EP1131928A1 - Block joint source-channel coding - Google Patents

Block joint source-channel coding

Info

Publication number
EP1131928A1
EP1131928A1 EP99943017A EP99943017A EP1131928A1 EP 1131928 A1 EP1131928 A1 EP 1131928A1 EP 99943017 A EP99943017 A EP 99943017A EP 99943017 A EP99943017 A EP 99943017A EP 1131928 A1 EP1131928 A1 EP 1131928A1
Authority
EP
European Patent Office
Prior art keywords
constellation
channel
source
points
phase
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP99943017A
Other languages
German (de)
French (fr)
Inventor
Philippe Leclair
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Comsis
Original Assignee
Comsis
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Comsis filed Critical Comsis
Publication of EP1131928A1 publication Critical patent/EP1131928A1/en
Withdrawn legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
    • H03M13/63—Joint error correction and other techniques
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
    • G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
    • G10L19/08—Determination or coding of the excitation function; Determination or coding of the long-term prediction parameters
    • G10L19/10—Determination or coding of the excitation function; Determination or coding of the long-term prediction parameters the excitation function being a multipulse excitation
    • G10L19/113—Regular pulse excitation
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
    • H03M13/25—Error detection or forward error correction by signal space coding, i.e. adding redundancy in the signal constellation, e.g. Trellis Coded Modulation [TCM]
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
    • H03M13/65—Purpose and implementation aspects
    • H03M13/6577—Representation or format of variables, register sizes or word-lengths and quantization
    • H03M13/6594—Non-linear quantization
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/18—Phase-modulated carrier systems, i.e. using phase-shift keying
    • H04L27/183—Multiresolution systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/32—Carrier systems characterised by combinations of two or more of the types covered by groups H04L27/02, H04L27/10, H04L27/18 or H04L27/26
    • H04L27/34—Amplitude- and phase-modulated carrier systems, e.g. quadrature-amplitude modulated carrier systems
    • H04L27/3488—Multiresolution systems

