Hierarchical audio coding, decoding method and system

• US 8,694,325 B2
• Filed: 10/26/2010
• Issued: 04/08/2014
• Est. Priority Date: 11/27/2009
• Status: Active Grant

• Current Assignee: ZTE Corporation
• Original Assignee: ZTE Corporation
• Inventors: Lin, Zhibin, Deng, Zheng, Yuan, Hao, Lu, Jing, Qiu, Xiaojun, Li, Jiali, Chen, Guoming, Peng, Ke, Liu, Kaiwen
• Primary Examiner(s): GUERRA-ERAZO, EDGAR X
• Application Number:
  US13/505,064
• Publication Number:
  US 20120226505A1
• Time in Patent Office: 1,260 Days
• Field of Search: 704/203, 704/206, 704/229, 704500-504
• US Class Current: 704/500
• CPC Class Codes: G10L 19/002 Dynamic bit allocation for perceptual audio coders G10L19/032
• CPC Class Codes: G10L 19/0204 using subband decomposition
• CPC Class Codes: G10L 19/0212 using orthogonal transformation
• CPC Class Codes: G10L 19/028 Noise substitution, i.e. substituting non-tonal spectral components by noisy source comfort noise for discontinuous speech transmission G10L19/012
• CPC Class Codes: G10L 19/038 Vector quantisation, e.g. TwinVQ audio
• CPC Class Codes: G10L 19/24 Variable rate codecs, e.g. for generating different qualities using a scalable representation such as hierarchical encoding or layered encoding
• CPC Class Codes: H03M 7/3082 Vector coding for television signals, see H04N19/94

Abstract

A hierarchical audio coding, decoding method and system are provided. The method includes dividing frequency domain coefficients of an audio signal after MDCT into a plurality of coding sub-bands, quantizing and coding amplitude envelope values of coding sub-bands; allocating bits to each coding sub-band of the core layer, quantizing and coding core layer frequency domain coefficients to obtain coded bits of core layer frequency domain coefficients; calculating the amplitude envelope value of each coding sub-band of the core layer residual signal; allocating bits to each coding sub-band of the extended layer, quantizing and coding the extended layer coding signal to obtain coded bits of the extended layer coding signal; multiplexing and packing amplitude value envelope coded bits of each coding sub-band composed by core layer and extended layer frequency domain coefficients, core layer frequency coefficients coded bits, and extended layer coding signal coded bits, then transmitting to the decoding end.

Claims

1. A hierarchical audio coding method, comprising following steps:

dividing frequency domain coefficients of an audio signal for which Modified Discrete Cosine Transform (MDCT) has been performed into core layer frequency domain coefficients and extended layer frequency domain coefficients, and dividing the core layer frequency domain coefficients and extended layer frequency domain coefficients into a plurality of coding sub-bands, and quantizing and coding amplitude envelope values of the coding sub-bands;

allocating bits to each coding sub-band of a core layer according to the amplitude envelope value of each coding sub-band of the core layer, and then quantizing and coding the core layer frequency domain coefficients to obtain coded bits of the core layer frequency domain coefficients;

calculating an amplitude envelope value of each coding sub-band of a core layer residual signal according to the amplitude envelope value and a bit allocation number of each coding sub-band of the core layer;

allocating bits to each coding sub-band of an extended layer according to an extended layer coding signal composed of the core layer residual signal and the extended layer frequency domain coefficients and the amplitude envelope value of each coding sub-band composed of the extended layer coding signal, and then quantizing and coding the extended layer coding signal to obtain coded bits of the extended layer coding signal; and

multiplexing and packing amplitude value envelope coded bits of each coding sub-band composed of the core layer frequency domain coefficients and the extended layer frequency domain coefficients, coded bits of the core layer frequency domain coefficients, and coded bits of the extended layer coding signal, and then transmitting to a decoding end.

