WO1996034383A1 - Procede et dispositif de codage et decodage a faible debit binaire - Google Patents

Procede et dispositif de codage et decodage a faible debit binaire Download PDF

Info

Publication number
WO1996034383A1
WO1996034383A1 PCT/US1996/003235 US9603235W WO9634383A1 WO 1996034383 A1 WO1996034383 A1 WO 1996034383A1 US 9603235 W US9603235 W US 9603235W WO 9634383 A1 WO9634383 A1 WO 9634383A1
Authority
WO
WIPO (PCT)
Prior art keywords
codeword
quantizer
prequantizer
signal
plural
Prior art date
Application number
PCT/US1996/003235
Other languages
English (en)
Inventor
James Patrick Ashley
Weimin Peng
Peng Zhang
Original Assignee
Motorola Inc.
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 Motorola Inc. filed Critical Motorola Inc.
Priority to BR9606292A priority Critical patent/BR9606292A/pt
Priority to DE19680399T priority patent/DE19680399T1/de
Priority to MX9606368A priority patent/MX9606368A/es
Priority to PL96318012A priority patent/PL318012A1/xx
Priority to AU66349/96A priority patent/AU683058B2/en
Priority to JP8532495A priority patent/JPH10503633A/ja
Priority to GB9625742A priority patent/GB2304508A/en
Publication of WO1996034383A1 publication Critical patent/WO1996034383A1/fr
Priority to SE9604443A priority patent/SE9604443L/xx
Priority to FI965131A priority patent/FI965131A/fi

Links

Classifications

    • 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
    • G10L19/02—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 spectral analysis, e.g. transform vocoders or subband vocoders
    • G10L19/032—Quantisation or dequantisation of spectral components
    • G10L19/038—Vector quantisation, e.g. TwinVQ audio