Definitions

  • the invention relates generally to coding with data compression.
  • data compression is motivated by more efficient management of the channel resource during data transmission or by a reduction in the size of the memory in a data logger.
  • Data compression is based on an extraction of the redundancy of the symbols to be transmitted according to a predetermined law.
  • the redundancy is added to the reception, during the reconstruction of the data.
  • the compression is said to be without distortion.
  • some distortion is accepted in the reconstruction process.
  • image coding for example, transform coding achieves fairly low coding rates.
  • speech coding analysis and synthesis coding achieves low coding rates.
  • Transform compression provides very low coding rates with reduced distortions. Compression is then achieved through a representation of the information, and discrimination is made through an unequal distribution of the frequency components. In general, the goal of any source coding is precisely the discovery of this structure in the information to be coded.
  • the non-repeat information becomes more sensitive to the effects of noise and consequently the performances compared to the model without noise deteriorate quickly.
  • the invention finds applications in very numerous fields, in particular in the coding of vo x and the coding of images.
  • digitized speech is transmitted at low bit rates expressed in kbit / s, with correct reconstruction quality.
  • the digitized image is transmitted at low bit rates, expressed in tens, hundreds or even thousands of kbit / s, and this with a correct reconstruction quality.
  • the field of images always requires considerable amounts of information to manage.
  • the invention relates more particularly to block coding according to which the source and channel coding operations are integrated into a single entity.
  • This coding technique is called joint source-channel coding.
  • the implementation of this technique is based on a redefinition of the quantification of the coefficients of the source linked to the property of bi] action between the no ore ⁇ e levels of the quantifier and the number of states of the modulations used.
  • compression we always refer to two algorithms: the first generates a representation of the source in the form of a limited number of bits; - the second reconstructs the source from the representation bits, possibly tainted with errors.
  • the compression schemes are classified into lossless schemes and lossy schemes.
  • the lossless or distortion-free diagrams are diagrams where the reconstruction of the data on decoding gives a source identical to the source to be transmitted. This is typical of compressing written text or computer generated information.
  • the perfect reconstruction constraint cannot be verified. Distortion between the original and the reconstructed is tolerated to obtain a higher compression ratio.
  • the reconstructed voice signal samples are not necessarily the same as those of the transmitted signal.
  • a certain degree of distortion is tolerated without compromising the understanding of the message sent.
  • the higher the compression ratio the greater the loss introduced by the coding process.
  • Group 1 Joint Source-Channel Codings according to which the source and channel coding operations are integrated into a single entity.
  • Group 2 Concatenated Source-Channel Encodings whereby a given source encoder is concatenated to a given channel encoder and the source and channel encoding bit rates are determined, so as to maximize the performance of the set.
  • Group 3 Joint Codings with Constraints according to which the coder and / or the decoder are modified so as to take into account channel errors; for example, a source encoder optimized for a noise-free channel is re-optimized to take into account the statistics of the channel.
  • the main objective of the invention is to provide a joint source-channel coding of group 1 in particular for the coefficients at the output of a vocoder, by avoiding adding redundancy by means of a code for correcting errors, as in a group 2 coder, and to be very sensitive to the slightest variation in channel characteristics, as with a group 3 coder.
  • a joint source-channel coding comprising a quantization and a phase modulation with a constellation of NC points
  • the constellation of the pnase modulation is a rotated constellation deduced from a predetermined rotation of a uniform phase constellation at NC state points so that the NC points of the constellation are projected onto any one of the axes of the constellation in NC all different projections
  • the quantization presents a non-uniform quantization law of NS levels which are NS projections among NC projections different from the constellation rotated.
  • the numbers NC and NS are integers, with NS ⁇ NC.
  • the constellation of NC points is a phase modulation constellation, or an amplitude modulation constellation
  • the invention also relates to a speech or image coder whose quantization means, also called coders, producing coefficients and parameters to be multiplexed in frames are joint source-channel coders implementing conforming joint source-channel codings to the invention.
  • the joint source-channel coders for long-term analysis filter gain coefficients, excitation grid position parameters, and short predictor filter coefficients high rank terms include QAM-type phase constellation modulators, and joint source-channel coders for short-term short-term predictor filter coefficients, long-term analysis filter delay coefficients, and amplitude block parameters include phase constellation modulators
  • the modulators with phase constellation of MAQ and MAQ types are respectively preceded by quantifiers having non-uniform quantization laws the levels of which are respectively subsets of projections relating to points of a constellation of phase predetermined turned to at least four dimensions.
  • the predetermined tour phase constellation is that resulting from a rotation of the constellation of the
  • the invention quantifies the coefficients at the output of the vocoder to transform them into samples taking a finite number of possible values. These samples were chosen to take into account the disturbances of the channel. In addition, this coding does not show sensitive bits as is the case with conventional tandem techniques. Thus, the resistance of the coding of the invention to channel noise is greater.
  • the vocoder of the invention provides a compressed speech signal resistant to disturbances of the transmission channel, allowing the production of transmitters and receivers with complexity and limited cost, in particular as regards coding and decoding.
  • the bandwidth of the transmission channel is reduced by several units compared to the prior art.
  • FIG. 1A and 1B are diagrams of phase constellations of an MDP4 modulation not turned and turned respectively;
  • FIG. 2 is a diagram showing the transformation between a constellation ⁇ e channel corresponding to the phase modulation MDP4 turned and a source constellation according to the invention;
  • - Figure 3 is a schematic block diagram of a joint source-channel encoder and a joint channel-source ⁇ ecoder according to the invention connected by a faded transmission channel;
  • FIGS. 4A and 4B are diagrams of phase constellations of an MAQ-16 modulation not turned and turned respectively;
  • FIG. 5 is a block diagram of a speech coder according to the prior art for cellular GSM mobile radio system
  • FIG. 6 is a block diagram of a joint source-channel coder in blocks according to the invention for a cellular GSM mobile radio system.
  • coefficients at the output of an encoder are quantified according to a joint code in blocks.
  • This quantization scheme labels the coefficients and output of the vocoder, called subsequently sample, not by bits, but by symbols which are transmitted directly on the channel.
  • h (t) ' denotes the pulse response ⁇ of the transmission filter and T is the inverse of the modulation speed expressed in bauds
  • the symbols a ⁇ are directly transmitted on the channel and represent all or part of a quantified sample from the vocoder.
  • the relation that there is between the quantized samples of the vocoder and the symbols transmitted has, identifies the coding in blocks.
  • the code is the set of all the possible symbol sequences a k that can be transmitted on the channel.
  • the encoder transforms samples from the source not yet quantified, or more finely quantified, into code words. For this, the coder searches for the code word closest to the sequence of samples sent by the vocoder according to a certain predetermined distance criterion. Very often, this distance criterion is the Euclidean distance between the sample leaving the voco ⁇ eur and the symbol of the source. These source symbols are then transformed into channel symbols a.
  • the channel ⁇ istor ⁇ these co ⁇ e words by adding to the noise, or by introducing interference between symbols, for example.
  • the decoder then reconstructs the vocoder samples from the noisy observations. For this, he chooses the code word located - at the shortest distance according to a certain criterion of the word received. Then, it associates the quantified sample which corresponds to it. Finally, thanks to the coefficients of the reconstructed vocoder, the voice decoder reconstructs the spoken word.
  • the invention implements the following joint source-channel coding in the particular mode of joint source-channel block codes.
  • the coding In joint coding, during digital transmission through fading channels, such as radio channels for mobile radiotelephones, the coding consists of passing from a source code, called source dictionary, composed of points of a space a DS dimensions to a channel code, called a channel dictionary, composed of the same number of points, ma s in a space with DC dimensions. Assuming that DC> DS, the points of the source dictionary in a number equal to the points of the channel dictionary must all be different, but in a space of smaller dimension than the channel space. For transmission in a fading channel, the dimensions where the fading takes place are precisely similar to the missing dimensions of the source dictionary. The resolution of the problem of transmission in a Rayleigh channel calls upon the notion of diversity which consists in distributing the information in the greatest possible number of components of the code word, so as to recover it in the components which are not not fainting.
  • the approach followed by the invention is the reverse of the classic spouse approach. Diversity is used when going back from the channel to the source.
  • the invention introduces the notion of turned constellations. Modulation techniques using point arrays have become very powerful tools for the design of digital transmission systems with high spectral efficiency both for a Gaussian channel and for a fading channel, and in particular for a Rayleigh channel.
  • a diversity order L of a multidimensional constellation is the minimum number of different components between any two points of the constellation.
  • diversity is designated by L.
  • To introduce diversity into a multidimensional constellation it is rotated so that all pairs of points have a maximum number of different components.
  • each point of the uniform phase constellation has one of its coordinates equal to another point of the constellation, and in FIG. 1B, each point of the constellation turned to each of its two coordinates , XI, X2 different from the corresponding coordinates of the other points of the constellation being rotated.
  • the turned constellation offers an additional degree of protection.
  • the rotated constellation gives an gain of 8 dB compared to the unpurned MDP4 constellation.
  • DC is an integer such that DC> 2
  • the principle is the same.
  • a block coded modulation seen as a finite subset of a network of points is "rotated" by a rotation at dimensional DC so as to have the greatest possible diversity up to DC. In this case, the expected gains are even greater than those for a 2-dimensional constellation.
  • the invention establishes an application between the levels resulting from a source coding and the points of the channel constellation in order to obtain the desired robustness in terms of distortion of the source reconstructed on reception.
  • the joint source-channel block coding of the invention exploits the diversity of the constellations turned to find this application.
  • Source coding is derived from channel coding.
  • the invention applies the property that, if a code comes from a constellation turned to maximum diversity, which is equal to the dimension of the constellation, then the projections on any one of the axes of the constellation constitute a set with a cardinal equal to that of the code.
  • This property is the basis of joint block coding according to the invention, and is deduced from the very definition of the diversity of a constellation.
  • the source dictionary is constituted by the points or levels indicated by a cross on the axis 0-X1, while the channel dictionary consists of the points [XI, X2] of the constellation MDP4 tour.
  • the method of the invention is advantageous in that it adapts the scalar quantizer to the distribution of the source by simple modification to the angle of rotation.
  • FIG. 2 an example of a quantification scale for a source sample is shown in thick lines.
  • Figure 3 shows an encoder 1 in a transmitting device and a decoder 2 in a receiving device according to the invention, connected through a fading channel 3.
  • the coder 1 comprises a source 11, a joint source-channel coding circuit in blocks 12 and an interleaver 13.
  • the source 11 is for example a vocoder.
  • the coding circuit 12 includes a scalar quantizer 121 followed by a modulator
  • the scalar quantizer is associated with a programmable circuit 123 which establishes a non-uniform quantization lo according to the source dictionary as shown at the bottom of FIG. 2.
  • the quantization lo is chosen as a function of the initial constellation (FIG. 1A) of the modulator 122 after a rotation thereof with a predetermined angle ⁇ in order to maximize the cardinality of the points projected on a predetermined axis O-Xl of the constellation CC channel.
  • Each sample provided by the vocoder 11 and representing for example a parameter of the vocoding model implemented in the vocoder 11 is quantified in the quantizer 121 according to the quantization law called by rounding, by corresponding to the level or point of the source constellation CS , closest to the sample.
  • the modulator 122 converts by "reverse" projection each point of the constellation CS transmitted by the quantifier into the corresponding symbol of the channel constellation CC which, according to FIG. IB or 2, is defined by Cartesian coordinates XI (t) and X2 (t), or polar, amplitude p (t) and phase ⁇ (t) that are different from the coordinates of all the other symbols of the constellation CC.
  • the interleaver 13 interleaves the symbols produced by the modulator 122 to introduce time diversity therein in a known manner.
  • the transmission channel 3 is shown diagrammatically in FIG. 3 by a generator 31 generating multiplicative noise due to fading and a generator
  • a deinterleaver 21 deinterleaves the interleaved symbols according to the initial order, and an estimation circuit 22 estimates the amplitude and the instantaneous phase of the channel.
  • a demodulator 23 demodulates each deinterleaving symbol in points of the source constellation CS, which had been emitted, by carrying out the transformation of the constellations CC to CS of FIG. 2 by projection on the predetermined axis O-Xl.
  • each set of projections resulting from the projection of the points ⁇ e the constellation MAQ-16 on one of the axes 01, OQ before application of the rotation matrix has a cardinal of 4, i.e. there is no has only 4 different component values relative to each axis 01, OQ for the 16 points.
  • the rotation and projection operations for designing a scalar quantifier with a bijection between the points of the channel constellation and the points or levels of the linear source constellation is a characteristic of the invention.
  • the invention is applied to the full-rate speech coder included in a base station or in a mobile station of the cellular mobile radio network according to the GSM standard.
  • the base station is connected to the switched telephone network via ISDN 64 kbit / s channels
  • the speech signal initially sampled at 8 kHz on 8 bits is compressed into a speech signal at 13 kbit / s to be transmitted in a radio channel.
  • the coder COP essentially comprises a segmentation circuit SEG for segmenting a speech signal SP at 64 kbit / s, an analysis filter LPC (Lmear Predictive Coding) with linear predictor PL, a long-term prediction analysis_ filter LTP (Long Term prediction) and an RPE excitation signal calculation circuit
  • the LPC short-term predictor filter models variations in the short-term vocal tract.
  • the LPC predictor generates eight LAR (Log Area) coefficients
  • LAR (2) and LAR (3) 5-bit quantization
  • LAR (4) and LAR (5) 4-bit quantization
  • LAR (6) and LAR (7) quantization on 3 Dits, i.e. 6 bits.
  • the LTP long-term analysis filter models rapid variations in the vocal tract. It provides an LTP-Lag delay coefficient and an LTP-Gam gain coefficient, per sub-frame of 5 milliseconds, ie a speed of 1.8 kbit / s; the long-term coefficients on a 20 ms frame are: LTP-Lag: quantization on 7 bits, ie 4 x 7
  • the RPE calculation circuit produces an excitation signal consisting of a block of pulses which are regularly distributed over time and coded at 9.4 kbit / s in position and amplitude by means of a grid.
  • the excitation signal is composed of gate position parameters Mi at the output of a low-pass filter FBP, ⁇ e parameters ⁇ e amplitude block X axi, and values relating to _13 pulses x ⁇ (0) to x ⁇ (12 ) by subframe î of 5 ms.
  • Excitation pulses x ⁇ (0) to x ⁇ (12): 3-bit quantization of the 13 pulses, i.e. 4 x 39 156 bits.
  • the above coefficients and parameters are multiplexed in an MX multiplexer at the output of the encoder to produce frames at 260 bits every 20 ms, ie at the coding rate of 13 kbit / s. Then in a COC channel coder followed by an interleaver, these bits are separated into three different relative importance classes. The bits of the first two classes are protected by different error correcting codes, and the bits of the last class, the least important, are concatenated without protection.
  • the channel encoder delivers a 456-bit code unit of data every 20 ms, a rate of 22.8 kbit / s. This selective protection by class is known as an Unequal Protection technique against errors.
  • a joint source-channel speech coder in blocks COPa in accordance with the invention for GSM network comprises the functional circuits SEG, PLa, LPCa, LTPa, FPB and RPEa of the coder COP, but with the following changes.
  • the channel coder COC is deleted, and the sets of quantizers, also called coders, QUI, QU2_ and QU3 at the output of the PL, LTP and RPE circuits are replaced by joint source-channel coding circuits in blocks. of the type of that 12 described with reference to FIG. 3 in the circuits PLa, LTPa and RPEa, with the exception of the pulse output of the RPEa calculation circuit.
  • the decoders (not shown) in the LPC, LTP and RPE circuits respectively performing the inverse functions of the coders (quantifiers) in the PL, LTP and RPE circuits are replaced by joint channel-source decoding circuits of the type of that 23-24 shown in Figure 3.
  • the modulation is carried out on 16 or 32 levels for example.
  • An MAQ -64 constellation is provided for each of the low rank coefficients LAR (0) and LAR (l)
  • the MAQ -256 constellation is the Cartesian product of two MAQ-l ⁇ constellations.
  • the process of the invention can be generalized. Instead of considering a projection of a CC channel constellation of dimension 2 or 4 towards a source constellation CS of dimension 1, we can consider in the same way, the projection of a channel constellation of dimension DC towards a source constellation of dimension DS, with DODS.
  • the source dictionary and the channel dictionary are chosen to be identical.
  • the task of the channel code is to provide the necessary protection to prevent loss in the quality, during transmission.
  • the bits which result from the quantization of the parameters are conventionally coded by analog-digital conversion into binary code words and then coded in the channel coder to be transmitted in the channel. The separation has become almost universal between source coding and channel coding.
  • the method of the invention performs all of the source coding and the channel coding differently.
  • the parameters to be transmitted are quantified in a non-uniform scalar quantizer which results from the projection of a rotated channel constellation.
  • the DC dimension depends on the degree of importance of the coefficients to be transmitted. The higher the coefficient, the greater the DC.
  • the coefficients of less importance correspond to scalar channel constellations, while the others will correspond to channel constellations with 2, 4 or even 8 dimensions for the most important coefficient.
  • the performance of the coding of the invention is illustrated in terms of reduction in bandwidth for an auoio quality identical to that existing at present.
  • Another approach is to present joint source-channel block coding as a means of building a communications system . which, for a fixed bandwidth, is capable of transmitting with a lower average power.
  • the coding of the invention is then advantageous, for example in communication systems by satellites where the smallest decibel gained is very significant on the cost of the payload.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Probability & Statistics with Applications (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computational Linguistics (AREA)
  • Nonlinear Science (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Compression, Expansion, Code Conversion, And Decoders (AREA)

Abstract

The invention concerns joint source-channel coding comprising quantization and phase modulation with a constellation of NC points. The constellation of the phase modulation is a rotated constellation (CC) derived from a predetermined rotation of a uniform phase constellation with NC points such that the NC points of the constellation are projected on any one of the axes (O-X1, O-X2) of the constellation in NC projections all different from one another. The quantization provides a non-uniform quantization law (CS) of NS levels which are NS projections among NC different projections of the rotated constellation (CC). Such coding can be used in particular in a speech encoder for mobile radio communication cellular system to reduce transmission channel band width.

