Abstract: A method of encoding a bitstream includes obtaining at least one current input word of the bitstream to encode, determining at least one nominal codeword associated with the at least one input word according to an error-control code, selecting one of the at least one nominal codeword or a substitute codeword in order to keep a current running digital sum and/or a running alternate sum bounded, and outputting an encoded word comprising the selected codeword. Embodiments make it possible to generate an encoded bitstream which is “DC free” and “Nyquist free” while providing error correction with a fixed coding rate.

Abstract: Signal processing methods, signal processing device(s), signal processing system(s) and non-transitory information storage means are provided herein for shaping the spectrum of an input bitstream. In at least one embodiment, a signal processing method includes performing a plurality of line coding steps on respective parts of an input bitstream to obtain a plurality of respective line-coded signals, and multiplexing the plurality of line-coded signals to obtain a spectrum-shaped output signal. Embodiments provide a modulation scheme that makes it possible to shape the spectrum of a bitstream to be transmitted in order to match a channel transfer function of a communication medium.

Abstract: A method of configuring a beam forming antenna in a communication network that comprises a first node where the beam forming antenna is located, second nodes and at least one destination node, communication links being established between said first node and said at least one destination node through at least some of said second nodes. The method comprises: emitting a signal by the beam forming antenna configured with a first set of antenna parameters; the same signal being sent from the first node to several second nodes; obtaining, for a plurality of communication links through which the signal has been sent, at least one physical magnitude representing the distortion caused by each communication link to the signal; aggregating said physical magnitudes of said plurality of communication links; and obtaining a second set of antenna parameters for configuring the beam forming antenna in accordance with said aggregated physical magnitudes.

Abstract: A method and device for decoding a plurality of packets of same data. Each packet is encoded using a product code. An encoded packet comprises data blocks is organized into rows and columns. Each of the rows and columns has a respective error detecting code. For a data packet, Erroneous row and column data are detected by their associated error detecting code. In the data packet, a potentially erroneous data block belonging to a row and a column of data blocks both having been detected as erroneous is detected. In another data packet of the same data, a non-erroneous data block corresponding to the identified potentially erroneous data block is determined. The data block is determined as non-erroneous if it belongs to at least one row or column of data having been detected as non-erroneous. The identified potentially erroneous block is replaced with the determined non-erroneous block.

Abstract: A method for processing a plurality of received copies of the same original data, the method comprising performing maximum ratio combining on equivalent data portions of the received copies to derive at least one further equivalent data portion. The received and the constructed copies are then divided into sub-portions and combinations of divided sub-portions are assembled to provide reconstructed data portions. An error rate assessment is performed on the reconstructed data portions; and on this basis one is selected as an output.

Abstract: The invention relates to a method of reconstructing pixel values of a video frame for concealing corrupted pixel values. The method comprising receiving, by a receiving unit, of a signal from a communication channel and delivering by the receiving unit of video packets comprising pixel values possibly corrupted with errors; associating confidence levels with pixel values comprised in the video packets as delivered by the receiving unit; and reconstructing pixel values usable for display from the received pixel values, wherein a reconstructed value for a given pixel is obtained from the received values of a set of pixels, including the given pixel, weighted by their associated confidence levels. The invention allows for better reconstruction of corrupted pixel values and reduces the perceived distortion when displaying the video frame.

Abstract: A method of encoding a bitstream includes obtaining at least one current input word of the bitstream to encode, determining at least one nominal codeword associated with the at least one input word according to an error-control code, selecting one of the at least one nominal codeword or a substitute codeword in order to keep a current running digital sum and/or a running alternate sum bounded, and outputting an encoded word comprising the selected codeword. Embodiments make it possible to generate an encoded bitstream which is “DC free” and “Nyquist free” while providing error correction with a fixed coding rate.

Abstract: Improving decoding of a set of k data symbols received from several receivers, the data symbols being encoded by a systematic block error correcting code of dimension k and size n. The set of data symbols is received along with a corresponding subset of parity symbols, forming a partial data block comprising m symbols. A partial data block transmitted by one emitter, comprising a set of k data symbols and a subset of (m?k) parity symbols, is received from each receiver. For each received partial data block, a subset of parity symbols is generated and an item of reliability information is computed as a function of the received parity symbols and parity symbols generated from a received set of data symbols. The items of computed reliability information are compared with each other to select one received set of data bits.

Abstract: A computer implemented method of controlling the decoding of codewords received by a linear block code pipelined decoder from an input buffer, the pipelined decoder comprising at least two decoding stages. The method comprises iteratively: loading the decoding stages of the pipelined decoder, executing a decoding step, determining the number of residual errors in the codewords and outputting error free codewords. The method allows the different decoding stages to be loaded with any codeword coming from the buffer or from any decoding stage of the decoder. Accordingly, the occupation rate of the pipeline is improved.

Abstract: Signal processing methods, signal processing device(s), signal processing system(s) and non-transitory information storage means are provided herein for shaping the spectrum of an input bitstream. In at least one embodiment, a signal processing method includes performing a plurality of line coding steps on respective parts of an input bitstream to obtain a plurality of respective line-coded signals, and multiplexing the plurality of line-coded signals to obtain a spectrum-shaped output signal. Embodiments provide a modulation scheme that makes it possible to shape the spectrum of a bitstream to be transmitted in order to match a channel transfer function of a communication medium.

