CN101127211A

CN101127211A - Method for decoding audio frequency signal and system for transmitting audio frequency signal

Info

Publication number: CN101127211A
Application number: CNA2007101368716A
Authority: CN
Inventors: 杰斯·赛森; 朱因韦·陈
Original assignee: Zyray Wireless Inc
Current assignee: Broadcom Corp; Zyray Wireless Inc
Priority date: 2006-07-12
Filing date: 2007-07-12
Publication date: 2008-02-20

Abstract

A system and method for encoding and decoding speech signals that includes a specially-designed Code Excited Linear Prediction (CELP) encoder and a vector quantization (VQ) based Noise Feedback Coding (NFC) decoder or that includes a specially-designed VQ-based NFC encoder and a CELP decoder. The VQ based NFC decoder may be a VQ based two-stage NFC (TSNFC) decoder. The specially-designed VQ-based NFC encoder may be a specially-designed VQ based TSNFC encoder. In each system, the encoder receives an input speech signal and encodes it to generate an encoded bit stream. The decoder receives the encoded bit stream and decodes it to generate an output speech signal. A system and method is also described in which a single decoder receives and decodes both CELP-encoded audio signals as well as VQ-based NFC-encoded audio signals.

Description

The method of decoded audio signal and the system of transmit audio signals

Technical field

The present invention relates to be used for the system of Code And Decode voice and/or sound signal.

Background technology

In 20 years, Qualcomm Code Excited Linear Prediction (QCELP) (CELP) technology has become most popular and topmost speech coding technology in the past.The research of concentrating has been carried out in voice quality and implementation efficiency aspect to the CELP principle.The research paper of relevant CELP reaches hundreds of even several thousand pieces.In fact, CELP has become the basis of the most international voice coding standard of issue since 1988.

Recently, demonstrate encode equal output voice quality and encoding and decoding complexity gradually based on two stage noise feedback encodings (TSNFC) of vector quantization (VQ) with CELP.The BroadVoice of the Botong company exploitation in california that gulf ^16 (BV16) are based on the TSNFC codec of VQ, and it is by CableLab ^Give standardization, as the PacketCable that is applied to cable phone (cable telephone) ^TM1.5 the mandatory audio coder ﹠ decoder (codec) of standard.BV16 still is SCTE (cable Telecommunications Engineer association) standard, ANSI American National Standard, and the ITU-T Recommendation recommendation codec in the standard J.161.In addition, BV16 and BroadVoice ^32 (BV32, another TSNFC codecs based on VQ of the Botong company exploitation in California that gulf) are PacketCable ^TM2.0 the part of standard.The invention people who announces on Dec 27th, 2005 based on an example of the TSNFC codec of VQ is described in the total United States Patent (USP) 6980951 of Chen, and this paper quotes wherein full content at this.

CELP and TSNFC are considered to two kinds of distinct voice coding methods.Therefore, be used for common a kind of structure that only centers on wherein of system of encoded voice and/or sound signal, but not make up around two kinds of methods simultaneously.But, use celp coder to carry out interoperability with TSNFC demoder (as BV16 or BV32), perhaps use TSNFC scrambler and CELP demoder, can access potential benefit.Yet, the current solution that does not also realize this purpose.

Summary of the invention

As hereinafter will describing in detail, the invention provides a kind of system and method, by this system and method, Qualcomm Code Excited Linear Prediction (QCELP) (CELP) scrambler can carry out compounding practice as two stage NFC (TSNFC) demoders based on VQ with noise feedback encoding (NFC) demoder based on vector quantization (VQ); By this system and method, can carry out compounding practice with the CELP demoder as TSNFC scrambler based on VQ based on the NFC scrambler of VQ.In addition, the present invention also provides a kind of system and method, by this system and method, and celp coder and all can carry out compounding practice with single demoder based on the NFC scrambler of VQ.

Specifically, one embodiment of the present of invention provide a kind of method of decoded audio signal.According to this method, received code bit stream at first.This coded bit stream is represented input audio signal, and as input speech signal, it is encoded by celp coder.This coded bit stream uses NFC demoder based on VQ as the TSNFC decoder decode based on VQ subsequently, generates output audio signal, as the output voice signal.This method also comprises and at first receives described input audio signal, uses celp coder this input audio signal of encoding then, generates described coded bit stream.