Definitions

  • the present invention relates, in general, to communication systems and, more particularly, to a method and apparatus for coding/decoding a signal for communication.
  • Source coding is used extensively in modern communications in order to provide more efficient ways of communicating a signal.
  • One common type of source coder is a vocoder, in which a voice signal is coded for transmission.
  • a vocoder is a linear predictive coder (LPC).
  • LPC linear predictive coder
  • each input speech vector is mapped to LPC parameters, which in turn are mapped to a best-fit codeword of a finite codebook.
  • An example of such a codeword would be one which upon filtering/decoding results in the least residual energy when compared with the original input speech.
  • FIG. 1 is a diagram illustrating the logical structure of a codebook for use in accordance with an embodiment of the invention
  • FIG. 2 is a flow-chart illustrating a preferred method for low-rate encoding in accordance with the invention
  • FIG. 3 is a block diagram illustrating a first embodiment apparatus according to the invention.
  • a presently preferred embodiment of the invention is a method in which a prequantizer search is performed to determine a best prequantizer (“P") codeword (i.e., having a least prequantizer residual component), and a restricted quantizer search is performed to determine the best quantizer ("Q") codeword (i.e., having the least quantizer residual component).
  • the residual components are then compared to determine which is least, and the corresponding index of the codeword yielding the least residual is formatted for transmission, along with a mode indicator indicating which type of index (quantizer or prequantizer) is being sent.
  • the coded signal is fed to a vocoder which first determines which mode indicator was sent. If a quantizer (“V”) mode indicator (e.g., a 0 bit) is received, then the vocoder uses the "V" codeword corresponding to the received V index along with a previously determined P codeword to form a replica of the original signal.
  • V quantizer
  • the vocoder uses the P codeword corresponding to the received P index to form the signal replica.
  • P prequantizer
  • vector quantizer 100 is a two-level, tree-structured vector quantizer. This structure illustrates the prequantizer and quantizer hierarchy used in a typical predetermined vector quantizer.
  • the quantizer stage 110 includes 1 through N prequantizer codewords, e.g., codeword Pi 111 , each P codeword corresponding to one of plural prequantizer ("P") indices of the predetermined vector quantizer.
  • the lower quantizer stage 120 includes plural sets of plural quantizer (“V”) codewords 121 -123, 131 -133, each quantizer ("V”) codeword of a set corresponding to single one of plural quantizer (“V”) indices.
  • Each set of V codewords is associated with a P codeword.
  • the first set of V codewords 121 -123 is associated with P codeword 111 , in which P codeword 111 forms a centroid (typically of lower resolution) for the (typically higher resolution) V codeword vectors 121 -123.
  • the tree structure of FIG. 1 would be representative of each of the vector segments that are used in determining a -eflection coefficient vector estimate.
  • this reflection co-efficient vector estimate might typically be segmented as follows:
  • the following bid allocation is representative of what one would expect in a full or half-rate sub-frame in each of the respective three vector segments:
  • the P codewords represent the prequantizer cells of the codebook and are used to code the coarse spectral information, while the V codewords are used to code the fine or detailed spectral information.
  • the prequantizer index (P index) for each respective segment will typically be highly correlated to the previously determined prequantizer index.
  • Table 2 illustrates one approach for further reducing the number of bits required for transmission at lower rates.
  • the least significant bits for certain of the P and V codewords are dropped prior to transmission.
  • the 1/8th rate P codeword for segment 1 would only include the five most significant bits of the stored P codeword, which as shown in Table 1 is 6 bits in length.
  • certain other codewords would be sent in their entirety (e.g., the V codeword of segment 1 ).
  • the P and V codewords within the codebook can be defined, as well as the bit structure that is transmitted at the different rates. What should be appreciated is that by this alternate approach further reduction in the size of the codeword index (or indices) for transmission is achieved. Thus, for example, where only a 16 bit structure is permitted for 1/8th rate framing, only 12 bits of vector quantizer information are needed, leaving one bit for a mode indicator and three bits for the coding of difference energy.
  • the mode indicator bit would either be used as a quantizer mode indicator (V mode indicator) when only the V index is to be transmitted, and a prequantizer mode indicator (P mode indicator, e.g., a 1 bit) when the P index is being transmitted .
  • FIG. 2 a flow chart illustrating a preferred embodiment of a coding process according to the invention as shown.
  • This process commences with an initial determination that the vocoder is in 1/8th rate mode (Step 210).
  • a prequantizer search (step 212) is then performed on an input signal. This search is performed by comparing each P codeword of the codebook with the signal and determining the P codeword having the least prequantizer residual component (e.g., energy).
  • the vector quantizer is a plural-vector-segment vector quantizer (e.g., three segments)
  • a determination of the best P codeword for each vector segment is made based on the P codewords having the least prequantizer residual energy, then each of the best P codewords are vector added to form a prequantizer vector sum, and the least prequantizer residual component is determined using this vector sum.
  • the prequantizer search may be restricted to only those P codewords corresponding to the P indices that can be transmitted, where only a restricted number of bits can be used in transmission (such as shown in Table 2).
  • a restricted quantizer search is performed.
  • the input signal is again compared, but this time against each quantizer codeword (or V codewords, each corresponding to one of plural quantizer indices of the codebook) of one set, or one set per segment if multiple vector segments are used.
  • This set is determined by a prev iously determined P codeword, which forms the prequantizer (or centroid) of the set of V codewords.
  • the previously determined P codeword could be a predetermined initialization P codeword (which for convenience could simply be the P codeword having index 000000 for a Table 2 segment 1 set).
  • the last P codeword transmitted is used (both transmitting and receiving vocoders thereby possessing the P codeword).
  • the V codeword of the set having the least quantizer residual component is then determined, and its corresponding V index stored in memory.
  • the best V codeword of the predetermined set of V codewords is determined for each vector segment if there are plural segments, with a quantizer vector sum being formed by vector adding each of the best V codewords, and a least quantizer residual component being determined therefrom.
  • a least quantizer residual component could be determined for each V codeword pei segment, or more simply just that of the first segment; these value(s) would in turn then be used in step 220.
  • a least quantizer residual component could be determined for each V codeword pei segment, or more simply just that of the first segment; these value(s) would in turn then be used in step 220.
  • One skilled in the art will appreciate that other alternatives are also available in forming a residual measure for use in comparing and determining which index to use.
  • step 220 a comparison is made between the least quantizer residual component and the least prequantizer residual component to determine which one is less.
  • a prequantizer mode indicator P mode indicator
  • V mode indicator a quantizer mode indicator
  • step 230 a further determination is made as to whether a higher rate of coding is needed, for better resolution.
  • both residual components are compared against a predetermined threshold to determine if too much residual energy is present. This is an indication that an unacceptable signal quality will be present when decoded.
  • a predetermined threshold One skilled in the art will know how to design an appropriate threshold, based upon such factors as the application involved, user preferences and the like. If the threshold is exceeded, the vocoder is forced to a higher rate (step 232). If the threshold is not exceeded, the determined index and mode indicator, along with other desired signal parameters (e.g., the quantizer residual energy for certain vocoders) are then formed into the coded signal and transmitted (step 234). The process is then repeated for the next signal (step 236).
  • desired signal parameters e.g., the quantizer residual energy for certain vocoders
  • FIG. 3 illustrates one embodiment of a communication system in which low-rate coding according to the invention is used.
  • an input signal 301 such as speech is received by variable rate vocoder 310
  • the signal is fed to both prequantizer searcher 312 and quantizer searcher 314.
  • Both searchers perform a comparison in accordance with steps 212 and 214, respectively, of FIG. 2 (using the P and V indices to retrieve the P and V codewords, respectively, from codebook 31 1 ).
  • the determined P index and residual component (signal 313) and V index and residual component (signal 315) are fed to a coder 316 for determination of which low-rate mode (if either) is to be used.
  • a coded version of the signal is then formed using the apprrpriate index and mode indicator, along with other appropriate (conventional) parameters.
  • This coded signal (signal 317) is then modulated in modulator 320 and amplified and transmitted via transmitter 325 of unit 300.
  • a receiver communication unit 330 receives and demodulates the transmitted signal via receiver 335 and demodulator 340, outputting a coded version of the original signal.
  • a decoder e.g., vocoder 350
  • Controller 352 determines if the coded signal includes a V or P mode indicator, and if not which of the higher decoding rates is applicable.
  • controller 352 controls signal generator 354 to retrieve the P codeword corresponding to the received P index from codebook 351 , and generate a replica of the original signal using the P codeword, along with other received signal parameters (e.g., the residual energy).
  • the signal generate 354 is controlled to use bo* the previously received P codeword (i.e., the P codeword used in the prior frame) and the V codeword corresponding to the received V index, along with other signal parameters, to generate the signal replica.
  • the signal replica is outputted as output 356 for further processing, such as transduction into an audible signal.
  • the invention is not limited in application to just vocoders in cellular communication systems, but also applies to other source coders for other types of communication systems. Accordingly, the invention is intended to embrace all such alterations, modifications, and variations within the spirit and scope of the appended claims.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Human Computer Interaction (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Compression, Expansion, Code Conversion, And Decoders (AREA)
  • Transmission Systems Not Characterized By The Medium Used For Transmission (AREA)
  • Executing Machine-Instructions (AREA)