2. The method according to claim 1, wherein, in said step of dividing frequency domain coefficients of an audio signal for which Modified Discrete Cosine Transform (MDCT) has been performed into core layer frequency domain coefficients and extended layer frequency domain coefficients, and dividing the core layer frequency domain coefficients and extended layer frequency domain coefficients into a plurality of coding sub-bands,

the frequency domain coefficients after MDCT has been performed are divided into a plurality of uniformly-spaced coding sub-bands, or are divided into a plurality of non-uniform coding sub-bands according to auditory perceptive characteristics;

low-frequency coding sub-bands are grouped as core layer coding sub-bands and high-frequency coding sub-bands are grouped as extended layer coding sub-bands.

3. The method according to claim 1, wherein,

in said step of quantizing and coding the core layer frequency domain coefficients and the extended layer coding signal,

to-be-quantized vectors of a coding sub-band whose bit allocation number is less than a classification threshold are quantized and coded using pyramid lattice vector quantization, and to-be-quantized vectors of a coding sub-band whose bit allocation number is greater than the classification threshold are quantized and coded using sphere lattice vector quantization;

the bit allocation number is a number of bits allocated to a single frequency domain coefficient in one coding sub-band.

4. The method according to claim 1, wherein, in said step of calculating an amplitude envelope value of each coding sub-band of a core layer residual signal according to the amplitude envelope value and a bit allocation number of each coding sub-band of the core layer,

the amplitude envelope value of each coding sub-band of the core layer residual signal is calculated by a following way:

inversely quantizing the quantized value of the core layer frequency domain coefficient, and implementing a difference calculation with respect to the core layer frequency domain coefficient to obtain the core layer residual signal;

dividing the core layer residual signal into coding sub-bands by an identical way as for the core layer frequency domain coefficient, and calculating the amplitude envelope value of each coding sub-band of the core layer residual signal according to the amplitude envelope value and the bit allocation number of each coding sub-band of the core layer.

5. The method according to claim 4, wherein, the step of calculating the amplitude envelope value of each coding sub-band of the core layer residual signal comprises:

calculating an amplitude envelope quantization exponent of each coding sub-band of the core layer, and implementing a difference calculation with respect to a quantization exponent modification value of a corresponding coding sub-band to obtain the amplitude envelope quantization exponent of each coding sub-band of the core layer residual signal;

the quantization exponent modification value of each coding sub-band being greater than or equal to 0, and not decreasing when the bit allocation number of the corresponding coding sub-band of the core layer increases; and

when the bit allocation number of a coding sub-band of the core layer is 0, the quantization exponent modification value being 0, and when the bit allocation number of the coding sub-band of the core layer is a defined maximum bit allocation number, the amplitude envelope value of the core layer residual signal in the coding sub-band being 0.

6. The method according to claim 1, wherein, said step of obtaining the coded bits of the core layer frequency domain coefficients comprises:

calculating the amplitude envelope quantization exponent of each coding sub-band of the core layer, and allocating bits to each coding sub-band of the core layer; and

normalizing the frequency domain coefficients of each coding sub-band according to an quantization amplitude envelope value of each coding sub-band of the core layer reconstructed according to the amplitude envelope quantization exponent of each coding sub-band of the core layer, and quantizing and coding according to the bit allocation number of each coding sub-band of the core layer using pyramid lattice vector quantization and sphere lattice vector quantization respectively to obtain the coded bits of the core layer frequency domain coefficients;

wherein, said step of obtaining the coded bits of the extended layer coding signal comprises;

constituting the extended layer coding signal with the core layer residual signal and the extended layer frequency domain coefficients;

allocating bits to each coding sub-band of the extended layer according to the amplitude envelope quantization exponent of each coding sub-band of the extended layer coding signal; and

normalizing the coding signal of each coding sub-band according to an quantization amplitude envelope value of each coding sub-band of the extended layer reconstructed according to the amplitude envelope quantization exponent of each coding sub-band of the extended layer, and quantizing and coding according to the bit allocation number of each coding sub-band of the extended layer using pyramid lattice vector quantization and sphere lattice vector quantization method respectively to obtain the coded bits of the extended layer coding signal.