One embodiment of the present of invention provide a kind of system of transmit audio signals.This system comprises celp coder and based on the NFC demoder of VQ.This celp coder be used to encode input audio signal such as input speech signal generate coded bit stream.Based on the NFC demoder of the VQ described coded bit stream that is used to decode, generate output audio signal, as the output voice signal.Can comprise TSNFC demoder based on the NFC demoder of VQ based on VQ.

One embodiment of the present of invention provide the another kind of method that is used for decoded audio signal.According to this method, received code bit stream at first.This coded bit stream is represented input audio signal, and as input speech signal, it is encoded as the TSNFC scrambler based on VQ by the NFC scrambler based on VQ.This coded bit stream uses the CELP decoder decode subsequently, generates output audio signal, as the output voice signal.This method also comprises and at first receives described input audio signal, uses this input audio signal of NFC encoder encodes based on VQ then, generates described coded bit stream.

One embodiment of the present of invention provide the another kind of system that is used for transmit audio signals.This system comprises based on the NFC scrambler of VQ and CELP demoder.NFC scrambler based on VQ is used to the input audio signal of encoding, and as input speech signal, generates coded bit stream.The CELP demoder described coded bit stream that is used to decode generates output audio signal, as the output voice signal.Described NFC scrambler based on VQ can comprise the TSNFC scrambler based on VQ.

One embodiment of the present of invention also provide a kind of method that is used for decoded audio signal.According to this method, at first receive first coded bit stream.Described first coded bit stream representative is by first input audio signal of celp coder coding.This first coded bit stream is decoded in demoder, generates first output audio signal.This method also comprises reception second coded bit stream.The representative of this second coded bit stream by based on the NFC scrambler of VQ as second input audio signal based on the TSNFC encoder encodes of VQ.Described second coded bit stream is decoded in described demoder equally, generates second output audio signal.Described first and second input audio signals can comprise input speech signal, and described first and second output audio signals can comprise the output voice signal.

One embodiment of the present of invention also provide a kind of system that is used for transmit audio signals.This system comprises celp coder, based on the NFC scrambler of VQ, and demoder.Described celp coder first input audio signal that is used to encode generates first coded bit stream.Described NFC scrambler based on VQ second input audio signal that is used to encode generates second coded bit stream.Described demoder described first coded bit stream that is used to decode generates first output audio signal; And described second coded bit stream of decoding, generate second sound signal of publishing books.Described first and second input audio signals can comprise input speech signal, and described first and second output audio signals can comprise the output voice signal.Can comprise TSNFC scrambler based on the NFC scrambler of VQ based on VQ.

The operation of other features and advantages of the present invention and structure and a plurality of embodiment will be described in detail in conjunction with the accompanying drawings.It should be noted that the present invention is not limited only to specific embodiment described herein.Only be just to have enumerated these embodiment for purposes of illustration.Based on the instruction of this paper, other embodiment should be very clear for others skilled in the art.

Description of drawings

The invention will be further described below in conjunction with drawings and Examples, in the accompanying drawing:

Fig. 1 is the structural representation that comprises the conventional audio Code And Decode system of the scrambler of traditional two stage noise feedback encodings (TSNFC) based on vector quantization (VQ) and traditional TSNFC demoder based on VQ;

Fig. 2 comprises Qualcomm Code Excited Linear Prediction (QCELP) (CELP) scrambler and the audio coding of traditional TSNFC demoder based on VQ and the structural representation of decode system according to one embodiment of the invention;

Fig. 3 is the structural representation that comprises the conventional audio Code And Decode system of traditional C ELP scrambler and traditional C ELP demoder;

Fig. 4 is TSNFC scrambler and the audio coding of traditional C ELP demoder and the structural representation of decode system based on VQ of comprising according to one embodiment of the invention;

Fig. 5 be according to one embodiment of the invention be used for encode and quantize the functional block diagram of the system of pumping signal based on input audio signal;

Fig. 6 is the synoptic diagram according to the example structure of the excitation quantization modules in the TSNFC scrambler of one embodiment of the invention;

Fig. 7 is the synoptic diagram according to the example structure of the excitation quantization modules in the celp coder of one embodiment of the invention;