Description

Codage source-canal conjoint en blocs Joint source-channel coding in blocks
L'invention concerne de manière générale le codage avec compression de données.The invention relates generally to coding with data compression.
En général, la compression de données est motivée par une gestion plus efficace de la ressource canal lors d'une transmission de données ou par une réduction de la taille de la mémoire dans un enregistreur de données. La compression de données est basée sur une extraction de la redondance des symboles à transmettre selon une loi prédéterminée.In general, data compression is motivated by more efficient management of the channel resource during data transmission or by a reduction in the size of the memory in a data logger. Data compression is based on an extraction of the redundancy of the symbols to be transmitted according to a predetermined law.
La redondance est rajoutée à la réception, lors de la reconstruction des données. Lorsque la reconstruction est faite sans erreur, la compression est dite sans distorsion. Dans le but d'atteindre des taux de compression plus élevés, une certaine distorsion est acceptée dans le processus de reconstruction. Pour le codage des images, par exemple, le codage par transformée permet d'atteindre des taux de codage assez faibles. Pour le codage de la parole, le codage par analyse et synthèse permet d'atteindre des taux de codage faibles.The redundancy is added to the reception, during the reconstruction of the data. When the reconstruction is done without error, the compression is said to be without distortion. In order to achieve higher compression rates, some distortion is accepted in the reconstruction process. For image coding, for example, transform coding achieves fairly low coding rates. For speech coding, analysis and synthesis coding achieves low coding rates.
Une compression par transformée confère des taux de codage très faibles avec des distorsions réduites. La compression est alors atteinte grâce à une représentation de l'information, et la discrimination est faite par l'intermédiaire d'une répartition inégale des composantes en fréquence. En général, le but de tout codage de source est justement la découverte de_ cette structure dans 1 ' information a coder .Transform compression provides very low coding rates with reduced distortions. Compression is then achieved through a representation of the information, and discrimination is made through an unequal distribution of the frequency components. In general, the goal of any source coding is precisely the discovery of this structure in the information to be coded.
D'une façon générale, plus on comprime l'information, moins on utilise le canal et par conséquent moins on expose l'information aux effets du bruit.In general, the more you compress the information, the less you use the channel and therefore therefore the less the information is exposed to the effects of noise.
Mais, dans le cas d'une utilisation du canal moins fréquente, l'information non reαondante devient plus sensible aux effets du bruit et par conséquent les performances par rapport au modèle sans bruit se dégradent rapidement.But, in the case of a less frequent use of the channel, the non-repeat information becomes more sensitive to the effects of noise and consequently the performances compared to the model without noise deteriorate quickly.
Il est important de quantifier le degré de sensibilité aux erreurs de transmission.It is important to quantify the degree of sensitivity to transmission errors.
L' invention trouve des applications dans de très nombreux domaines, notamment dans le codage de la vo x et le codage des images.The invention finds applications in very numerous fields, in particular in the coding of vo x and the coding of images.
Pour le codage de la voix, la parole numérisée est transmise à de faibles débits exprimes en kbit/s, avec une qualité de reconstruction correcte.For voice coding, digitized speech is transmitted at low bit rates expressed in kbit / s, with correct reconstruction quality.
Pour le codage des images, l'image numérisée est transmise à de faibles débits, exprimes en dizaines, centaines voire milliers de kbit/s, et ceci avec une qualité de reconstruction correcte. Par contre, le domaine des images exige toujours des quantités d'information considérables a gérer.For image coding, the digitized image is transmitted at low bit rates, expressed in tens, hundreds or even thousands of kbit / s, and this with a correct reconstruction quality. On the other hand, the field of images always requires considerable amounts of information to manage.
L'invention concerne plus particulièrement le codage en blocs selon lequel les opérations de codage de source et de canal sont intégrées dans une seule entité. Cette technique de codage est appellée codage source-canal conjoint.The invention relates more particularly to block coding according to which the source and channel coding operations are integrated into a single entity. This coding technique is called joint source-channel coding.
La mise en oeuvre de cette tecnnique s'appuie sur une redéfinition de la quantification des coefficients de la source liée à la propriété de bi] action entre le no ore αe niveaux du quantificateur et le nombre d'états des modulations employées . En matière de compression, on fait toujours référence à deux algorithmes : le premier génère une représentation de la source sous la forme d'un nombre limité de bits ; - le deuxième reconstruit la source a partir des bits de la représentation, éventuellement entaches d' erreurs .The implementation of this technique is based on a redefinition of the quantification of the coefficients of the source linked to the property of bi] action between the no ore αe levels of the quantifier and the number of states of the modulations used. In terms of compression, we always refer to two algorithms: the first generates a representation of the source in the form of a limited number of bits; - the second reconstructs the source from the representation bits, possibly tainted with errors.
Selon les contraintes imposées, les schémas de compression sont classés en schémas sans pertes et schémas avec pertes.According to the constraints imposed, the compression schemes are classified into lossless schemes and lossy schemes.
Les schémas sans pertes ou sans distorsion sont des schémas où la reconstruction des données au décodage donne une source identique a la source a transmettre. C'est le cas typique de la compression de texte écrit ou de l'information générée par ordinateur .The lossless or distortion-free diagrams are diagrams where the reconstruction of the data on decoding gives a source identical to the source to be transmitted. This is typical of compressing written text or computer generated information.
Dans certaines situations, la contrainte de reconstruction parfaite ne peut être vérifiée. Une distorsion entre l'original et le reconstruit est tolérée pour obtenir un taux de compression plus élevé .In certain situations, the perfect reconstruction constraint cannot be verified. Distortion between the original and the reconstructed is tolerated to obtain a higher compression ratio.
L'application la plus fréquente de ce genre de technique est le codage de la voix. Dans ce cas, les échantillons du signal de voix reconstruits ne sont pas nécessairement les mêmes que ceux du signal émis. Un certain degré de distorsion est tolère sans que la compréhension du message émis soit altérée. Intuitivement, plus le taux de compression est élevé, plus grande est la perte introduite par le processus de codage .The most common application of this kind of technique is voice coding. In this case, the reconstructed voice signal samples are not necessarily the same as those of the transmitted signal. A certain degree of distortion is tolerated without compromising the understanding of the message sent. Intuitively, the higher the compression ratio, the greater the loss introduced by the coding process.
Les solutions de compression dites "conjointes" peuvent être classifiées selon trois groupes : Groupe 1 : Codages Source-Canal Conjoints selon lesquels les opérations de codage de source et de canal sont intégrées dans une seule entité.The so-called "joint" compression solutions can be classified into three groups: Group 1: Joint Source-Channel Codings according to which the source and channel coding operations are integrated into a single entity.
Groupe 2 - Codages Concatenes Source-Canal selon lesquels un codeur de source donne est concaténé à un codeur de canal donne et les taux binaires de codage de la source et du canal sont déterminés, de façon à maximiser les performances de 1 ' ensemble . - Groupe 3 - Codages Conjoints avec Contraintes selon lesquels le codeur et/ou le décodeur sont modifies de manière a prendre en compte les erreurs de canal ; par exemple, un codeur de source optimise pour un canal sans bruit est re-optimise afin de tenir compte de la statistique du canal.Group 2 - Concatenated Source-Channel Encodings whereby a given source encoder is concatenated to a given channel encoder and the source and channel encoding bit rates are determined, so as to maximize the performance of the set. - Group 3 - Joint Codings with Constraints according to which the coder and / or the decoder are modified so as to take into account channel errors; for example, a source encoder optimized for a noise-free channel is re-optimized to take into account the statistics of the channel.
L'objectif principal de l'invention est de fournir un codage source-canal conjoint du groupe 1 notamment pour les coefficients en sortie d'un vocodeur, en évitant d'ajouter de la redondance par l'intermédiaire d'un code correcteur d'erreurs, comme dans un codeur du groupe 2, et d'être très sensible a la moindre variation des caractéristiques du canal, comme avec un codeur du groupe 3.The main objective of the invention is to provide a joint source-channel coding of group 1 in particular for the coefficients at the output of a vocoder, by avoiding adding redundancy by means of a code for correcting errors, as in a group 2 coder, and to be very sensitive to the slightest variation in channel characteristics, as with a group 3 coder.
A cette fin, un codage source-canal conjoint comprenant une quantification et une modulation de phase avec une constellation de NC points, est caractérisé en ce que la constellation de la modulation de pnase est une constellation tournée déduite d'une rotation prédéterminée d'une constellation de phase uniforme à NC points d'état afin que les NC points de la constellation soient projetés sur l'un quelconque des axes de la constellation en NC projections toutes différentes, et la quantification présente une loi de quantification non uniforme de NS niveaux qui sont NS projections parmi NC projections différentes de la constellation tournée. Les nombres NC et NS sont des entiers, avec NS < NC .To this end, a joint source-channel coding comprising a quantization and a phase modulation with a constellation of NC points, is characterized in that the constellation of the pnase modulation is a rotated constellation deduced from a predetermined rotation of a uniform phase constellation at NC state points so that the NC points of the constellation are projected onto any one of the axes of the constellation in NC all different projections, and the quantization presents a non-uniform quantization law of NS levels which are NS projections among NC projections different from the constellation rotated. The numbers NC and NS are integers, with NS <NC.
Selon la réalisation préférée, la constellation de NC points est une constellation de modulation de phase, ou une constellation de modulation d'amplitudeAccording to the preferred embodiment, the constellation of NC points is a phase modulation constellation, or an amplitude modulation constellation
2 de deux porteuses en quadrature de type MAQ ou MAQ .2 of two quadrature carriers of the MAQ or MAQ type.
L'invention concerne également un codeur de parole ou d'image dont les moyens de quanti ication, appelés également codeurs, produisant des coefficients et paramètres à multiplexer en trames sont des codeurs source-canal conjoint mettant en oeuvre des codages source-canal conjoint conformes à 1 ' invention .The invention also relates to a speech or image coder whose quantization means, also called coders, producing coefficients and parameters to be multiplexed in frames are joint source-channel coders implementing conforming joint source-channel codings to the invention.