7. The method according to claim 1, wherein, in said step of multiplexing and packing amplitude value envelope coded bits of each coding sub-band composed of the core layer frequency domain coefficients and the extended layer frequency domain coefficients, coded bits of the core layer frequency domain coefficients, and coded bits of the extended layer coding signal, and then transmitting to the decoding end,

multiplexing and packing are implemented according to a following code stream format:

writing side information bits of the core layer after a frame header of a code stream, writing the amplitude value envelope coded bits of the coding sub-band of the core layer into a bit stream multiplexer MUX, and then writing the coded bits of the core layer frequency domain coefficients into the MUX; and

writing side information bits of the extended layer into the MUX, writing the amplitude value envelope coded bits of the coding sub-band of the extended layer frequency domain coefficients into the MUX, and then writing the coded bits of the extended layer coding signal into the MUX;

the bits whose number meets a required code rate are sent to the decoding end according to the required code rate.

8. The method according to claim 7, wherein,

the side information of the core layer includes Huffman coding flag bits of the coding sub-band amplitude envelope composed of MDCT coeffcients, Huffman coding flag bits of the core layer frequency domain coefficients, and bits of a number of iterations;

the side information of the extended layer includes Huffman coding flag bits of the extended layer coding signal and bits of a number of iterations.

9. The method according to claim 7, wherein, said step of sending the bits whose number meets the required code rate to the decoding end comprises:

coding the core layer and the extended layer within a range of a whole frequency band in which coding is needed according to a number of bits for a defined maximum coding rate such that the maximum coding rate is reached within a range of the coding frequency band;

after the coded bits of the core layer, writing the coded bits of the extended layer coding signal into the bit stream multiplexer according to an initial value of importance of each coding sub-band in a descending order; and

sending the bits whose number meets the required code rate to the decoding end based on a principle that the coded bits of the coding sub-band with a smaller initial value of importance are firstly rejected;

wherein, the code rate corresponding to the core layer is marked as L0 layer, each code rate corresponding to the extended layer is marked as L1_1 layer, L1_2 layer, . . . , up to L1_K layer according to a number of rejected bits of the extended layer respectively, wherein, the code rate of the L1 _K layer is precisely the maximum coding rate.

10. The method according to claim 1, wherein,

after said step of dividing frequency domain coefficients of an audio signal for which Modified Discrete Cosine Transform (MDCT) has been performed into core layer frequency domain coefficients and extended layer frequency domain coefficients, the method further comprises:

dividing the frequency domain coefficients of the extended layer, in an ascending order of frequencies, into M parts, which are frequency domain coefficients of extended layers from extended layer 1 to extended layer M respectively, wherein, the frequency for the frequency domain coefficient in an extended layer 1 is lowest and the frequency for the frequency domain coefficient in an extended layer M is highest;

after said step of quantizing and coding the core layer frequency domain coefficients, the method further comprises;

inversely quantizing the quantized value of the core layer frequency domain coefficient, and implementing a difference calculation with respect to the core layer frequency domain coefficient to obtain the core layer residual signal;

in said step of calculating an amplitude envelope value of each coding sub-band of a core layer residual signal according to the amplitude envelope value and a bit allocation number of each coding sub-band of the core layer, the core layer residual signal is divided into coding sub-bands by a same way as for the core layer, and the amplitude envelope value of each coding sub-band of the core layer residual signal is calculated according to the amplitude envelope value and the bit allocation number of each coding sub-band of the core layer;

in said step of allocating the bits to each coding sub-band of an extended layer according to an extended layer coding signal composed of the core layer residual signal and the extended layer frequency domain coefficients and the amplitude envelope value of each coding sub-band composed of the extended layer coding signal, and then quantizing and coding the extended layer coding signal to obtain coded bits of the extended layer coding signal,

the coding signal of the extended layer 1 is constituted with the core layer residual signal and the frequency domain coefficient of the extended layer 1, the amplitude envelope quantization exponent of each coding sub-band is calculated according to the coding signal of extended layer 1, and bits are allocated to each coding sub-band of the extended layer 1;