Fig. 8 is the structural representation that can be used for realizing common demoder of the present invention;

Fig. 9 is the process flow diagram according to the method for transmit audio signals of one embodiment of the invention such as voice signal;

Figure 10 is the process flow diagram according to the method for transmit audio signals of another embodiment of the present invention such as voice signal;

Figure 11 is the structural representation according to the system of one embodiment of the invention, and single demoder wherein both had been used for decoding by CELP method bitstream encoded, and NFC method bitstream encoded by based on VQ is used to again decode;

Figure 12 is the process flow diagram according to the method for transmit audio signals of further embodiment of this invention such as voice signal;

Figure 13 is the synoptic diagram that is used to realize an example of computer system of the present invention.

Embodiment

Although use to exist significantly differently between Qualcomm Code Excited Linear Prediction (QCELP) (CELP) and the coder structure based on two stage noise feedback encodings (TSNFC) of vector quantization (VQ), the corresponding decoder architecture of two kinds of methods of embodiments of the invention hypothesis is identical.As a rule, the task of celp coder or TSNFC scrambler is to be that unit generates and quantizes pumping signal, excitation gain and long-term predictor parameter and short-term forecasting device parameter with the frame.Suppose that the CELP demoder is identical with the TSNFC demoder, for specific TSNFC decoder architecture (as the decoder architecture of BV16), then can design a kind of celp coder, it can reach the target that the TSNFC scrambler can reach, just can generate and quantize pumping signal, excitation gain and Prediction Parameters, make decoding that the TSNFC demoder can be correct bit stream by the celp coder compression.In other words, might design a kind of celp coder compatible mutually with specifying the TSNFC demoder.

This notion is described in Fig. 1 and Fig. 2.Specifically, Fig. 1 be comprise traditional based on VQ TSNFC scrambler 110 and the structural representation of the conventional audio Code And Decode system 100 of traditional TSNFC demoder 120 based on VQ.Scrambler 110 compressible input audio signal such as input speech signals generate the TSNFC coded bit stream based on VQ.Demoder 120 decodable codes generate output audio signal as the output voice signal based on the TSNFC coded bit stream of VQ.Scrambler 110 and demoder 120 can be by the BroadVoice of the Botong company exploitation in your gulf of california for example ^16 (BV16) codec and BroadVoice ^32 (BV32) codec is realized.

Fig. 2 is the structural representation according to audio coding of one embodiment of the invention and decode system 200, and it is equivalent to the legacy system 100 among Fig. 1 on function.In system 200, traditional TSNFC demoder 220 based on VQ is equivalent to the TSNFC demoder 120 based on VQ in the system 100.But traditional TSNFC scrambler 110 based on VQ is substituted by celp coder 210, and the latter comes out according to one embodiment of the invention specialized designs, can be compatible mutually with the TSNFC demoder 220 based on VQ.Because the CELP demoder is equivalent to the TSNFC demoder based on VQ, therefore can regards TSNFC demoder 220 as the CELP demoder, thereby design the celp coder 210 that can carry out compounding practice with demoder 220 based on VQ.

Embodiments of the invention are also supposed the CELP demoder for appointment, as ITU-TRecommendation demoder G.731, might design a kind of TSNFC scrambler based on VQ, and the bit stream of its generation is compatible mutually with the CELP demoder of appointment.

This notion is described in Fig. 3 and Fig. 4.Specifically, Fig. 3 is the structural representation that comprises the conventional audio Code And Decode system 300 of traditional C ELP scrambler 310 and traditional C ELP demoder 320.Scrambler 310 is used to compress input audio signal such as input speech signal, generates the CELP coded bit stream.The demoder 320 CELP coded bit stream that is used to decode generates output audio signal, as the output voice signal.Scrambler 310 and demoder 320 for example can use ITU-T G.723.1 codec realize.

Fig. 4 is the structural representation according to audio coding of one embodiment of the invention and decode system 400, and it is equivalent to the legacy system 300 among Fig. 3 on function.In system 400, traditional CELP demoder 420 is equivalent to the traditional C ELP demoder 320 in the system 300.But traditional celp coder 310 is substituted by the TSNFC scrambler 410 based on VQ, and the latter is custom-designed according to one embodiment of the invention, can be compatible mutually with CELP demoder 420.Owing to be equivalent to the CELP demoder based on the TSNFC demoder of VQ, therefore CELP demoder 420 can be regarded as TSNFC demoder based on VQ, thereby design can with the TSNFC scrambler 410 based on VQ of demoder 420 compounding practices.