Lorsque le codeur est un vocodeur pour système cellulaire de radiocommunications mobiles, les codeurs source-canal conjoint pour des coefficients de gain de filtre d'analyse à long terme, des paramètres de position de grille d'excitation et des coefficients de filtre prédicteur à court terme de rangs élevés comprennent des modulateurs à constellation de phase tournée de type MAQ, et les codeurs source-canal conjoint pour des coefficients de filtre prédicteur à court terme de rangs faibles, des coefficients de retard de filtre d'analyse à long terme et des paramètres de bloc d'amplitude comprennent des modulateurs à constellation de phaseWhen the coder is a vocoder for the cellular mobile radio system, the joint source-channel coders for long-term analysis filter gain coefficients, excitation grid position parameters, and short predictor filter coefficients high rank terms include QAM-type phase constellation modulators, and joint source-channel coders for short-term short-term predictor filter coefficients, long-term analysis filter delay coefficients, and amplitude block parameters include phase constellation modulators
2 tournée de type MAQ .2 MAQ-type tours.
Selon une variante, les modulateurs à constellation de phase tournée de types MAQ et MAQ sont respectivement précédés par des quantificateurs présentant des lois de quantification non uniformes dont les niveaux sont respectivement des sous- ensembles de projections relatives à des points d'une constellation de phase tournée prédéterminée à au moins quatre dimensions. Par exemple, la constellation de phase tournée prédéterminée est celle issue d'une rotation de la constellation de laAccording to a variant, the modulators with phase constellation of MAQ and MAQ types are respectively preceded by quantifiers having non-uniform quantization laws the levels of which are respectively subsets of projections relating to points of a constellation of phase predetermined turned to at least four dimensions. For example, the predetermined tour phase constellation is that resulting from a rotation of the constellation of the
2 modulation de phase uniforme MAQ -256.2 uniform phase modulation MAQ -256.
En sortie d'un codeur de parole, dit vocodeur, par analyse-synthèse ou par transformée, l'invention quantifie les coefficients en sortie du vocodeur pour les transformer en des échantillons prenant un nombre fini de valeurs possibles. Ces échantillons ont été choisis de façon à prendre en compte les perturbations du canal. De plus, ce codage ne fait pas apparaître de bits sensibles comme c'est le cas des techniques classiques en tandem. Ainsi, la résistance du codage de l'invention au bruit du canal est plus grande.At the output of a speech coder, called a vocoder, by analysis-synthesis or by transform, the invention quantifies the coefficients at the output of the vocoder to transform them into samples taking a finite number of possible values. These samples were chosen to take into account the disturbances of the channel. In addition, this coding does not show sensitive bits as is the case with conventional tandem techniques. Thus, the resistance of the coding of the invention to channel noise is greater.
Le vocodeur de l'invention fournit un signal de parole comprimé résistant aux perturbations du canal de transmission, permettant la réalisation d'émetteurs et de récepteurs à complexité et à coût limité, en particulier en ce qui concerne le codage et le décodage. La largeur de bande du canal de transmission est réduite de plusieurs unités par rapport à la technique antérieure.The vocoder of the invention provides a compressed speech signal resistant to disturbances of the transmission channel, allowing the production of transmitters and receivers with complexity and limited cost, in particular as regards coding and decoding. The bandwidth of the transmission channel is reduced by several units compared to the prior art.
D'autres caractéristiques et avantages de la présente invention apparaîtront plus clairement à la lecture de la description suivante de plusieurs réalisations ; référées de l'invention en référence aux dessins annexés correspondants dans lesquels :Other characteristics and advantages of the present invention will appear more clearly on reading the following description of several embodiments ; references of the invention with reference to the corresponding appended drawings in which:
- les figures 1A et 1B sont des diagrammes de constellations de phase d'une modulation MDP4 non tournée et tournée respectivement ; la figure 2 est un diagramme montrant la transformation entre une constellation αe canal correspondant à la modulation de phase MDP4 tournée et une constellation de source selon l'invention ; - la figure 3 est un bloc-diagramme schématique d'un codeur source-canal conjoint et d'un αecodeur canal-source conjoint selon l'invention relies par un canal de transmission à évanouissements ;- Figures 1A and 1B are diagrams of phase constellations of an MDP4 modulation not turned and turned respectively; FIG. 2 is a diagram showing the transformation between a constellation αe channel corresponding to the phase modulation MDP4 turned and a source constellation according to the invention; - Figure 3 is a schematic block diagram of a joint source-channel encoder and a joint channel-source αecoder according to the invention connected by a faded transmission channel;
- les figures 4A et 4B sont des diagrammes de constellations de phase d'une modulation MAQ-16 non tournée et tournée respectivement ;- Figures 4A and 4B are diagrams of phase constellations of an MAQ-16 modulation not turned and turned respectively;
- la figure 5 est un bloc-diagramme d'un codeur de parole selon la technique antérieure pour système cellulaire de radiocommunications mobiles GSM ; et - la figure 6 est un bloc-diagramme d'un codeur source-canal conjoint en blocs selon l'invention pour système cellulaire de radiocommunications mobiles GSM.- Figure 5 is a block diagram of a speech coder according to the prior art for cellular GSM mobile radio system; and FIG. 6 is a block diagram of a joint source-channel coder in blocks according to the invention for a cellular GSM mobile radio system.
Selon l'invention, des coefficients en sortie d'un codeur, tel que vocodeur, sont quantifies suivant un code conjoint en blocs. Ce scnema de quantification étiquette les coefficients er sortie du vocodeur, appelle par la suite échantillon, non pas par des bits, mais par des symboles qui sont transmis directement sur le canal. Ainsi, pour transmettre dans le canal le signal suivant :According to the invention, coefficients at the output of an encoder, such as vocoder, are quantified according to a joint code in blocks. This quantization scheme labels the coefficients and output of the vocoder, called subsequently sample, not by bits, but by symbols which are transmitted directly on the channel. Thus, to transmit the following signal in the channel:
x(t) = ∑akπ(t-kT)x (t) = ∑a k π (t-kT)
où h (t)' " désigne la réponse mpulsionnelle ~du filtre d'émission et T est l'inverse de la rapidité de modulation exprimée en bauds , les symboles a< sont directement émis sur le canal et représentent tout ou partie d'un échantillon quantifié issu du vocodeur. La relation qu' il y a entre les échantillons quantifiés du vocodeur et les symboles transmis a, identifie le codage en blocs.where h (t) '"denotes the pulse response ~ of the transmission filter and T is the inverse of the modulation speed expressed in bauds, the symbols a < are directly transmitted on the channel and represent all or part of a quantified sample from the vocoder. The relation that there is between the quantized samples of the vocoder and the symbols transmitted has, identifies the coding in blocks.
Le code est l'ensemble de toutes les séquences de symboles ak possibles pouvant être transmises sur le canal.The code is the set of all the possible symbol sequences a k that can be transmitted on the channel.
Le codeur transforme des échantillons de la source non encore quantifiés, ou quantifiés plus finement, en mots du code. Pour cela, le codeur cherche le mot du code le plus proche de la séquence d'échantillons émis par le vocodeur suivant un certain critère de distance prédéterminée. Très souvent, ce critère de distance est la distance euclidienne entre l'échantillon sortant du vocoαeur et le symbole de la source. Ces symboles de source sont ensuite transformes en symboles a du canal.The encoder transforms samples from the source not yet quantified, or more finely quantified, into code words. For this, the coder searches for the code word closest to the sequence of samples sent by the vocoder according to a certain predetermined distance criterion. Very often, this distance criterion is the Euclidean distance between the sample leaving the vocoαeur and the symbol of the source. These source symbols are then transformed into channel symbols a.
Les mots de code qui ne sont qu' une suite de symboles du canal ak, sont ensuite transmis en émettant le signal vu précédemment x(t) = ∑akh(t-kT) k où h(t) est la réponse impulsionnelle au filtre d'émission et T est l'inverse de la rapidité de modulation, c'est-à-dire du débit en symboles par seconde ou en bauds . Pendant la transmission, le canal αistorα ces mots de coαe en rajoutant au bruit, ou en introduisant de l'interférence entre symboles, par exemple .The code words which are only a series of symbols of the channel a k , are then transmitted by emitting the signal seen previously x (t) = ∑a k h (t-kT) k where h (t) is the response impulse to the emission filter and T is the inverse of the modulation speed, that is to say of the bit rate in symbols per second or in bauds. During transmission, the channel αistorα these coαe words by adding to the noise, or by introducing interference between symbols, for example.
Le décodeur reconstruit ensuite les échantillons du vocodeur à partir des observations bruitées. Pour cela, il choisit le mot de code situé -à la plus petite distance suivant un certain critère du mot reçu. Puis, il lui associe l'échantillon quantifié qui lui correspond. Enfin, grâce aux coefficients du vocodeur reconstruit, le décodeur de voix reconstruit la parole émise.The decoder then reconstructs the vocoder samples from the noisy observations. For this, he chooses the code word located - at the shortest distance according to a certain criterion of the word received. Then, it associates the quantified sample which corresponds to it. Finally, thanks to the coefficients of the reconstructed vocoder, the voice decoder reconstructs the spoken word.
L'invention met en oeuvre le codage conjoint source-canal suivant dans le mode particulier des codes conjoints source-canal en blocs.The invention implements the following joint source-channel coding in the particular mode of joint source-channel block codes.
En codage conjoint, lors de la transmission numérique à travers des canaux à évanouissements, tels que canaux radio pour radiotéléphones mobiles, le codage consiste à passer d'un code source, dit dictionnaire source, composé de points d'un espace a DS dimensions à un code canal, dit dictionnaire canal, composé du même nombre de points, ma s dans un espace à DC dimensions. En supposant que DC > DS , les points du dictionnaire source en nombre égal aux points du dictionnaire canal doivent être tous différents, mais dans un espace de plus petite dimension que l'espace canal. Pour la transmission dans un canal à évanouissements, les dimensions où ont lieu les évanouissements sont similaires justement aux dimensions manquantes du dictionnaire source. La résolution du problème de la transmission dans un canal de Rayleigh fait appel à la notion de diversité qui consiste à répartir l'information dans le plus grand nombre possible de composantes du mot de code, de façon à la récupérer dans les composantes qui ne sont pas en évanouissement.In joint coding, during digital transmission through fading channels, such as radio channels for mobile radiotelephones, the coding consists of passing from a source code, called source dictionary, composed of points of a space a DS dimensions to a channel code, called a channel dictionary, composed of the same number of points, ma s in a space with DC dimensions. Assuming that DC> DS, the points of the source dictionary in a number equal to the points of the channel dictionary must all be different, but in a space of smaller dimension than the channel space. For transmission in a fading channel, the dimensions where the fading takes place are precisely similar to the missing dimensions of the source dictionary. The resolution of the problem of transmission in a Rayleigh channel calls upon the notion of diversity which consists in distributing the information in the greatest possible number of components of the code word, so as to recover it in the components which are not not fainting.