Abstract

Système destiné à un codage ou décodage à faible débit binaire, comportant l'utilisation d'un indicateur de mode dans le signal codé pour indiquer s'il est envoyé une information émanant d'un préquantificateur (c'est-à-dire grossière) ou d'un quantificateur (c'est-à-dire fine). Dans un mode de réalisation d'émetteur (300), on fait une recherche de préquantificateur (212) et une recherche restreinte de quantificateur (214) sur les mots de code prédéfinis d'un préquantificateur ('P') et d'un quantificateur ('V'), respectivement, afin de déterminer les mots de code ayant l'énergie résiduelle la plus faible. L'indice du mot de code P ou V ayant la plus petite énergie résiduelle est ensuite déterminé (220), puis transmis en même temps qu'un indicateur de mode P ou V approprié. Dans un mode de réalisation de récepteur (330), l'indicateur de mode est utilisé pour régler le décodeur (par exemple un vocodeur) (350) en sorte de former une réplique du signal à partir d'un mot de code P (déterminé à partir de l'indice transmis) ou à partir d'un mot de code V (à partir de l'indice transmis) et d'un mot de code P préalablement déterminé (par exemple le point médian du mot de code V).
PCT/US1996/003235 1995-04-19 1996-03-07 Procede et dispositif de codage et decodage a faible debit binaire WO1996034383A1 (fr)

Priority Applications (9)