A potential benefit using celp coder and TSNFC demoder such as BV16 or BV32 demoder to carry out compounding practice is, in the past in 20 years, carrying out the research of concentrating aspect the quality improvement of CELP coding techniques and the complexity reduction.Therefore, use celp coder can make full use of the achievement in research that these concentrate research.On the other hand, in some cases, use the TSNFC scrambler also can provide some benefit and advantage.Therefore, the present invention has substantial technique effect and value.

It should be noted that, though the TSNFC encoder that is based on VQ of Shi Yonging in the above-described embodiments, the present invention also can use existing based on VQ single phase NFC demoder (with reference to the embodiment among the figure 2) or custom-designedly realize based on VQ single phase NFC scrambler (as shown in Figure 4).Therefore, for example, in one embodiment of the invention, can be used in combination custom-designed single phase NFC scrambler and the G.728 low CELP of the delay demoder of ITU-T Recommendation based on VQ.Those skilled in the art should be understood that G.728 codec is a kind of single phase prediction codec, and it uses the short-term forecasting device, and does not use long-term predictor.

The key distinction between CELP and the TSNFC scrambler is every kind of scrambler is how to encode and quantize pumping signal.Although every kind of method is used different incentive structures, there is overlapping part between these structures, thereby the coding of every kind of method and quantization operation are used interchangeably.Being used to carry out the corn module of these operations, for example being used to carry out the quantification of pre-filtering, estimation and linear predictive coding (LPC) coefficient, the functional module that pitch period is estimated operations such as (pitch period estimation), all is sharable.

This notion is described in Fig. 5, wherein showed according to one embodiment of the invention be used for encode and quantize the functional block diagram of the system 500 of pumping signal based on input audio signal.The same just as will be described in detail, according to the configuration mode of system 500, can use it for the celp coder 210 in the system 200 that realizes describing among Fig. 2, or the TSNFC scrambler 410 based on VQ in the system of describing among Fig. 4 400.

As shown in Figure 5, system 500 comprises pre-filtering module 502, lpc analysis module 504, LPC quantization modules 506, weighting block 508, rough pitch period (coarse pitch period) estimation module 510, pitch period accurate module 512, fundamental tone beat (pitch tap) estimation module 514 and excitation quantization modules 516.The working method of these modules will be carried out concise and to the point description below.

Pre-filtering module 502 is used to receive input audio signal such as input speech signal, and input audio signal is carried out filtering, generates the pre-filtering input audio signal.Lpc analysis module 504 is used to receive the pre-filtering input audio signal, therefrom generates the LPC coefficient.LPC quantization modules 506 is used for receiving the LPC coefficient from lpc analysis module 504, it is quantized the LPC coefficient of generating quantification.As shown in Figure 5, the LPC coefficient of these quantifications will mail to excitation quantization modules 516.

Weighting block 508 is used to receive the pre-filtering sound signal, therefrom generates weighting sound signal such as weighted speech signal.Rough pitch period estimation module 510 is used to receive the weighting sound signal, and selects rough pitch period based on this weighting sound signal.Pitch period accurate module 512 receives the rough pitch period of selecting, and it is carried out accurately, generates pitch period.Fundamental tone beat estimation module 514 is used to receive pre-filtering sound signal and accurate pitch period, generates one or more fundamental tone beats based on these input informations.As shown in Figure 5, pitch period and fundamental tone beat all will mail to excitation quantization modules 516.

Those skilled in the art should be very familiar to the function of each module in above-mentioned module 502,504,506,508,510,512,514 and 516, and can realize these modules.

Excitation quantization modules 516 receives LPC coefficient, pitch period and the fundamental tone beat of pre-filtering sound signal, quantification.Excitation quantization modules 516 is also encoded to pumping signal based on these input informations and is quantized.According to some embodiments of the present invention, excitation quantization modules 516 can use CELP technology (for example when system 500 is celp coder 210 a part of) to carry out the excitation coding and quantize, or is used to use TSNFC technology (for example when system 500 is based on TSNFC scrambler 410 a part of of VQ) to carry out that excitation is encoded and quantification.On principle, also can use other technologies.For example, a kind of selection is the open loop quantification acquisition pumping signal by long-term forecasting residual (long-term prediction residual).