En ce qui concerne le lien entre la diversité et le codage conjoint, l'approche suivie par l'invention est l'inverse de l'approche classique du conjoint. La diversité est utilisée en remontant du canal vers la source. Pour cela, l'invention introduit la notion de constellations tournées. Les techniques de modulation utilisant des réseaux de points sont devenues des outils très puissants pour la conception de systèmes de transmission numérique avec une haute efficacité spectrale tant pour un canal gaussien, que pour un canal à évanouissements, et notamment pour un canal de Rayleigh.With regard to the link between diversity and joint coding, the approach followed by the invention is the reverse of the classic spouse approach. Diversity is used when going back from the channel to the source. For this, the invention introduces the notion of turned constellations. Modulation techniques using point arrays have become very powerful tools for the design of digital transmission systems with high spectral efficiency both for a Gaussian channel and for a fading channel, and in particular for a Rayleigh channel.
La notion d'ordre de diversité est introduite pour atteindre des gains très importants dans le cas d'un canal à évanouissements, et donc pour le cas de l'optimisation conjointe. Un ordre de diversité L d'une constellation multidimensionnelle est le nombre minimal de composantes différentes entre deux points quelconques de la constellation. Dans ce qui suit, la diversité est désignée par L. Pour introduire de la diversité dans une constellation multidimensionelle, celle-ci est tournée de façon à ce que tous les couples de points aient un nombre maximal de composantes différentes. Par exemple pour une constellation à deux dimensions, le passage du diagramme de phase pour une Modulation par Déplacement de Phase à 4 états MDP4 dans la figure 1A au diagramme de phase dans la figure 1B montre que, grâce à une rotation de la constellation MDP4 d'un angle α, un ordre de diversité L=2 est obtenu. Dans la figure 1A, chaque point de la constellation de phase uniforme a l'une de ses cordonnées égale à un autre point de la constellation, et dans la figure 1B, chaque point de la constellation tournée a chacune de ses deux coordonnées , XI, X2 différente des coordonnées correspondantes des autres points de la constellation tournée .The notion of diversity order is introduced to achieve very significant gains in the case of a fading channel, and therefore for the case of joint optimization. A diversity order L of a multidimensional constellation is the minimum number of different components between any two points of the constellation. In what follows, diversity is designated by L. To introduce diversity into a multidimensional constellation, it is rotated so that all pairs of points have a maximum number of different components. For example for a two-dimensional constellation, the passage from the phase diagram for a Modulation by Phase Displacement with 4 states MDP4 in FIG. 1A to the phase diagram in FIG. 1B shows that, thanks to a rotation of the constellation MDP4 d 'an angle α, an order of diversity L = 2 is obtained. In FIG. 1A, each point of the uniform phase constellation has one of its coordinates equal to another point of the constellation, and in FIG. 1B, each point of the constellation turned to each of its two coordinates , XI, X2 different from the corresponding coordinates of the other points of the constellation being rotated.
Si l'amplitude p(t) du signal modulé est soumise suivant une direction quelconque φ(t) à une réduction drastique due aux effets d'un évanouissement, la distance entre deux points de la constellation tournée montrée à la figure 1B n'est pas réduite de la même façon que la constellation uniforme non tournée montrée à la figure 1A. Par conséquent la constellation tournée offre un degré de protection supplémentaire. Pour atteindre ce degré d'indépendance dans les évanouissements des composantes en phase et en quadrature XI (t) = [p(t) .cosφ(t) ] et X2(t) = [p (t ) . sinφ ( t ) ] , il est connu d'entrelacer le signal modulé en sortie du modulateur inclus dans le codeur de l'émetteur. La constellation tournée confère un gain de 8 dB par rapport à la constellation MDP4 non tournée. Dans le cas d'une constellation de phase à DC dimensions, où DC est un entier tel que DC > 2, le principe est le même. Une modulation codée en blocs vue comme un sous-ensemble fini d'un réseau de points est "tournée" par une rotation à DC dimensions de façon à avoir une diversité la plus grande possible jusqu'à DC . Dans ce cas, les gains attendus sont encore plus importants que ceux pour une constellation à 2 dimensions.If the amplitude p (t) of the modulated signal is subjected in any direction φ (t) to a reduction drastic due to the effects of a fainting, the distance between two points of the turned constellation shown in Figure 1B is not reduced in the same way as the uniform non-turned constellation shown in Figure 1A. Therefore the turned constellation offers an additional degree of protection. To achieve this degree of independence in the fading of phase and quadrature components XI (t) = [p (t) .cosφ (t)] and X2 (t) = [p (t). sinφ (t)], it is known to interleave the modulated signal at the output of the modulator included in the encoder of the transmitter. The rotated constellation gives an gain of 8 dB compared to the unpurned MDP4 constellation. In the case of a phase constellation with DC dimensions, where DC is an integer such that DC> 2, the principle is the same. A block coded modulation seen as a finite subset of a network of points is "rotated" by a rotation at dimensional DC so as to have the greatest possible diversity up to DC. In this case, the expected gains are even greater than those for a 2-dimensional constellation.
L'invention établit une application entre les niveaux issus d'un codage de source et les points de la constellation canal pour obtenir la robustesse souhaitée en terme de distorsion de la source reconstruite à la réception. Le codage conjoint source-canal en blocs de l'invention exploite la diversité des, constellations tournées pour trouver cette application.The invention establishes an application between the levels resulting from a source coding and the points of the channel constellation in order to obtain the desired robustness in terms of distortion of the source reconstructed on reception. The joint source-channel block coding of the invention exploits the diversity of the constellations turned to find this application.
A cette fin, une démarche inverse à la démarche classique est suivie. Le codage de source est déduit du codage de canal. Pour établir le lien entre codage conjoint et diversité, l'invention applique la propriété que, si un code est issu d'une constellation tournée a diversité maximale, laquelle est égale a la dimension de la constellation, alors les projections sur l'un quelconque des axes de la constellation constituent un ensemble présentant un cardinal égal à celui du code.To this end, an approach opposite to the conventional approach is followed. Source coding is derived from channel coding. To establish the link between coding conjunct and diversity, the invention applies the property that, if a code comes from a constellation turned to maximum diversity, which is equal to the dimension of the constellation, then the projections on any one of the axes of the constellation constitute a set with a cardinal equal to that of the code.
Cette propriété est à la base du codage conjoint en blocs selon l'invention, et est déduite de la définition même de la diversité d'une constellation.This property is the basis of joint block coding according to the invention, and is deduced from the very definition of the diversity of a constellation.
Toujours selon l'exemple de la constellation tournée a 2 dimensions pour modulation MDP4, la figure 2 montre un ensemble de NS = 4 points obtenus par projection des NC = 4 points [01, 00, 11, 10] de la constellation canal CC dans la figure IB sur l'unStill according to the example of the constellation turned to 2 dimensions for MDP4 modulation, FIG. 2 shows a set of NS = 4 points obtained by projection of the NC = 4 points [01, 00, 11, 10] of the constellation CC channel in figure IB on one
0-X1 des deux axes de la constellation MDP4 tournée, afin d'obtenir la constellation source CS .0-X1 of the two axes of the MDP4 constellation rotated, in order to obtain the source constellation CS.
A partir de l'ensemble CS à NC = 4 points obtenus par projection est construit un quantificateur scalaire à un facteur d'échelle près.From the set CS at NC = 4 points obtained by projection is constructed a scalar quantifier to within a scale factor.
Dans l'exemple illustré à la figure 2, le dictionnaire source est constitue par les points ou niveaux indiques par une croix sur l'axe 0-X1, alors que le dictionnaire canal est constitué des points [XI, X2 ] de la constellation MDP4 tournée. Le procède de l'invention est avantageux en ce qu'il adapte le quantificateur scalaire à la distribution de la source par simple modification αe l'angle de rotation. Dans la figure 2, un exemple d'échelle de quantification pour échantillon de source est montre en traits épais. _ _In the example illustrated in FIG. 2, the source dictionary is constituted by the points or levels indicated by a cross on the axis 0-X1, while the channel dictionary consists of the points [XI, X2] of the constellation MDP4 tour. The method of the invention is advantageous in that it adapts the scalar quantizer to the distribution of the source by simple modification to the angle of rotation. In Figure 2, an example of a quantification scale for a source sample is shown in thick lines. _ _
La Figure 3 montre un codeur 1 dans un dispositif d'émission et un décodeur 2 dans un dispositif de réception selon l'invention, reliés a travers un canal à évanouissements 3.Figure 3 shows an encoder 1 in a transmitting device and a decoder 2 in a receiving device according to the invention, connected through a fading channel 3.
Le codeur 1 comporte une source 11, un circuit de codage source-canal conjoint en blocs 12 et un entrelaceur 13. La source 11 est par exemple un vocodeur .The coder 1 comprises a source 11, a joint source-channel coding circuit in blocks 12 and an interleaver 13. The source 11 is for example a vocoder.
Le circuit de codage 12 comprend un quantificateur scalaire 121 suivi d'un modulateurThe coding circuit 12 includes a scalar quantizer 121 followed by a modulator
122. Le quantificateur scalaire est associé à un circuit programmable 123 qui établit une lo de quantification non uniforme selon le dictionnaire source comme montré en bas de la figure 2. La lo de quantification est choisie en fonction αe la constellation initiale (figure 1A) du modulateur 122 après une rotation de celle-ci avec un angle prédéterminé α afin de maximiser le cardinal des points projetés sur un axe prédéterminé O-Xl de la constellation canal CC . Chaque échantillon fourni par le vocodeur 11 et représentant par exemple un paramètre du modèle de vocodage mis en oeuvre dans le vocodeur 11 est quantifié dans le quantificateur 121 selon la loi de quantification dite par arrondi, en correspondant au niveau ou point de la constellation source CS, le plus proche de l'échantillon. Puis le modulateur 122 convertit par projection "inverse" chaque point de la constellation CS transmis par le quantificateur en le symbole correspondant de la constellation canal CC qui, selon la figure IB ou 2, est défini par des coordonnées cartésiennes XI (t) et X2(t), ou polaires, amplitude p(t) et phase φ(t) qu sont différentes des coordonnées de tous les autres symboles de la constellation CC .122. The scalar quantizer is associated with a programmable circuit 123 which establishes a non-uniform quantization lo according to the source dictionary as shown at the bottom of FIG. 2. The quantization lo is chosen as a function of the initial constellation (FIG. 1A) of the modulator 122 after a rotation thereof with a predetermined angle α in order to maximize the cardinality of the points projected on a predetermined axis O-Xl of the constellation CC channel. Each sample provided by the vocoder 11 and representing for example a parameter of the vocoding model implemented in the vocoder 11 is quantified in the quantizer 121 according to the quantization law called by rounding, by corresponding to the level or point of the source constellation CS , closest to the sample. Then the modulator 122 converts by "reverse" projection each point of the constellation CS transmitted by the quantifier into the corresponding symbol of the channel constellation CC which, according to FIG. IB or 2, is defined by Cartesian coordinates XI (t) and X2 (t), or polar, amplitude p (t) and phase φ (t) that are different from the coordinates of all the other symbols of the constellation CC.