the coding signal of the extended layer 1 is quantized and coded according to the bit allocation number of each coding sub-band of the extended layer 1;

the quantized value of the coding signal of an extended layer i is inversely quantized, and a difference calculation is implemented with respect to the coding signal of the extended layer i to obtain a residual signal of the extended layer i;

the residual signal of the extended layer i is divided into coding sub-bands by a same way as for the extended layer i, and the amplitude envelope value of each coding sub-band of the residual signal of the extended layer i is calculated according to the amplitude envelope value and the bit allocation number of each coding sub-band of the extended layer i;

the coding signal of an extended layer i+1 is constituted with the residual signal of the extended layer i and the frequency domain coefficient of the extended layer i+1, and bits are allocated to each coding sub-band of the extended layer i+1 according to the amplitude envelope quantization exponent of each coding sub-band of the coding signal of the extended layer i+1; and

the coding signal of each coding sub-band of the extended layer i+1 is normalized, vector quantized and coded according to a quantization amplitude envelope value and a number of coded bits of each coding sub-band of the extended layer i+1 reconstructed according to the amplitude envelope quantization exponent of each coding sub-band of the extended layer i+1 to obtain the coded bits of the coding signal of the extended layer i+1;

said i=1, . . . . N−

1, and N is a positive integer greater than 2.

11. The method according to claim 1, wherein, when multiplexing and packing amplitude value envelope coded bits of each coding sub-band composed of the core layer frequency domain coefficients and the extended layer frequency domain coefficients, the coded bits of the core layer frequency domain coefficients, and the coded bits of the extended layer coding signal,

multiplexing and packing are implemented according to a following code stream format:

after the coded bits of the core layer, according to a number of the extended layer, writing the bits of an extended layer with a smaller serial number into a bit stream multiplexer MUX, and then writing the bits of an extended layer with a greater serial number into the bit stream multiplexer MUX;

in an identical extended layer, firstly writing side information bits of the extended layer into the MUX, writing the amplitude value envelope coded bits of the coding sub-band of the extended layer frequency domain coefficients into the MUX, and then writing the coded bits of the extended layer coding signal into the MUX;

sending the bits whose number meets a required code rate to the decoding end according to the required code rate.

12. A hierarchical audio decoding method, comprising following steps:

de-multiplexing a bit stream sent by a coding end, and decoding amplitude envelope coded bits of core layer frequency domain coefficients to obtain an amplitude envelope value of each coding sub-band of a core layer;

allocating bits to each sub-band of the core layer according to the amplitude envelope value of each coding sub-band of the core layer, and calculating an amplitude envelope value of a core layer residual signal;

decoding amplitude envelope coded bits of extended layer frequency domain coefficients to obtain an amplitude envelope value of the extended layer frequency domain coefficients;

allocating bits to each coding sub-band of an extended layer according to the amplitude envelope value of the core layer residual signal and amplitude envelope value of the extended layer frequency domain coefficients; and

respectively decoding coded bits of the core layer frequency domain coefficients and coded bits of an extended layer coding signal according to a bit allocation number of the core layer and a bit allocation number of the extended layer to obtain the core layer frequency domain coefficients and the extended layer coding signal, rearranging the extended layer coding signal according to an order of frequency bands and adding with the core layer frequency domain coefficients to obtain frequency domain coefficients of a whole bandwidth, performing inverse Modified Discrete Cosine Transform (IMDCT) on the frequency domain coefficients of the whole bandwidth to obtain an output audio signal.

13. The method according to claim 12, wherein, the step of calculating the amplitude envelope value of the core layer residual signal comprises:

calculating an amplitude envelope quantization exponent of each coding sub-band of the core layer, and implementing a difference calculation with respect to a quantization exponent modification value of a corresponding coding sub-band to obtain the amplitude envelope quantization exponent of each coding sub-band of the core layer residual signal;

calculating the amplitude envelope value of each sub-band of the residual signal of an extended layer i according to the amplitude envelope and the bit allocation number of each coding sub-band of the extended layer i;

the quantization exponent modification value of each coding sub-band being greater than or equal to 0, and not decreasing when the bit allocation number of the corresponding coding sub-band of the core layer increases; and

when the bit allocation number of a coding sub-band of the core layer is 0, the quantization exponent modification value being 0, and when the bit allocation number of the coding sub-band is a defined maximum bit allocation number, the amplitude envelope value of the core layer residual signal in the coding sub-band being 0.