No matter under which kind of situation, the structure of pumping signal (model that long-term forecasting just is residual) can not be changed by decoder architecture and bit stream definition decision.An example according to the general code device structure 800 of one embodiment of the invention is described in Fig. 8, and its content will be described in more detail below.

Those skilled in the art should be understood that and encourages quantization modules 516 and to select the pumping signal parameter of using in the scrambler by the any-mode estimation.The quality of the voice signal of rebuilding depends primarily on the method for using when excitation quantizes.TSNFC and CELP have all proved can provide high-quality signal under appropriate complexity situation, and open-loop method has lower complexity usually fully, but the signal quality that provides is also lower.

It should be noted that in some cases, Parameter Optimization and/or quantification quantized to carry out with excitation that in this sense, the functional module that is positioned at excitation quantization modules 516 outsides of showing among Fig. 5 also can be regarded the part of excitation quantification as.It should be noted that more the parameter relevant with fundamental tone is partially or completely to quantize to combine with excitation together to estimate and/or quantize sometimes.Therefore, those skilled in the art should be understood that the present invention is not limited only to the particular arrangement mode shown in Figure 5 and the definition mode of functional module, and other arrangement mode and definition also are feasible.

Fig. 6 is the synoptic diagram according to the example structure 600 of the excitation quantization modules in the TSNFC scrambler of one embodiment of the invention, Fig. 7 be according to one embodiment of the invention celp coder in the synoptic diagram of example structure 700 of excitation quantization modules.All can be used for the excitation quantization modules 516 in the realization system 500 in these structures.

At first, among Fig. 6 among structure 600 and Fig. 7 the difference between the structure 700 seemed to get rid of the possibility of all exchanges.But the high-rise module of respective decoder may have the structure (for example structure of describing among Fig. 8) of closely similar (even if inequality), and this explanation might be exchanged.What need explanation once more is, the implementation method of exchanging design is not insignificant (non-trivial), and its process need carries out some considerations.

Structure 600 among Fig. 6 is used to carry out one type TSNFC excitation quantification.The type is according to N _s(z) the short-term moulding (short-term shaping) of the whole quantizing noises of realization is shown in module 620; The type is according to N ₁(z) the long-term moulding (long-term shaping) of realization quantizing noise is shown in module 640.Provided LPC (short-term) fallout predictor in module 610, fundamental tone (for a long time) fallout predictor is arranged in module 630.The name that the mode of operation of structure 600 was announced on January 30th, 2007 is called in the United States Patent (USP) 7171355 of " Method andApparatus for One-Stage and Two-Stage Noise Feedback Coding of Speech andAudio Signals " to be introduced, and the present invention quotes wherein full content at this.For making this paper more brief, just no longer content has wherein been described here.

Structure 700 among Fig. 7 has been described an embodiment of the structure of carrying out CELP excitation quantification.Structure 700 is according to 1/W _s(z) realize the short-term moulding of quantizing noise, shown in module 720, but it does not carry out the long-term moulding of quantizing noise.In the CELP term, wave filter W _s(z) be commonly referred to " perceptual weighting filter ".The long-term shaping of quantizing noise is omitted, and this is because it does not quantize to carry out with the CELP of pumping signal usually.But this can be by adding long-term weighting filter and W _s(z) polyphone is realized.Being the short-term forecasting device in the module 710, is long-term predictor in the module 730.It should be noted that the fallout predictor in the

module

610 and 630 in the structure 600 in these fallout predictors difference corresponding diagram 6.The mode that structure 700 is carried out CELP excitation quantification is known for a person skilled in the art, is not described in detail here.