Abstract: A method and device for decoding a plurality of packets of same data. Each packet is encoded using a product code. An encoded packet comprises data blocks is organized into rows and columns. Each of the rows and columns has a respective error detecting code. For a data packet, Erroneous row and column data are detected by their associated error detecting code. In the data packet, a potentially erroneous data block belonging to a row and a column of data blocks both having been detected as erroneous is detected. In another data packet of the same data, a non-erroneous data block corresponding to the identified potentially erroneous data block is determined. The data block is determined as non-erroneous if it belongs to at least one row or column of data having been detected as non-erroneous. The identified potentially erroneous block is replaced with the determined non-erroneous block.

Abstract: Improving decoding of a set of k data symbols received from several receivers, the data symbols being encoded by a systematic block error correcting code of dimension k and size n. The set of data symbols is received along with a corresponding subset of parity symbols, forming a partial data block comprising m symbols. A partial data block transmitted by one emitter, comprising a set of k data symbols and a subset of (m?k) parity symbols, is received from each receiver. For each received partial data block, a subset of parity symbols is generated and an item of reliability information is computed as a function of the received parity symbols and parity symbols generated from a received set of data symbols. The items of computed reliability information are compared with each other to select one received set of data bits.

Abstract: A computer implemented method of controlling the decoding of codewords received by a linear block code pipelined decoder from an input buffer, the pipelined decoder comprising at least two decoding stages. The method comprises iteratively: loading the decoding stages of the pipelined decoder, executing a decoding step, determining the number of residual errors in the codewords and outputting error free codewords. The method allows the different decoding stages to be loaded with any codeword coming from the buffer or from any decoding stage of the decoder. Accordingly, the occupation rate of the pipeline is improved.

Abstract: A method for processing a plurality of received copies of the same original data, the method comprising performing maximum ratio combining on equivalent data portions of the received copies to derive at least one further equivalent data portion. The received and the constructed copies are then divided into sub-portions and combinations of divided sub-portions are assembled to provide reconstructed data portions. An error rate assessment is performed on the reconstructed data portions; and on this basis one is selected as an output.

Abstract: A method is provided for configuring an overall encoding scheme comprising a first encoding for generating first encoded packets from source packets and a second encoding for generating second encoded packets by combination of first encoded packets according to a combination scheme applied by predetermined nodes of the network. The method obtains a number N of defective paths among paths used to transmit the first and second encoded packets to at least one destination node, a path being defective if an associated quality of transmission is below a predetermined threshold. A combination scheme is selected from among at least two possible predetermined combination schemes for which a number of encoded second packets is greater than or equal to N.

Abstract: A method for assessing reception quality of a common transmission by a set of receivers comprising deriving a reception quality for different combinations of the set, and assessing each against a threshold which varies according to the number of receivers in the respective combination. A single assessment of quality can be produced which takes into account multiple different receivers, and multiple different combinations of such receivers. Such assessment allows transmission power and/or the number of active receivers to be set according to a given reception criteria.

Abstract: A method is proposed for decoding a plurality of data packets received through a meshed communications network for the retrieval of source packets transmitted by one or more source nodes, the meshed communications network comprising relay nodes generating combined packets, each combined packet consisting of a linear combination of source packets. During the decoding by a destination node, this method consists in performing two decoding operations of which the first is a decoding by group of packets and the second is a decoding that takes account of pieces of likelihood information resulting from the first decoding. The fact of first of all carrying out a decoding with groups of packets makes it possible to exploit the repetitions (or redundancies) of packets in a meshed network to optimize the bit error rate during the decoding of source data transmitted on the communications network.

Abstract: One embodiment of the invention pertains to a method for allocating bandwidth in a communications network, characterized in that it comprises steps for: a) determining a subset of elementary networks from a set of possible elementary networks in a communications network, the elementary networks of the subset each being capable of simultaneously implementing a elementary network encoding scheme; b) determining at least one critical destination node in the communications network; c) allocating at least one bandwidth portion released by an application of the elementary network encoding scheme in at least one elementary network of the subset, for a transmission of at least one additional data unit to the at least one critical destination node.

Abstract: A method is proposed for determining an erasures vector associated with a data block to be decoded built out of received copies, and using levels of reliability of transmission associated with symbols contained in different copies of a same block of received pieces of data. Such a method advantageously makes reduces the error rate at output of a decoder by minimizing the operations to be performed to determine the erasures vector, marking the erasures and preventing the generation of an excessively large number of erasures (even when the transmission of certain received copies is unreliable) to have a number of erasures that does not exceed the correction capacity of the decoder.

Abstract: A method for managing allocation of paths for transmission, in packets of symbols, of at least a first data content, each content being transmitted from a source node up to a destination node in a meshed communications network comprising a plurality of nodes, each destination node receiving a plurality of copies of a same packet of symbols of the content transmitted to said destination node by a source node, each copy being received by the destination node via a distinct transmission path directly from the source node or via at least one relay node of the meshed communications network, the packets being encoded with an error correction code enabling decoding by erasure of packets symbols.