14. The method according to claim 12, wherein, said step of obtaining the core layer frequency domain coefficients and the extended layer coding signal comprises:

decoding, inversely quantizing and inversely normalizing the coded bits of the core layer frequency domain coefficients according to the bit allocation number, the quantization amplitude envelope value and side information of each coding sub-band of the core layer to obtain the core layer frequency domain coefficients; and

decoding, inversely quantizing and inversely normalizing the coded bits of the extended layer coding signal according to the bit allocation number of each coding sub-band of the extended layer, the quantization amplitude envelope value and side information of the extended layer coding signal to obtain the extended layer coding signal.

15. The method according to claim 12, wherein, in said step of decoding the coded bits of the extended layer coding signal,

the coded bits of the extended layer coding signal are decoded in a following order:

in different extended layers, an order for decoding various layers is;

firstly decoding the coded bits of each coding sub-band and the amplitude envelope of a lower extended layer with a smaller extended layer serial number, and then decoding the coded bits of each coding sub-band and the amplitude envelope of a higher extended layer with a greater extended layer serial number, a number of decoded bits is calculated during the decoding, and decoding is terminated when the number of the decoded bits meets requirements on a total number of bits;

in an identical extended layer, an order for decoding various sub-bands is determined by an initial value of importance of each coding sub-band, a coding sub-band with a greater importance is preferentially decoded, and if there are two coding sub-bands with the same importance, then a lower-frequency coding sub-band is preferentially decoded, a number of decoded bits is calculated during the decoding, and decoding is terminated when the number of the decoded bits meets requirements on the total number of bits.

16. A hierarchical audio coding system, comprising:

a Modified Discrete Cosine Transform (MDCT) unit, an amplitude envelope calculating unit, an amplitude envelope quantizing and coding unit, a core layer bit allocating unit, a core layer frequency domain coefficient vector quantizing and coding unit, and a bit stream multiplexer;

the system further comprises;

an extended layer coding signal generating unit, a residual signal amplitude envelope generating unit, an extended layer bit allocating unit, and an extended layer coding signal vector quantizing and coding unit;

wherein;

the MDCT unit is configured to perform Modified Discrete Cosine Transform on an input audio signal to generate core layer frequency domain coefficients and extended layer frequency domain coefficients;

the amplitude envelope calculating unit is configured to group to-be-coded parts in the core layer frequency domain coefficients and extended layer frequency domain coefficients as core layer frequency domain coefficient coding sub-bands and extended layer frequency domain coefficient coding sub-bands, and calculate an amplitude envelope value of each coding sub-band;

the amplitude envelope quantizing and coding unit is configured to quantize and code the amplitude envelope values of the core layer frequency domain coefficient coding sub-bands and extended layer frequency domain coefficient coding sub-bands to generate an amplitude envelope quantization exponent and coded bits of amplitude envelope of each coding sub-band of the core layer frequency domain coefficients and the extended layer frequency domain coefficients;

the core layer bit allocating unit is configured to allocate bits to the core layer to obtain a bit allocation number of each coding sub-band of the core layer;

the core layer frequency domain coefficient vector quantizing and coding unit is configured to normalize, vector quantize and code the frequency domain coefficients of each coding sub-band of the core layer using a quantization amplitude envelope value and bit allocation number of each coding sub-band of the core layer reconstructed according to the amplitude envelope quantization exponent of each coding sub-band of the core layer to obtain coded bits of the core layer frequency domain coefficients;

the extended layer coding signal generating unit is configured to generate a residual signal, and obtain an extended layer coding signal composed of the residual signal and extended layer frequency domain coefficients;