The task that excitation quantizes among Fig. 6 and Fig. 7 is to select clauses and subclauses from VQ decoding book (VQ decoding book 650 Fig. 6 and the VQ decoding book 770 among Fig. 7), but it also comprises the quantized value of selecting excitation gain, is expressed as " g ".For convenience of description, suppose this parameter in Fig. 6 structure 600 and the structure 700 among Fig. 7 in quantize respectively.In Fig. 6 and Fig. 7, from VQ decoding book, select vector to finish by the mean square deviation q (n) of minimum quantization error on input vector length usually.If use identical VQ decoding book in TSNFC and celp coder, and the module that is arranged on the outside the pale of civilization portion of energizing quantity is identical, and then two scramblers can provide compatible bit stream, although two excitation quantizing processs are different in itself.In addition, two bit streams also can be compatible mutually with TSNFC demoder or CELP demoder.

Although the present invention is described in conjunction with TSNFC that enumerates specially among Fig. 6 and Fig. 7 and CELP structure respectively, should be understood that all changes that it is applicable to TSNFC, NFC and CELP.As mentioned above, excitation quantizes even can use the additive method that is used for quantizing pumping signal to substitute.Above mentioned the concrete example that the open loop of fundamental tone prediction residual quantizes.

Fig. 8 is the structural representation that can be used for realizing common demoder 800 of the present invention.Yet the present invention is not limited only to the decoder architecture among Fig. 8, also can use other suitable structures.

As shown in Figure 8, decoder architecture 800 comprises bit demodulation multiplexer 802, be used to receive incoming bit stream, and from this incoming bit stream optionally the output encoder bit give pumping signal demoder 804, long-term forecasting parameter decoder 810 and short-term forecasting parameter decoder 812.Pumping signal demoder 804 is used for from bit demodulation multiplexer 802 received code bits, from wherein decoding pumping signal.Long-term forecasting parameter decoder 810 is used for from bit demodulation multiplexer 802 received code bits, from wherein decoding pitch period and fundamental tone beat.Short-term forecasting parameter decoder 812 is used for from bit demodulation multiplexer 802 received code bits, from wherein decoding the LPC coefficient.Long-term composite filter 806 is used to receive pumping signal corresponding to the fundamental tone composite filter, according to above-mentioned pitch period and fundamental tone beat this pumping signal is carried out filtering.Short-term composite filter 808 is used to receive from pumping signal after the filtering of long-term composite filter 808 corresponding to the LPC composite filter, and according to the LPC coefficient this signal is carried out filtering.Short-term composite filter 808 is output as output audio signal.

The a plurality of methods that realize according to the embodiment of the invention hereinafter will be described.These methods only are to be used for for example, are not to be to be used to limit the scope of the invention.

Fig. 9 is the process flow diagram 900 according to the method for transmit audio signals of one embodiment of the invention such as voice signal.Method in the process flow diagram 900 can be carried out by the system among Fig. 2 200.

As shown in Figure 9, the method in the process flow diagram 900 starts from step 902, and celp coder receives input audio signal such as input speech signal.In step 904, celp coder this input audio signal of encoding generates coded bit stream.Be similar to the celp coder 210 among Fig. 2, this celp coder is through custom-designed, and it is compatible mutually with the NFC demoder based on VQ.Therefore the bit stream that generates in the step 904 can be received by the NFC demoder based on VQ and decoding.In step 906, send out this coded bit stream from celp coder.In step 908, receive this coded bit stream based on the NFC demoder of VQ.Based on the NFC demoder of VQ can be for example based on the TSNFC demoder of VQ.In step 910,, generate output audio signal as the output voice signal based on this coded bit stream of NFC decoder decode of VQ.

Figure 10 is the process flow diagram 1000 according to the method for transmit audio signals of another embodiment of the present invention such as voice signal.Method in the process flow diagram 1000 can be carried out by the system among Fig. 4 400.

As shown in figure 10, the method in the process flow diagram 1000 starts from step 1002, based on NFC demoder reception input audio signal such as the input speech signal of VQ.Based on the NFC scrambler of VQ can be for example based on the TSNFC scrambler of VQ.In step 1004,, generate coded bit stream based on this input audio signal of NFC encoder encodes of VQ.Being similar to the NFC scrambler 410 based on VQ among Fig. 4, is through custom-designed based on the NFC scrambler of VQ, and it can be compatible mutually with the CELP demoder.Therefore, the bit stream that generates in step 1004 can be received the decode by the CELP demoder.