L' entrelaceur 13 entrelace les symboles produits par le modulateur 122 pour y introduire de la diversité temporelle d'une manière connue. Le canal de transmission 3 est schématise dans la figure 3 par un générateur 31 générant du bruit multiplicatif dû aux évanouissements et un générateurThe interleaver 13 interleaves the symbols produced by the modulator 122 to introduce time diversity therein in a known manner. The transmission channel 3 is shown diagrammatically in FIG. 3 by a generator 31 generating multiplicative noise due to fading and a generator
32 générant un bruit additif dû à des signaux parasites notamment.32 generating additive noise due in particular to parasitic signals.
A la réception dans le décodeur 2, un désentrelaceur 21 désentrelace les symboles entrelacés selon l'ordre initial, et un circuit d'estimation 22 estime l'amplitude et la phase instantanée du canal. Un démodulateur 23 démodule chaque symbole désentrelace en points de la constellation source CS, qui avait été émis, en effectuant la transformation des constellations CC vers CS de la figure 2 par projection sur l'axe prédéterminé O-Xl.On reception in the decoder 2, a deinterleaver 21 deinterleaves the interleaved symbols according to the initial order, and an estimation circuit 22 estimates the amplitude and the instantaneous phase of the channel. A demodulator 23 demodulates each deinterleaving symbol in points of the source constellation CS, which had been emitted, by carrying out the transformation of the constellations CC to CS of FIG. 2 by projection on the predetermined axis O-Xl.
En partant d'une constellation de canal du type à modulation d'amplitude de deux porteuses en quadrature MAQ à NC = 2 états avec k un entier positif, ou une version a plus grande dimension de ce type de modulation, l'application d'une matrice de rotation à la constellation de canal produit une constellation tournée.Starting from a channel constellation of the amplitude modulation type of two carriers in quadrature MAQ at NC = 2 states with k a positive integer, or a larger version of this type of modulation, the application of a rotation matrix at the channel constellation produces a rotated constellation.
Par exemple pour une constellation uniforme du type MAQ-lβ montrée à la figure 4A, deux projections sont possibles : l'une sur l'axe des composantes en phase I (t) et l'autre sur l'axe des composantes en quadrature Q(t). Chaque ensemble de projections résultant de la projection des points αe la constellation MAQ-16 sur l'un des axes 01, OQ avant application de la matrice de rotation, a un cardinal de 4, c'est-à-dire il n'y a que 4 valeurs _de composante différentes relativement à chaque axe 01, OQ pour les 16 points.For example, for a uniform constellation of the MAQ-lβ type shown in FIG. 4A, two projections are possible: one on the axis of the components in phase I (t) and the other on the axis of the components in quadrature Q (t). Each set of projections resulting from the projection of the points αe the constellation MAQ-16 on one of the axes 01, OQ before application of the rotation matrix, has a cardinal of 4, i.e. there is no has only 4 different component values relative to each axis 01, OQ for the 16 points.
Une matrice de rotation de diversité maximale appliquée à la constellation MAQ-lβ convertit celle- ci en une constellation tournée présentant des ensembles de projections sur les axes 01 et OQ, avec un cardinal de NS = 16, comme montré à la figure 4B, c'est-à-dire le nombre NS d'éléments de l'ensemble des NC projections sur l'axe 01 ou OQ est égal a celui de l'ensemble des mots de code du canal. Les opérations de rotation et projection pour concevoir un quantificateur scalaire avec une bijection entre les points de la constellation de canal et les points ou niveaux de la constellation linéaire de source est une caractéristique de l'invention.A rotation matrix of maximum diversity applied to the constellation MAQ-lβ converts it ci in a turned constellation presenting sets of projections on axes 01 and OQ, with a cardinality of NS = 16, as shown in FIG. 4B, i.e. the number NS of elements of the set of NC projections on axis 01 or OQ is equal to that of the set of code words for the channel. The rotation and projection operations for designing a scalar quantifier with a bijection between the points of the channel constellation and the points or levels of the linear source constellation is a characteristic of the invention.
A titre d'exemple, l'invention est appliquée au codeur αe parole plein débit inclus dans une station de base ou dans une station mobile du reseau cellulaire de radiocommunications mobiles selon la norme GSM. Bien que la station de base soit reliée au réseau téléphonique commuté par des canaux RNIS à 64 kbit/s, le signal de parole échantillonne initialement à 8 kHz sur 8 bits est compressé en un signal de parole à 13 kbit/s pour être transmis dans un canal de radiocommunication. Un codeur de parole GSM COP montre schématiquement a la figure 5 transforme des trames de parole, dites également segments ou blocs, de 20 ms de 160 échantillons de 8 bits, en des blocs de 260 bits, soit un déoit de 260/20 = 13 kbit/s. Le codeur COP comprend essentiellement un circuit de segmentation SEG pour segmenter un signal de parole SP à 64 kbit/s, un filtre d'analyse LPC (Lmear Prédictive Coding) avec prédicteur linéaire PL, un filtre d'analyse_ à prédiction à long terme LTP (Long Term prédiction) et un circuit de calcul de signal d'excitation RPEBy way of example, the invention is applied to the full-rate speech coder included in a base station or in a mobile station of the cellular mobile radio network according to the GSM standard. Although the base station is connected to the switched telephone network via ISDN 64 kbit / s channels, the speech signal initially sampled at 8 kHz on 8 bits is compressed into a speech signal at 13 kbit / s to be transmitted in a radio channel. A GSM COP speech coder shown diagrammatically in FIG. 5 transforms speech frames, also called segments or blocks, of 20 ms of 160 samples of 8 bits, into blocks of 260 bits, that is to say a failure of 260/20 = 13 kbit / s. The coder COP essentially comprises a segmentation circuit SEG for segmenting a speech signal SP at 64 kbit / s, an analysis filter LPC (Lmear Predictive Coding) with linear predictor PL, a long-term prediction analysis_ filter LTP (Long Term prediction) and an RPE excitation signal calculation circuit
(Regular Puise Excitation) . Le filtre prédicteur à court terme LPC modelise les variations du conduit vocal a court terme. Le prédicteur LPC génère huit coefficients LAR (Log Area(Regular Puise Excitation). The LPC short-term predictor filter models variations in the short-term vocal tract. The LPC predictor generates eight LAR (Log Area) coefficients
Ratio) avec un débit de 1,8 kbit/s qui sont quantifies de la façon suivante :Ratio) with a speed of 1.8 kbit / s which are quantified as follows:
LAR(O) et LAR(l) : quantification sur β bits, LAR (O) and LAR (l): quantification on β bits,
LAR(2) et LAR(3) : quantification sur 5 bits, LAR(4) et LAR(5) : quantification sur 4 bits, LAR (2) and LAR (3): 5-bit quantization, LAR (4) and LAR (5): 4-bit quantization,
LAR(6) et LAR(7) : quantification sur 3 Dits, soit 6 bits.LAR (6) and LAR (7): quantization on 3 Dits, i.e. 6 bits.
Le filtre d'analyse a long terme LTP modelise les variations rapides du conduit vocal. Il fournit un coefficient de retard LTP-Lag et un coefficient de gain LTP-Gam, par sous-trame de 5 millisecondes, soit un débit de 1,8 kbit/s ; les coefficients a long terme sur une trame de 20 ms sont: LTP-Lag : quantification sur 7 bits, soit 4 x 7 The LTP long-term analysis filter models rapid variations in the vocal tract. It provides an LTP-Lag delay coefficient and an LTP-Gam gain coefficient, per sub-frame of 5 milliseconds, ie a speed of 1.8 kbit / s; the long-term coefficients on a 20 ms frame are: LTP-Lag: quantization on 7 bits, ie 4 x 7
LTP-Gam : quantification sur 2 bits, soit 4 x 2 = 8 bits.LTP-Gam: 2-bit quantization, i.e. 4 x 2 = 8 bits.
Le circuit de calcul RPE produit un signal d'excitation constitué par un bloc d'impulsions qui sont régulièrement reparties dans le temps et codées a 9,4 kbit/s en position et en amplitude au moyen d'une grille. Le signal d'excitation est compose de paramètres de position de grille Mi en sortie d'un filtre passe-bas FBP, αe paramètres αe bloc d'amplitude X axi, et de valeurs relatives a _13 impulsions xι(0) à xι(12) par sous-trame î de 5 ms . Les coefficients d'excitation sur une trame de 20 ms sont : Mi : quantification sur 2 bits, soit 4 x 2 = 8The RPE calculation circuit produces an excitation signal consisting of a block of pulses which are regularly distributed over time and coded at 9.4 kbit / s in position and amplitude by means of a grid. The excitation signal is composed of gate position parameters Mi at the output of a low-pass filter FBP, αe parameters αe amplitude block X axi, and values relating to _13 pulses xι (0) to xι (12 ) by subframe î of 5 ms. The excitation coefficients on a 20 ms frame are: Mi: 2-bit quantization, i.e. 4 x 2 = 8
XMaxi : quantification sur 6 bits, soit 4 x 6 = 24 bits. Impulsions d'excitation xι(0) à xι(12) : quantification sur 3 bits des 13 impulsions, soit 4 x 39 = 156 bits.XMaxi: quantization on 6 bits, i.e. 4 x 6 = 24 bits. Excitation pulses xι (0) to xι (12): 3-bit quantization of the 13 pulses, i.e. 4 x 39 = 156 bits.
Les coefficients et paramètres précités sont multiplexes dans un multiplexeur MX en sortie du codeur pour produire des trames à 260 bits toutes les 20 ms, soit au débit de codage de 13 kbit/s. Puis dans un codeur de canal COC suivi d'un entrelaceur, ces bits sont séparés en trois classes d' importance relative différentes. Les bits des deux premières classes sont protégés par des codes correcteurs d'erreurs différents, et les bits de la dernière classe, la moins importante, sont concatenes sans protection. Le codeur de canal délivre une unité de données de 456 bits codes toutes les 20 ms, soit un débit de 22,8 kbit/s. Cette protection sélective par classe est connue comme une technique de Protection Inégale contre les erreurs.The above coefficients and parameters are multiplexed in an MX multiplexer at the output of the encoder to produce frames at 260 bits every 20 ms, ie at the coding rate of 13 kbit / s. Then in a COC channel coder followed by an interleaver, these bits are separated into three different relative importance classes. The bits of the first two classes are protected by different error correcting codes, and the bits of the last class, the least important, are concatenated without protection. The channel encoder delivers a 456-bit code unit of data every 20 ms, a rate of 22.8 kbit / s. This selective protection by class is known as an Unequal Protection technique against errors.