the residual signal amplitude envelope generating unit is configured to obtain an amplitude envelope quantization exponent of each coding sub-band of the residual signal according to an amplitude envelope quantization exponent of each coding sub-band of the core layer and a bit allocation number of a corresponding coding sub-band;

the extended layer bit allocating unit is configured to allocate bits to an extended layer according to the amplitude envelope quantization exponent of each coding sub-band of the residual signal and the amplitude envelope quantization exponent of each coding sub-band of the extended layer frequency domain coefficients to obtain a bit allocation number of each coding sub-band of the extended layer coding signal;

the extended layer coding signal vector quantizing and coding unit is configured to normalize, vector quantize and code the coding signal of each coding sub-band of the extended layer using a quantization amplitude envelope value and bit allocation number of each coding sub-band of the extended layer reconstructed according to the amplitude envelope quantization exponent of each coding sub-band of the extended layer to obtain coded bits of the extended layer coding signal;

the bit stream multiplexer is configured to pack side information bits of the core layer, the coded bits of the amplitude envelope of each coding sub-band of the core layer, the coded bits of the core layer frequency domain coefficients, side information bits of the extended layer, the coded bits of each amplitude envelope of the extended layer frequency domain coefficients, and the coded bits of the extended layer coding signal.

17. The system according to claim 16, wherein,

the extended layer coding signal generating unit further comprises a residual signal generating unit and an extended layer coding signal synthesizing unit;

the residual signal generating unit is configured to inversely quantize quantized values of the core layer frequency domain coefficients, and implement a difference calculation with respect to the core layer frequency domain coefficient to obtain a core layer residual signal;

the extended layer coding signal synthesizing unit is configured to synthesize the core layer residual signal and the extended layer frequency domain coefficients in an order of frequency bands to obtain the extended layer coding signal.

18. The system according to claim 16, wherein,

the residual signal amplitude envelope generating unit further comprises a quantization exponent modification value obtaining unit and a residual signal amplitude envelope quantization exponent calculating unit;

the quantization exponent modification value obtaining unit is configured to calculate a quantization exponent modification value of each coding sub-band of the residual signal according to the bit allocation number of each coding sub-band of the core layer, wherein, the quantization exponent modification value of each coding sub-band is greater than or equal to 0, and does not decrease when the bit allocation number of the corresponding coding sub-band of the core layer increases;

when the bit allocation number of a coding sub-band of the core layer is 0, the quantization exponent modification value is 0, and when the bit allocation number of the coding sub-band is a defined maximum bit allocation number, the amplitude envelope value of the residual signal in the coding sub-band is 0;

the residual signal amplitude envelope quantization exponent calculating unit is configured to implement a difference calculation between an amplitude envelope quantization exponent of each coding sub-band of the core layer and a quantization exponent modification value of a corresponding coding sub-band to obtain the amplitude envelope quantization exponent of each coding sub-band of the core layer residual signal.

19. The system according to claim 16, wherein,

the bit stream multiplexer is configured to write the coded bits of the extended layer coding signal into a code stream according to an initial value of importance of each coding sub-band in a descending order, and for the coding sub-bands of same importance, preferably write the coded bits of a lower-frequency coding sub-band into the code stream.

20. A hierarchical audio decoding system, comprising:

a bit stream de-multiplexer, a core layer amplitude envelope decoding unit, a core layer bit allocating unit, and a core layer decoding and inverse quantizing unit;

the system further comprises;

an extended layer amplitude envelope decoding unit, a core layer residual signal amplitude envelope generating unit, an extended layer bit allocating unit, an extended layer coding signal decoding and inverse quantizing unit, a frequency domain coefficient generating unit, a noise filling unit and an inverse Modified Discrete Cosine Transform (IMDCT) unit;

wherein;

the core layer amplitude envelope decoding unit is configured to decode amplitude envelope coded bits of coding sub-bands of a core layer output by the bit stream de-multiplexer to obtain an amplitude envelope quantization exponent of each coding sub-band of the core layer;

the core layer bit allocating unit is configured to allocate bits to the core layer to obtain a bit allocation number of each coding sub-band of the core layer;