In step 1006, send out coded bit stream from NFC scrambler based on VQ.In step 1008, the CELP demoder receives this coded bit stream.In step 1010, this coded bit stream of CELP decoder decode generates output audio signal as the output voice signal.

According to principle of the present invention and above-described details, in one embodiment of the invention, can use single common decoder architecture to receive the decode by the sound signal of celp coder coding and oneself by sound signal based on the NFC encoder encodes of VQ.Such embodiment will be described in Figure 11.

Specifically, Figure 11 is the structural representation according to the system 1100 of one embodiment of the invention, celp coder 1110 bitstream encoded that both had been used to decode of single demoder 1130 wherein, and NFC scrambler 1120 bitstream encoded based on VQ are used to again decode.The operation of system 1100 will be described with reference to the process flow diagram among Figure 12 1200 among Figure 11.

As shown in figure 12, the method in the process flow diagram 1200 starts from step 1202, and celp coder 1110 receives the decode first input audio signal as first voice signal, generates first coded bit stream.In step 1204, celp coder 1110 mails to demoder 1130 with first coded bit stream.In step 1206, NFC scrambler 1120 reception and second input audio signals of encoding based on VQ as second voice signal, generate second coded bit stream.In step 1208, second coded bit stream is mail to demoder 1130 based on the NFC scrambler 1120 of VQ.

In step 1210, demoder 1130 receives the decode first coded bit stream, generates first output audio signal, as the first output voice signal.In step 1212, demoder 1130 also receives the decode second coded bit stream, generates second output audio signal as the second output voice signal.Therefore the demoder 1130 CELP coded bit stream of can either decoding can be decoded again based on the NFC coded bit stream of VQ.

Hereinafter a kind of general-purpose computing system will be described.The present invention can use the combination of hardware or software and hardware to realize.Therefore, the present invention can realize in computer system or other disposal system environment.Figure 13 has described an example of this computer system 1300.In the present invention, all processing modules among Fig. 2, Fig. 4-Figure 12 can both be carried out on one or more different computer systems 1300 with step, to realize the whole bag of tricks of the present invention.Computer system 1300 comprises one or more processors such as processor 1304.Processor 1304 can make special use or nextport universal digital signal processor NextPort.Processor 1304 is connected to communication construction 1302 (for example bus or network).Various softwares are realized and will be provided description in conjunction with this example computer system.Reading after these describe, those skilled in the art should grasp and how to use other computer systems and/or computer architecture to realize the present invention.

Computer system 1300 also can comprise primary memory 1306, and preferably random access storage device (RAM) also can comprise supplementary storage 1320.Supplementary storage 1320 can comprise for example hard disk drive 1322 and/or removable memory driver 1324, as disc driver, tape drive, CD drive or similar storer.Removable memory driver 1324 is read and write removable memory module 1328 according to known mode.Removable memory module 1328 can be disk, tape, CD or similar storage unit, and it is read and write by removable memory driver 1324.Should be understood that removable memory module 1328 comprises the computing machine storage medium, stores computer software and/or data on it.

In another embodiment of the present invention, supplementary storage 1320 can comprise other similar devices, allows computer program or other instruction load in computer system 1300.These devices can comprise for example removable memory module 1330 and interface 1326.These devices also can comprise program tape and tape interface (as the similar device in the video game device), removable storage chip (as EPROM or PROM) and relevant slot, be used for software and data are mail to from removable memory module 1330 other removable memory modules 1330 and the interface 1326 of computer system 1300.

Computer system 1300 also can include communication interface 1340.Communication interface 1340 can transmit software and data between computer system 1300 and external unit.Communication interface 1340 can be modulator-demodular unit, network interface (as Ethernet card), communication port, PCMCIA groove card etc.Software and data are to transmit by communication interface 1340 with the form of signal, and these signals can be electric signal, electromagnetic signal, light signal or other signals that can be received by communication interface 1340.These signals mail to communication interface 1340 by communication path 1342.Communication path 1342 carrying signals, it can use electric wire or cable, telephone wire, cellular phone link, RF link and other communication channels to realize.

In this article, term " computer program medium " and " computing machine medium " be used in reference to generation as removable memory module 1328 and 1330, be installed in the hard disk on the hard disk drive 1322, and by the medium of signal of communication interface 1340 receptions and so on.These computer programs are used for providing software to computer system 1300.