En se référant maintenant a la figure 6, un codeur de parole source-canal conjoint en blocs COPa conforme à l' invention pour reseau GSM comprend les circuits fonctionnels SEG, PLa, LPCa, LTPa, FPB et RPEa du codeur COP, mais avec les modifications suivantes. Comparativement à la figure 5, le codeur de canal COC est supprimé, et les ensembles de quantificateurs, dits également codeurs, QUI, QU2_et QU3 en sortie des circuits PL, LTP et RPE sont remplacés par des circuits de codage source-canal conjoint en blocs du type de celui 12 décrit en référence à la figure 3 dans les circuits PLa, LTPa et RPEa, à l'exception de la sortie d'impulsion du circuit de calcul RPEa. De même, les décodeurs (non représentés) dans les circuits LPC, LTP et RPE effectuant respectivement les fonctions inverses des codeurs (quantificateurs) dans les circuits PL, LTP et RPE sont remplacés par des circuits de décodage canal-source conjoint du type de celui 23-24 montré à la figure 3.Referring now to FIG. 6, a joint source-channel speech coder in blocks COPa in accordance with the invention for GSM network comprises the functional circuits SEG, PLa, LPCa, LTPa, FPB and RPEa of the coder COP, but with the following changes. Compared to FIG. 5, the channel coder COC is deleted, and the sets of quantizers, also called coders, QUI, QU2_ and QU3 at the output of the PL, LTP and RPE circuits are replaced by joint source-channel coding circuits in blocks. of the type of that 12 described with reference to FIG. 3 in the circuits PLa, LTPa and RPEa, with the exception of the pulse output of the RPEa calculation circuit. Similarly, the decoders (not shown) in the LPC, LTP and RPE circuits respectively performing the inverse functions of the coders (quantifiers) in the PL, LTP and RPE circuits are replaced by joint channel-source decoding circuits of the type of that 23-24 shown in Figure 3.
Un modulateur d'amplitude MDA sélectionne pour chacune des 4 x 13 = 52 impulsions d'excitation xi une amplitude parmi 8 niveaux au moyen d'une modulation à déplacement d'amplitude d'une porteuse.An amplitude modulator MDA selects for each of the 4 x 13 = 52 excitation pulses xi an amplitude from 8 levels by means of amplitude displacement modulation of a carrier.
Selon d'autres variantes, la modulation est effectuée sur 16 ou 32 niveaux par exemple. Pour les 4 gains du prédicteur à long terme LTP-According to other variants, the modulation is carried out on 16 or 32 levels for example. For the 4 gains of the long-term predictor LTP-
Gain, les 4 paramètres de position de grille d'excitation Mi et les 4 coefficients de filtre prédicteur à court terme LAR ( 4 ) à LAR (7), sont prévus des circuits de codage source-canal conjoint en blocs SCI, SC2, SC3 et SC4 comportant chacun un quantificateur de taux correspondant dont la loi de quantification résulte d'une projection axiale d'une constellation MAQ à DC = 2 dimensions ayant subi une rotation de π/5 : MAQ-128 pour chacun des 4 coefficients de filtre d'analyse à long terme LTP-Gain, the 4 excitation grid position parameters Mi and the 4 short-term predictive filter coefficients LAR (4) to LAR (7), are provided for source-channel coding circuits combined in blocks SCI, SC2, SC3 and SC4 each comprising a corresponding rate quantifier whose quantization law results from an axial projection of a constellation MAQ at DC = 2 dimensions having undergone a rotation of π / 5: MAQ-128 for each of the 4 filter coefficients d long-term analysis LTP-
Gain dans le circuit SCI, MAQ-4 pour chacun des 4 paramètres de position de grille Mi dans le circuitGain in the SCI circuit, MAQ-4 for each of the 4 grid position parameters Mi in the circuit
SC2, MAQ-16 pour chacun des coefficients de rangs élevés LAR(4) et LAR(5) dans le circuit SC3 et finalement MAQ-8 pour chacun des coefficients de rangs élevés LAR(β) et LAR(7) dans le circuit SC4._SC2, MAQ-16 for each of the high rank coefficients LAR (4) and LAR (5) in the circuit SC3 and finally MAQ-8 for each of the high rank coefficients LAR (β) and LAR (7) in the circuit SC4 ._
Les autres coefficients des circuits de codage source-canal conjoint en blocs SC5, SCβ et SC7 comportent chacun un quantificateur de taux correspondant dont la loi de αuantification résulte d'une projection axiale d'une constellation MAQxMAQ =The other coefficients of the joint source-channel coding circuits in blocks SC5, SCβ and SC7 each include a corresponding rate quantifier from which the law of αuantification results an axial projection of a constellation MAQxMAQ =
22
MAQ à DC = 4 dimensions ayant subi une rotation engendrée par la matrice carrée de rotation MR suivante :MAQ at DC = 4 dimensions having undergone a rotation generated by the following square matrix of rotation MR:
a b c da b c d
-b -c -d a-b -c -d a
-c -d a b-c -d a b
-a a b c-a a b c
avec : a = 0, 6935209 b = 0,5879371 c = -0,1379495 d = 0,3928468with: a = 0.6935209 b = 0.5879371 c = -0.1379495 d = 0.3928468
Une constellation MAQ -64 est prévue pour chacun des coefficients de rangs faibles LAR(0) et LAR(l)An MAQ -64 constellation is provided for each of the low rank coefficients LAR (0) and LAR (l)
2 dans le circuit SC5, une MAQ -32 pour chacun des coefficients de rangs faibles LAR(2) et LAR(3) dans2 in circuit SC5, an MAQ -32 for each of the low rank coefficients LAR (2) and LAR (3) in
2 le circuit SC6, une MAQ -128 pour chacun des 4 coefficients de retard de filtre d'analyse à long2 the SC6 circuit, an MAQ -128 for each of the 4 long analysis filter delay coefficients
2 terme LTP-Lag dans le circuit SC7 et une MAQ -64 pour chacun des 4 paramètres de bloc d'amplitude Xmaxi dans le circuit SC8.2 LTP-Lag term in the SC7 circuit and an MAQ -64 for each of the 4 amplitude block parameters Xmaxi in the SC8 circuit.
En variante, chaque constellation linéaire de source à 32, 64, 128 niveaux est l'un des sous- ensembles à NS = 32, 64, 128 points résultant des NC = 256 projections sur n'importe quel des DC = 4 axes de coordonnées des points d'une constellation tournée qui est déduite de la rotation de la constellationAs a variant, each linear source constellation at 32, 64, 128 levels is one of the subsets at NS = 32, 64, 128 points resulting from the NC = 256 projections on any of the DC = 4 coordinate axes points of a rotated constellation which is deduced from the rotation of the constellation
MAQ--256 à NC = 256 points suivant la matrice —deMAQ - 256 to NC = 256 points according to the matrix - of
2 rotation MR. La constellation MAQ -256 est le produit cartésien de deux constellations MAQ-lβ. Ainsi, les coefficients et paramètres codés à partir d'une constellation à DC = 4 dimensions sont estimés plus sensibles aux erreurs que ceux codes a partir d'une constellation a DC = 2 dimensions, et par conséquent sont transmis avec une qualité supérieure .2 rotation MR. The MAQ -256 constellation is the Cartesian product of two MAQ-lβ constellations. Thus, the coefficients and parameters coded from a constellation with DC = 4 dimensions are estimated more sensitive to errors than those coded from a constellation a DC = 2 dimensions, and therefore are transmitted with higher quality.
La largeur de bande de fréquence utilisée dans l'ensemble codeur de source COP-codeur de canal COC selon la technique de Protection Inégale de données connue par la norme du système GSM, et la largeur de bande de fréquence utilisée dans le codeur source- canal conjoint COPa de l'invention, exprimées en nombre de symboles complexes par trame de 20 ms , sont respectivement : 456/2 = 228 sympoles complexes, une paire de points représentant un point du plan complexe, et 52/2 + 12 + (12 x 2) = 62 symboles complexes. A performances audio identiques, le codeur source-canal conjoint COPa de l' invention offre un rapport de réduction de largeur de bande d'un facteur de 268/62 = 3, 67.The frequency bandwidth used in the source encoder COP-COC channel encoder assembly according to the Unequal Data Protection technique known by the GSM system standard, and the frequency bandwidth used in the source-channel encoder joint COPa of the invention, expressed in number of complex symbols per 20 ms frame, are respectively: 456/2 = 228 complex sympoles, a pair of points representing a point of the complex plane, and 52/2 + 12 + (12 x 2) = 62 complex symbols. For identical audio performances, the joint source-channel coder COPa of the invention offers a bandwidth reduction ratio by a factor of 268/62 = 3.67.
Le procède de l'invention peut être généralise. Au lieu de considérer une projection d'une constellation canal CC de dimension 2 ou 4 vers une constellation source CS de dimension 1, on peut considérer de la même façon, la projection d'une constellation canal de dimension DC vers une constellation source de dimension DS, avec DODS.The process of the invention can be generalized. Instead of considering a projection of a CC channel constellation of dimension 2 or 4 towards a source constellation CS of dimension 1, we can consider in the same way, the projection of a channel constellation of dimension DC towards a source constellation of dimension DS, with DODS.
Lorsque les dimensions des espaces source et canal sont les mêmes, le dictionnaire source et le dictionnaire canal sont choisis identiques.When the dimensions of the source and channel spaces are the same, the source dictionary and the channel dictionary are chosen to be identical.
Dans la majorité des systèmes de codage de la voix connus, les effets du bruit sont très souvent négligés. Le code de canal a pour tâche de fournir la protection nécessaire pour éviter la perte dans la qualité, lors de la transmission. Dans les codeurs de source connus, les bits qui résultent de la quantification des paramètres sont codes classiquement par conversion analogique-numérique en des mots de code binaire puis codés dans le codeur de canal pour être transmis dans le canal. La séparation est devenue quasiment universelle entre codage de source et codage de canal.In the majority of known voice coding systems, the effects of noise are very often overlooked. The task of the channel code is to provide the necessary protection to prevent loss in the quality, during transmission. In known source coders, the bits which result from the quantization of the parameters are conventionally coded by analog-digital conversion into binary code words and then coded in the channel coder to be transmitted in the channel. The separation has become almost universal between source coding and channel coding.
Le procédé de l'invention effectue l'ensemble du codage de source et du codage de canal différemment.The method of the invention performs all of the source coding and the channel coding differently.
Les paramètres à transmettre sont quantifies dans un quantificateur scalaire non uniforme qui resuite de la projection d'une constellation canal tournée. Dans ce cas, la dimension de l'espace source est DS=1. Pour l'espace canal, la dimension DC dépend du degré d'importance des coefficients a transmettre. Plus un coefficient est important, plus DC est grand. Ainsi, les coefficients de moindre importance correspondent à des constellations canal scalaires, alors que les autres correspondront à des constellations canal à 2, 4 voire 8 dimensions pour le coefficient le plus important.The parameters to be transmitted are quantified in a non-uniform scalar quantizer which results from the projection of a rotated channel constellation. In this case, the dimension of the source space is DS = 1. For the channel space, the DC dimension depends on the degree of importance of the coefficients to be transmitted. The higher the coefficient, the greater the DC. Thus, the coefficients of less importance correspond to scalar channel constellations, while the others will correspond to channel constellations with 2, 4 or even 8 dimensions for the most important coefficient.
Dans la réalisation présentée ci-dessus dans le cadre du système GSM, la performance du codage de l'invention est illustrée en terme de réduction de largeur de bande pour une qualité auoio identique a celle existant actuellement.In the embodiment presented above within the framework of the GSM system, the performance of the coding of the invention is illustrated in terms of reduction in bandwidth for an auoio quality identical to that existing at present.
Une autre approche consiste à présenter le codage source-canal conjoint en blocs comme un moyen de construire .un système de communications .numériques qui, pour une largeur de bande fixée, est capable d'émettre avec une puissance moyenne plus faible. Le codage de l'invention est alors avantageux par exemple dans les systèmes de communications par satellites où le moindre décibel gagné est très conséquent sur le coût de la charge utile. Another approach is to present joint source-channel block coding as a means of building a communications system . which, for a fixed bandwidth, is capable of transmitting with a lower average power. The coding of the invention is then advantageous, for example in communication systems by satellites where the smallest decibel gained is very significant on the cost of the payload.