Application Number Priority Date Filing Date Title
BR9606292A BR9606292A (pt) 1995-04-19 1996-03-07 Processo para vocodificação de taxa baixa sistema para codificação de taxa baixa vocodificador de taxa cariavel unidade de comunicação método para processar sinais vocodificadores e taxa baixa
DE19680399T DE19680399T1 (de) 1995-04-19 1996-03-07 Verfahren und Vorrichtung zur niederratigen Kodierung und Dekodierung
MX9606368A MX9606368A (es) 1995-04-19 1996-03-07 Metodo y aparato para codificar y decodificar a baja velocidad.
PL96318012A PL318012A1 (en) 1995-04-19 1996-03-07 Method of and device for low speed encoding and decoding
AU66349/96A AU683058B2 (en) 1995-04-19 1996-03-07 Method and apparatus for low rate coding and decoding
JP8532495A JPH10503633A (ja) 1995-04-19 1996-03-07 低レートコーディングおよびデコーディングのための方法および装置
GB9625742A GB2304508A (en) 1995-04-19 1996-03-07 Method and apparatus for low rate coding and decoding
SE9604443A SE9604443L (sv) 1995-04-19 1996-12-03 Metod och anordning för låghastighetskodning och -avkodning
FI965131A FI965131A (fi) 1995-04-19 1996-12-19 Menetelmä ja laite pieninopeuksista koodausta ja dekoodausta varten

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US42478095A 1995-04-19 1995-04-19
US08/424,780 1995-04-19

Publications (1)

Publication Number Publication Date
WO1996034383A1 true WO1996034383A1 (fr) 1996-10-31

Family

ID=23683842

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1996/003235 WO1996034383A1 (fr) 1995-04-19 1996-03-07 Procede et dispositif de codage et decodage a faible debit binaire

Country Status (14)

Country Link
JP (1) JPH10503633A (fr)
CN (1) CN1150853A (fr)
AU (1) AU683058B2 (fr)
BR (1) BR9606292A (fr)
CA (1) CA2191972A1 (fr)
DE (1) DE19680399T1 (fr)
FI (1) FI965131A (fr)
FR (1) FR2733345B1 (fr)
GB (1) GB2304508A (fr)
IT (1) IT1285178B1 (fr)
MX (1) MX9606368A (fr)
PL (1) PL318012A1 (fr)
SE (1) SE9604443L (fr)
WO (1) WO1996034383A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008030673A2 (fr) * 2006-07-31 2008-03-13 Qualcomm Incorporated Systèmes, procédés et appareil destinés à limiter le facteur de gain

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104637483A (zh) * 2015-02-03 2015-05-20 中国电子科技集团公司第五十八研究所 基于多通道的低速率语音编解码系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4963030A (en) * 1989-11-29 1990-10-16 California Institute Of Technology Distributed-block vector quantization coder
EP0480083A1 (fr) * 1989-06-21 1992-04-15 Interdigital Technology Corporation Procédé et système de compression d'un signal de communication
FR2709366A1 (fr) * 1993-03-26 1995-03-03 Motorola Inc Procédé de stockage de vecteurs de coefficient de réflexion.

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE6901749U (de) * 1969-01-17 1969-05-22 Huettenwerke Siegerland Ag Diffusor fuer bundgluehoefen
JPH02287399A (ja) * 1989-04-28 1990-11-27 Fujitsu Ltd ベクトル量子化制御方式
US5010574A (en) * 1989-06-13 1991-04-23 At&T Bell Laboratories Vector quantizer search arrangement

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0480083A1 (fr) * 1989-06-21 1992-04-15 Interdigital Technology Corporation Procédé et système de compression d'un signal de communication
US4963030A (en) * 1989-11-29 1990-10-16 California Institute Of Technology Distributed-block vector quantization coder
FR2709366A1 (fr) * 1993-03-26 1995-03-03 Motorola Inc Procédé de stockage de vecteurs de coefficient de réflexion.

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
LI ET AL.: "Fast vector quantization based on lattice prequantization", INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING 1989, vol. 2, 23 May 1989 (1989-05-23) - 26 May 1989 (1989-05-26), GLASGOW, GB, pages 759 - 761, XP000090221 *
MIYANO ET AL.: "IMPROVED 4.8KB/S CELP CODING USING TWO-STAGE VECTOR QUANTIZATION WITH MULTIPLE CANDIDATES (LCELP)", SPEECH PROCESSING, vol. 1, 23 March 1992 (1992-03-23) - 26 March 1992 (1992-03-26), SAN FRANCISCO, CA US, pages 321 - 324, XP000341148 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008030673A2 (fr) * 2006-07-31 2008-03-13 Qualcomm Incorporated Systèmes, procédés et appareil destinés à limiter le facteur de gain
WO2008030673A3 (fr) * 2006-07-31 2008-06-26 Qualcomm Inc Systèmes, procédés et appareil destinés à limiter le facteur de gain
US9454974B2 (en) 2006-07-31 2016-09-27 Qualcomm Incorporated Systems, methods, and apparatus for gain factor limiting