Computer program (being also referred to as computer control logic) is stored in primary memory 1306 and/or the supplementary storage 1320.Computer program also can receive by communication interface 1340.After these computer programs are performed, can make computer system 1300 can realize the method that this paper tells about.Specifically, these computer programs can make processor 1300 realize the process that the present invention told about, any means of telling about as mentioned after being performed.Therefore, this computer program is represented the controller of computer system 1300.Realize that when using software when of the present invention, this software can be stored in the computer program, uses removable memory driver 1324, interface 1326 or communication interface 1340 to be loaded in the computer system 1300.

In another embodiment, the present invention also can mainly be realized by hardware, for example uses nextport hardware component NextPort such as special IC (ASIC) and gate array.The realization that is used to carry out the hardware state machine of function described herein should be conspicuous for a person skilled in the art.

The present invention is by having showed specific function of the present invention and relation thereof by method step.The scope of described method step and order are to define arbitrarily for convenience of description.As long as can carry out specific function and order, also can use other boundary and order.Therefore any boundary or order described or choosing fall into scope and spirit essence of the present invention.The present invention also is described some important function by functional module.The boundary of described functional module and the relation of various functional modules are to define arbitrarily for convenience of description.As long as can carry out specific function, also can use other boundary or relation.Therefore described other boundary or relation also falls into scope and spirit essence of the present invention.Those of ordinary skills also as can be known, the functional module among the application and other displaying property module and assembly can be embodied as the processor and the aforesaid combination in any of discrete component, special IC, the appropriate software of execution.

In addition, although more than be the description of the present invention being carried out by some embodiment, those skilled in the art know that the present invention is not limited to these embodiment, under the situation that does not break away from the spirit and scope of the present invention, can carry out various changes or equivalence replacement to these features and embodiment.

Claims

1. a method that is used for decoded audio signal is characterized in that, described method comprises:

The received code bit stream, wherein said coded bit stream representative is via the input audio signal of celp coder coding;

Use generates output audio signal based on the NFC demoder of the VQ described coded bit stream of decoding.

2. method according to claim 1 is characterized in that described input audio signal comprises input speech signal, and described output audio signal comprises the output voice signal.

3. method according to claim 1 is characterized in that, uses based on the decode step of described coded bit stream of the NFC demoder of VQ to comprise: use based on the two stage NFC demoders of the VQ described coded bit stream of decoding.

4. method according to claim 1 is characterized in that, described method also comprises:

Receive described input audio signal;

Use the celp coder described input audio signal of encoding, generate described coded bit stream.

5. a system that is used for transmit audio signals is characterized in that, described system comprises:

Celp coder is used to the input audio signal of encoding, and generates coded bit stream;

Based on the NFC demoder of VQ, the described coded bit stream that is used to decode generates output audio signal.

6. system according to claim 5 is characterized in that described input audio signal comprises input speech signal, and described output audio signal comprises the output voice signal.

7. a method that is used for decoded audio signal is characterized in that, described method comprises:

The received code bit stream, the representative of wherein said coded bit stream is via the input audio signal based on the NFC encoder encodes of VQ;

Use the described coded bit stream of CELP decoder decode, generate output audio signal.

8. a system that is used for transmit audio signals is characterized in that, described system comprises:

NFC scrambler based on VQ is used to the input audio signal of encoding, and generates coded bit stream;

The CELP demoder, the described coded bit stream that is used to decode generates output audio signal.

9. a method that is used for decoded audio signal is characterized in that, described method comprises:

Receive first coded bit stream, wherein said first coded bit stream representative is via first input audio signal of celp coder coding;

Described first coded bit stream of decoding generates first output audio signal in demoder;

Receive second coded bit stream, the representative of wherein said second coded bit stream is via second input audio signal based on the NFC encoder encodes of VQ;

Described second coded bit stream of decoding generates second output audio signal in described demoder.

10. a system that is used for transmit audio signals is characterized in that, described system comprises:

Celp coder, first input audio signal that is used to encode generates first coded bit stream;

Based on the NFC scrambler of VQ, second input audio signal that is used to encode generates second coded bit stream;

Demoder, described first coded bit stream that is used to decode generates first output audio signal, and described second coded bit stream of decoding, and generates second output audio signal.