Claims

REVENDICATIONS
1 - Codage source-canal conjoint comprenant une quantification et une modulation de phase avec une constellation de NC points, caractérise en ce que la constellation de la modulation de phase (122) est une constellation tournée (CC) déduite d'une rotation prédéterminée (α) d'une constellation de phase uniforme à NC points d'état afin que les NC points de la constellation soient projetés sur l'un quelconque des axes (O-I, O-Q) de la constellation en NC projections toutes différentes, et la quantification1 - Joint source-channel coding comprising a quantization and a phase modulation with a constellation of NC points, characterized in that the constellation of the phase modulation (122) is a rotated constellation (CC) deduced from a predetermined rotation ( α) of a uniform phase constellation at NC state points so that the NC points of the constellation are projected onto any one of the axes (OI, OQ) of the constellation in NC all different projections, and the quantification
(121) présente une loi de quantification non uniforme(121) presents a non-uniform quantification law
(CS) de NS niveaux qui sont NS projections parmi NC projections différentes de la constellation tournée(CS) of NS levels which are NS projections among NC different projections of the constellation toured
(CC) .(CC) .
2 - Codage conforme à la revendication 1, dans lequel la constellation de NC points est une constellation de modulation de phase, ou une constellation de modulation d'amplitude de deux2 - Coding according to claim 1, in which the constellation of NC points is a phase modulation constellation, or an amplitude modulation constellation of two
2 porteuses en quadrature de type MAQ ou MAQ .2 MAQ or MAQ type quadrature carriers.
3 - Codeur de parole (COPa) ou d'image comprenant des moyens de quantification pour produire des coefficients et paramètres à multiplexer en trames, caractérisé en ce que les moyens de quantification sont des codeurs source-canal conjoint3 - Speech or image coder (COPa) comprising quantization means for producing coefficients and parameters to be multiplexed in frames, characterized in that the quantization means are joint source-channel coders
(SCI à SC8) mettant en oeuvre des codages source- canal conjoint conformes à la revendication 1 ou 2.(SCI to SC8) implementing joint source-channel codings according to claim 1 or 2.
4 - Codeur conforme à la revendication 3, dans lequel les codeurs source-canal conjoint (SCI, SC2, SC3, SC4) pour des coefficients de gain de filtre d'analyse à long terme (LTP-Gam), des paramètres de position de grille d'excitation (Mi) et des coefficients de filtre prédicteur à court terme de rangs élevés (LAR (4) à LAR (7)) comprennent des modulateurs à constellation de phase tournée de type MAQ, et les codeurs source-canal conjoint (SCS, SC6, SC7, SC8) pour des coefficients de filtre prédicteur à court terme de rangs faibles (LAR(O) à LAR(3)), des coefficients de retard de filtre d'analyse à long terme (LTP-Lag) et des paramètres de bloc d'amplitude (Xmaxi) comprennent des modulateurs à constellation4 - Encoder according to claim 3, in which the joint source-channel encoders (SCI, SC2, SC3, SC4) for long-term analysis filter gain coefficients (LTP-Gam), parameters of excitation gate position (Mi) and short-term predictor filter coefficients of high ranks (LAR (4) to LAR (7)) include modulators with phase constellation of MAQ type, and source-channel coders joint (SCS, SC6, SC7, SC8) for short-term short-term predictor filter coefficients (LAR (O) to LAR (3)), long-term analysis filter delay coefficients (LTP-Lag ) and amplitude block parameters (Xmaxi) include constellation modulators
2 de phase tournée de type MAQ .2 of phase MAQ type tour.
5 - Codeur conforme à la revendication 4, dans lequel les modulateurs à constellation de phase5 - Encoder according to claim 4, in which the phase constellation modulators
2 tournée de types MAQ et MAQ sont respectivement précédés par des quantificateurs présentant des lois de quantification non uniformes dont les niveaux sont respectivement des sous-ensembles de projections relatives à des points d'une constellation de phase tournée prédéterminée à au moins quatre dimensions. 2 rounds of MAQ and MAQ types are respectively preceded by quantifiers having non-uniform quantization laws whose levels are respectively subsets of projections relating to points of a constellation of phase tour predetermined at least four dimensions.
EP99943017A 1999-09-21 1999-09-21 Block joint source-channel coding Withdrawn EP1131928A1 (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/FR1999/002241 WO2001022676A1 (en) 1999-09-21 1999-09-21 Block joint source-channel coding

Publications (1)

Publication Number Publication Date
EP1131928A1 true EP1131928A1 (en) 2001-09-12

Family

ID=9542424

Family Applications (1)

Application Number Title Priority Date Filing Date
EP99943017A Withdrawn EP1131928A1 (en) 1999-09-21 1999-09-21 Block joint source-channel coding

Country Status (2)

Country Link
EP (1) EP1131928A1 (en)
WO (1) WO2001022676A1 (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8089892B2 (en) 2005-12-15 2012-01-03 Thomson Licensing Adaptive joint source and channel coding scheme for H.264 video multicasting over wireless networks

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE69427024T2 (en) * 1993-12-03 2001-10-18 Koninkl Philips Electronics Nv DIGITAL MEDIATION SYSTEM AND RECEIVER FOR USE IN IT
SE506379C3 (en) * 1995-03-22 1998-01-19 Ericsson Telefon Ab L M Lpc speech encoder with combined excitation
ATE192259T1 (en) * 1995-11-09 2000-05-15 Nokia Mobile Phones Ltd METHOD FOR SYNTHESIZING A VOICE SIGNAL BLOCK IN A CELP ENCODER
FR2776447B1 (en) * 1998-03-23 2000-05-12 Comsis JOINT SOURCE-CHANNEL ENCODING IN BLOCKS

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See references of WO0122676A1 *

Also Published As

Publication number Publication date
WO2001022676A1 (en) 2001-03-29

Similar Documents

Publication Publication Date Title
EP0891656B1 (en) Data block convolutional coding device and method, and corresponding decoding method and device
US20210319286A1 (en) Joint source channel coding for noisy channels using neural networks
Wade Signal coding and processing
EP0448492B1 (en) Apparatus for the transmission of digital data with at least two levels of protection, and corresponding receiver
JP2509784B2 (en) Apparatus and method for transmitting on fading channel
Kurka et al. Successive refinement of images with deep joint source-channel coding
US20210351863A1 (en) Joint source channel coding based on channel capacity using neural networks
EP0481549B1 (en) Coding/decoding system and method for digital signals transmitted by coded modulation
EP1168739B1 (en) Method and apparatus for estimating the impulse response of a transmission channel, in particular for a cellular mobile telephone
EP2176861B1 (en) Hierarchical coding of digital audio signals
EP2609704B1 (en) Method and device for transmission and reception in a multi-input multi-output channel, distributing a code word between multiple mapping matrices, and corresponding computer program
GB2576702A (en) Joint source channel coding of information sources using neural networks
EP2070199A1 (en) Method of spatio-temporal coding for multi-antenna communication system of pulsed uwb type
FR2754127A1 (en) Video coder and decoder using adaptive lattice quantiser
FR2899745A1 (en) SPATIO-TEMPORAL ENCODING METHOD FOR MULTI-ANTENNA COMMUNICATION SYSTEM OF IMPULSE UWB TYPE
EP0594505A1 (en) Method for hierarchical coding/decoding of a digital signal and corresponding system used for digital television
EP1668794A1 (en) Method for the multi-antennae emission of a signal by unitary space-time codes, receiving method, and corresponding signal
EP2180626B1 (en) Distributed turbo encoder for block fading channels
JP4220365B2 (en) Transmitting apparatus, receiving apparatus, data transmitting method, and data receiving method
EP0961264B1 (en) Emitting/receiving device for the selection of a source coder and methods used therein
WO2001022676A1 (en) Block joint source-channel coding
FR2776447A1 (en) Digital word compressing coding method
Lervik et al. Robust digital signal compression and modulation exploiting the advantages of analog communications
Lu et al. Reducing signal distortion due to transmission error via multiresolution digital modulations
Fuldseth et al. Combined video coding and multilevel modulation

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20010312

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE

RBV Designated contracting states (corrected)

Designated state(s): DE FR GB

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN

18D Application deemed to be withdrawn

Effective date: 20040401