted, so as to obtain a noise factor, and the noise factor is multiplied with the data issuing from at least one of the first encoder, second encoder, and systematic output.

Abstract: An original information sequence is encoded by a use of an error correcting code. There is associated with the encoded sequence K frequency symbols in a space consisting of a series of 2p increasing frequencies, each of the K symbols representing N encoded symbols, with p, K and N being strictly positive integers. There is applied to the K symbols a reversible transformation including a multiplication by an invertible matrix of size N×N, and signals obtained from the inverse transform signals are sent. There exists a K-tuplet of positive integers n1, n2, . . . , nk at least one of which is strictly positive, such that, for an integer I varying from 1 to K, after periodic extraction of one frequency out of 2ni among the frequencies of the ith of the K symbols, thus forming a reduced frequency symbol of 2p?ni frequencies, K reduced frequency symbols are obtained, representing the original information sequence, with a view to complete or partial decoding.

Abstract: In a data transmission method, said data are organized in blocks of K frames, each of said frames being composed of N binary words, each of these binary words being associated with an “elementary symbol”. Each of the M=K·N elementary symbols per block modulates a different sub-carrier amongst a series of sub-carriers forming an OFDM modulation, the allocation of each elementary symbol to a given sub-carrier being effected so that, when the OFDM signal is demodulated after it is received, a chosen subset is extracted from amongst said K frames. Application to the transmission of data representing compressed images in accordance with the JPEG2000 standard.

Abstract: The method of transmitting data in the form of symbols sent over an electromagnetic channel is characterized in that a measurement of significance is attributed to each group of data to be transmitted, in that the transmission reliability of the carriers is estimated dynamically, and in that the most significant data are sent over the most reliable carriers at each instant, the other data being sent over the carriers of decreasing reliability, in decreasing order of significance of the data.

Abstract: The invention concerns a method of adjusting at least one parameter for the compression of data representing physical quantities, the compressed data then being coded according to a coding mode in order to protect it from transmission errors. The methods includes, as from a required compressed data size, determining at least one characteristic of the coding mode, determining an effective size of the compressed data according to the required size and the at least one characteristic, and adjusting at least one compression parameter according to the effective size.

Abstract: In order to decode a received digital signal which has been watermarked with a known watermark, at least part of the digital signal is decoded by using a parametrisable iterative decoder, the watermark is extracted from the decoded signal, the extracted watermark is compared with the known watermark, and at least one parameter of the decoder is modified if necessary, according to the result of the comparison.

Abstract: The present invention concerns a method and an apparatus of generating a reference event in a receiving node receiving frames of information, when a frame of information is received, the method and the apparatus sample the frame of information in order to form sampled data, process sampled data in order to detect among said sampled data a specific data, monitor the number of processed sampled data until the detection of a specific data and generate a reference event according to the result of processing and monitoring.

Abstract: In order to evaluate the noise related to data issuing from a turbo-encoder: estimates of the noise related to its systematic output, to the data issuing from its first elementary encoder and to the data issuing from its second elementary encoder are determined (70); at least two of the estimates of the noise related to the systematic output, to the data issuing from the first encoder, and to the data issuing from the second encoder, are added (72); the results of these additions are divided (74) by the number of augends added, and then inverted (76), so as to obtain a noise factor; and the noise factor is multiplied with the data issuing from at least one of the first encoder, second encoder and systematic output.

Abstract: In order to evaluate the noise related to data issuing from a turbo-encoder, estimates of the noise related to its systematic output, to the data issuing from its first elementary encoder, and to the data issuing from its second elementary encoder are determined. At least two of the estimates of the noise related to the systematic output, to the data issuing from the first encoder, and to the data issuing from the second encoder, are added. The results of these additions are divided by the number of augends added, and then inverted, so as to obtain a noise factor, and the noise factor is multiplied with the data issuing from at least one of the first encoder, second encoder, and systematic output.

Abstract: The present invention concerns channel codes particularly well adapted to transmission in channels in,which errors tend to occur in bursts. Moreover, the codes according to one embodiment of the invention using an algebraic geometric curve are easy to decode and have a relatively high minimum distance. The invention also relates to the corresponding encoding and decoding methods, as well as the devices and apparatuses adapted to implement those methods. Application is in particular to mass storage, and to systems of communication by OFDM.

Abstract: In order to decode a received digital signal which has been watermarked with a known watermark, at least part of the digital signal is decoded by using a parametrisable iterative decoder, the watermark is extracted from the decoded signal, the extracted watermark is compared with the known watermark, and at least one parameter of the decoder is modified if necessary, according to the result of the comparison.

Abstract: In a data transmission method, said data are organized in blocks of K frames, each of said frames being composed of N binary words, each of these binary words being associated with an “elementary symbol”. Each of the M=K·N elementary symbols per block modulates a different sub-carrier amongst a series of sub-carriers forming an OFDM modulation, the allocation of each elementary symbol to a given sub-carrier being effected so that, when the OFDM signal is demodulated after it is received, a chosen subset is extracted from amongst said K frames. Application to the transmission of data representing compressed images in accordance with the JPEG2000 standard.

Abstract: An original information sequence is encoded (E1) by means of an error correcting code; there is associated (E2) with the encoded sequence K frequency symbols in a space consisting of a series of 2p increasing frequencies, each of the K symbols representing N encoded symbols, p, K and N being strictly positive integers; there is applied (E3) to the K symbols a reversible transformation including a multiplication by an invertible matrix of size N×N; and signals obtained from the inverse transform signals are sent (E4). There exists a K-tuplet of positive integers n1, n2, . . . , nK at least one of which is strictly positive, such that, for an integer i varying from 1 to K, after periodic extraction of one frequency out of 2ni amongst the frequencies of the ith of the K symbols, thus forming a reduced frequency symbol with 2p-ni frequencies, K reduced frequency symbols are obtained, representing the original information sequence, with a view to complete or partial decoding.

Abstract: The invention concerns a method of adjusting at least one parameter for the compression of data representing physical quantities, the compressed data then being coded according to a coding mode in order to protect them against transmission errors, characterised in that it includes, as from a required compressed data size (RT), the steps of:

Abstract: In order to evaluate the noise related to data issuing from a turbo encoder: estimates of the noise related to its systematic output, to the data issuing from its first elementary encoder and to the data issuing from its second elementary encoder are determined (70); at least two of the estimates of the noise related to the systematic output, to the data issuing from the first encoder, and to the data issuing from the second encoder, are added (72); the results of these additions are divided (74) by the number of augends added, and then inverted (76), so as to obtain a noise factor; and the noise factor is multiplied with the data issuing from at least one of the first encoder, second encoder and systematic output.