the core layer decoding and inverse quantizing unit is configured to calculate a quantization amplitude envelope value of each coding sub-band of the core layer according to the amplitude envelope quantization exponent of each coding sub-band of the core layer, and decode, inversely quantize and inversely normalize the coded bits of the core layer frequency domain coefficients output by the bit stream de-multiplexer using the bit allocation number and the quantization amplitude envelope value of each coding sub-band of the core layer to obtain the core layer frequency domain coefficients;

the extended layer amplitude envelope decoding unit is configured to decode the amplitude envelope coded bits of the core layer frequency domain coefficients output by the bit stream de-multiplexer to obtain an amplitude envelope quantization exponent of each coding sub-band of the core layer frequency domain coefficients;

the core layer residual signal amplitude envelope generating unit is configured to calculate an amplitude envelope quantization exponent of each coding sub-band of the residual signal according to the amplitude envelope quantization exponent of each coding sub-band of the core layer and the bit allocation number of a corresponding coding sub-band;

the extended layer bit allocating unit is configured to allocate bits to an extended layer according to the amplitude envelope quantization exponent of each coding sub-band of the residual signal and the amplitude envelope quantization exponent of each coding sub-band of the extended layer frequency domain coefficients to obtain a bit allocation number of each coding sub-band of the extended layer coding signal;

the extended layer coding signal decoding and inverse quantizing unit is configured to calculate a quantization amplitude envelope value of each coding sub-band of the extended layer using the amplitude envelope quantization exponent of each coding sub-band of the extended layer, and decode, inversely quantize and inversely normalize the coded bits of the extended layer coding signal output by the bit stream de-multiplexer using the bit allocation number and the quantization amplitude envelope value of each coding sub-band of the extended layer to obtain the extended layer coding signal;

the frequency domain coefficient generating unit is configured to reorder the extended layer coding signal output by the extended layer coding signal decoding and inverse quantizing unit according to an order of the frequency bands, and then perform sum calculation with the core layer frequency domain coefficients output by the core layer coding signal decoding and inverse quantizing unit to obtain frequency domain coefficient output values;

the noise filling unit is configured to perform noise filling on the coding sub-bands to which no coded bit is allocated during a process of coding;

the IMDCT unit is configured to perform IMDCT on all frequency domain coefficients for which noise filling has been performed to obtain an output audio signal.

21. The system according to claim 20, wherein,

the core layer residual signal amplitude envelope generating unit further comprises a quantization exponent modification value obtaining unit and a residual signal amplitude envelope quantization exponent calculating unit;

the quantization exponent modification value obtaining unit is configured to calculate the quantization exponent modification value of each coding sub-band of the residual signal according to the bit allocation number of each coding sub-band of the core layer, and make adjustment according to an absolute value of the frequency domain coefficient of a corresponding coding sub-band, wherein, the quantization exponent modification value of each coding sub-band is greater than or equal to 0, and does not decrease when the bit allocation number of the corresponding coding sub-band of the core layer increases, and if the bit allocation number of a coding sub-band of the core layer is 0, the quantization exponent modification value is 0, and if the bit allocation number of the coding sub-band of the core layer is a defined maximum bit allocation number, the amplitude envelope value of the residual signal in the coding sub-band is 0;

the residual signal amplitude envelope quantization exponent calculating unit is configured to perform a difference calculation between the amplitude envelope quantization exponent of each coding sub-band and the quantization exponent modification value of a corresponding coding sub-band to obtain the amplitude envelope quantization exponent of each coding sub-band of the core layer residual signal.

22. The system according to claim 20, wherein,

the extended layer coding signal decoding and inverse quantizing unit is configured such that an order for decoding various sub-bands of the extended layer coding signal is determined by an initial value of importance of each coding sub-band, a coding sub-band with a greater importance is preferentially decoded, and if there are two coding sub-bands with same importance, then a lower-frequency coding sub-band is preferentially decoded, a number of decoded bits is calculated during the decoding, and decoding is terminated when the number of the decoded bits meets requirements on a total number of bits.