Also Published As

Publication number Publication date
FI965131A0 (fi) 1996-12-19
AU6634996A (en) 1996-11-18
ITRM960235A1 (it) 1997-10-15
FR2733345A1 (fr) 1996-10-25
CA2191972A1 (fr) 1996-10-31
PL318012A1 (en) 1997-05-12
DE19680399T1 (de) 1997-09-18
GB2304508A (en) 1997-03-19
IT1285178B1 (it) 1998-06-03
JPH10503633A (ja) 1998-03-31
CN1150853A (zh) 1997-05-28
AU683058B2 (en) 1997-10-30
ITRM960235A0 (fr) 1996-04-15
FR2733345B1 (fr) 1997-07-25
GB9625742D0 (en) 1997-01-29
SE9604443D0 (sv) 1996-12-03
FI965131A (fi) 1996-12-19
SE9604443L (sv) 1997-02-19
MX9606368A (es) 1997-03-29
BR9606292A (pt) 1997-11-18

Similar Documents

Publication Publication Date Title
AU716982B2 (en) Method for signalling a noise substitution during audio signal coding
US5692102A (en) Method device and system for an efficient noise injection process for low bitrate audio compression
US7280959B2 (en) Indexing pulse positions and signs in algebraic codebooks for coding of wideband signals
US6064954A (en) Digital audio signal coding
EP1905000B1 (fr) Utilisation selective de plusieurs modeles entropiques pour le codage et le decodage adaptatifs
EP1904999B1 (fr) Segmentation de frequence permettant d'obtenir des bandes de codage efficace de donnees multimedia numeriques
CN109313908B (zh) 用于对音频信号进行编码的音频编码器以及方法
JP3378238B2 (ja) ソフト適応性特性を含む音声コーディング
US7756698B2 (en) Sound decoder and sound decoding method with demultiplexing order determination
AU2002221389A1 (en) Indexing pulse positions and signs in algebraic codebooks for coding of wideband signals
JP2001500640A (ja) オーディオ信号の符号化方法
US20050091044A1 (en) Method and system for pitch contour quantization in audio coding
Gardner et al. QCELP: A variable rate speech coder for CDMA digital cellular
RU2223555C2 (ru) Адаптивный критерий кодирования речи
EP0747884A2 (fr) Atténuation de gain de dictionnaire en cas de pertes des paquets de données
US8498875B2 (en) Apparatus and method for encoding and decoding enhancement layer
AU683058B2 (en) Method and apparatus for low rate coding and decoding
EP1199710B1 (fr) Dispositif, procédé et mémoire avec une programme enregistrée de decodage de la voix en parties non vocales
JP3197156B2 (ja) ディジタル音声コーダ及びデコーダにおけるスペクトルパラメータを量子化及び逆量子化する方法及び装置
JP3308783B2 (ja) 音声復号化装置
EP0906664B1 (fr) Systeme de transmission de la parole
EP0896321B1 (fr) Méthode et système pour la compression et la décompression de son
US8502706B2 (en) Bit allocation for encoding track information
KR100468960B1 (ko) 음성부호화 시스템의 양방향 피치 강화 시스템
JPH0728500A (ja) 音声符号器及び音声復号器

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 96190352.X

Country of ref document: CN

AK Designated states

Kind code of ref document: A1

Designated state(s): AM AT AU BB BG BR BY CA CH CN CZ DE DK EE ES FI GB GE HU IS JP KE KG KP KR KZ LK LR LT LU LV MD MG MN MW MX NO NZ PL PT RO RU SD SE SG SI SK TJ TM TT UA UG US UZ VN

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): KE LS MW SD SZ UG AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN ML MR NE SN TD TG

WWE Wipo information: entry into national phase

Ref document number: 2191972

Country of ref document: CA

Ref document number: 96044433

Country of ref document: SE

WWE Wipo information: entry into national phase

Ref document number: PA/a/1996/006368

Country of ref document: MX

WWE Wipo information: entry into national phase

Ref document number: 965131

Country of ref document: FI

121 Ep: the epo has been informed by wipo that ep was designated in this application
WWP Wipo information: published in national office

Ref document number: 96044433

Country of ref document: SE

RET De translation (de og part 6b)

Ref document number: 19680399

Country of ref document: DE

Date of ref document: 19970918

WWE Wipo information: entry into national phase

Ref document number: 19680399

Country of ref document: DE

122 Ep: pct application non-entry in european phase