KR20080034817A - Apparatus and method for encoding and decoding signal - Google Patents

Abstract

An apparatus and a method for encoding and decoding signals are provided to encode and decode the signals with different characteristics by encoding low frequency band signals through different encoding methods and encoding high frequency band signals through band expansion encoding methods. An apparatus for encoding and decoding signals comprises a bit unpacking part, first and second decoding parts(111,112), a band expansion decoding part and a mixing part. The bit unpacking part withdraws low frequency band signals, middle/high frequency band signals, and band expansion information. The first decoding part decodes the encoded low frequency band signal using a first decoding method. The second decoding part decodes the encoded middle/high frequency band signal using a second decoding method. The band expansion decoding part recovers the high frequency band signals from the decoded low frequency and middle/high frequency band signals. A signal dividing part(100) divides the inputted signals into low, middle/high and high frequency bands.

Description

Apparatus and method for encoding and decoding signal

1 is a block diagram showing a schematic configuration of an encoding apparatus according to the present invention.

2 is a block diagram showing an embodiment of a configuration of an encoding apparatus according to the present invention.

FIG. 3 is a block diagram illustrating a first embodiment of the configuration of the speech coder illustrated in FIG. 2.

FIG. 4 is a block diagram illustrating a second embodiment of the configuration of the speech encoder illustrated in FIG. 2.

FIG. 5 is a block diagram illustrating a third embodiment of the configuration of the speech encoder illustrated in FIG. 2.

FIG. 6 is a block diagram illustrating a first embodiment of the configuration of the audio encoder illustrated in FIG. 2.

FIG. 7 is a block diagram illustrating a second embodiment of the configuration of the audio encoder illustrated in FIG. 2.

FIG. 8 is a block diagram illustrating an embodiment of a configuration of a band extension encoder illustrated in FIG. 2.

9 is a block diagram illustrating an embodiment of a configuration of a decoding apparatus according to the present invention.

The present invention relates to an apparatus and method for encoding / decoding a signal, and more particularly, to an efficient encoding / decoding apparatus and method for encoding / decoding at an optimal bit rate according to characteristics of a signal.

Conventional audio coders provide high quality audio signals at high bit rates above 48 kbps, but are inefficient for processing speech signals. Conventional speech coders can efficiently encode speech signals at low bit rates below 12 kbps, but encode various audio signals. Lack in.

An object of the present invention is to provide an encoding / decoding apparatus and method for encoding signals having different characteristics, such as voice signals and audio signals, at an optimal bit rate.

According to an aspect of the present invention, there is provided a decoding method comprising: extracting an encoded high frequency band signal and band extension information; Decoding the encoded low frequency band signal using a first decoding method; Decoding the encoded high frequency band signal using a second decoding method; Restoring a high frequency band signal from the decoded low frequency band signal and the high frequency band signal by using the extracted band extension information; And synthesizing the decoded low frequency band signal and the high frequency band signal and the reconstructed high frequency band signal.

According to an aspect of the present invention, there is provided an encoding method, comprising: dividing an input signal into a low frequency band, a medium frequency band, and a high frequency band; Encoding the low frequency band signal using a first encoding scheme performed on a time domain; Encoding the high frequency band signal using a second coding scheme performed on a frequency domain; Generating band extension information for reconstructing the signal of the high frequency band by using the low frequency band signal and the high frequency band signal; And generating a bitstream including the encoded low frequency band signal and the high frequency band signal and the band extension information.

According to an aspect of the present invention, there is provided a decoding apparatus comprising: a bit unpacking unit configured to extract an encoded low frequency band signal, an encoded medium frequency band signal, and band extension information from an input bitstream; A first decoder which decodes the encoded low frequency band signal using a first decoding method performed in a time domain; A second decoder which decodes the encoded mid-frequency band signal using a second decoding method performed on a frequency domain; A band extension decoding unit which restores a high frequency band signal from the decoded low frequency band signal and the high frequency band signal by using the extracted band extension information; And a synthesizer for synthesizing the decoded low frequency band signal and the high frequency band signal and the restored high frequency band signal.

According to an aspect of the present invention, there is provided an encoding apparatus, including: a signal splitter configured to divide an input signal into a low frequency band, a medium frequency band, and a high frequency band; A first encoder which encodes the low frequency band signal using a first encoding scheme performed in a time domain; A second encoder which encodes the high-frequency band signal using a second encoding scheme performed on a frequency domain; A band extension encoder for generating band extension information for restoring a signal of the high frequency band by using the low frequency band signal and the high frequency band signal; And a bit packing unit generating a bitstream including the encoded low frequency band signal, the high frequency band signal, and the band extension information.

Hereinafter, a coding / decoding apparatus and method according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram illustrating a schematic configuration of an encoding apparatus according to the present invention. The illustrated encoding apparatus includes a signal splitter 100, an encoder 110, and a bit packing unit 120.

Referring to FIG. 1, the signal splitter 100 splits an input signal to be encoded into a plurality of signals. The signal splitter 100 may divide a signal input using a band pass filter into a plurality of frequency bands, for example, a low frequency band, a medium frequency band, and a high frequency band signal. The bandwidths of the divided frequency bands may be the same or different from each other.

The signal splitter 100 preferably divides an input signal so that interference between a plurality of divided signals, for example, a plurality of frequency band signals, can be minimized.

The encoder 110 includes a plurality of encoders 111, 112, and 113 to perform encoding using an encoder having a structure capable of encoding the divided signals most efficiently. The encoder which can encode the signal most efficiently may mean an encoder having the highest compression efficiency when encoding the signal among the plurality of encoders 111, 112, and 113.

The encoder 110 may include an encoder as many as the number of the divided signals, and encodes the encoded signals by one-to-one correspondence with the divided signals and the plurality of encoders 111, 112, and 113 according to the characteristics of the divided signals. Can be done.

For example, when the signal splitter 100 divides an input signal into a plurality of frequency bands, the frequency band and the encoder to encode the band may correspond to each other according to the characteristics of the frequency bands. In addition, in order to reduce the bit rate, encoding may be performed by applying different encoders to the low frequency bands of the divided frequency bands and using an encoder using a bandwidth extension method to the high frequency bands.

Encoding units for encoding the divided signals are predetermined according to the characteristics of the respective signals, or encoding units for encoding the respective signals most efficiently by grasping the characteristics of the divided signals during the encoding process. You can decide.

The bit packing unit 120 generates a bitstream using the plurality of encoded signals.

Information about the plurality of divided signals, for example, divided information such as the number of divided signals and divided frequency band information may be included in a transmitted bitstream, and the decoding apparatus decodes the divided information using the divided information. A plurality of signals may be synthesized to recover the original signal.

FIG. 2 is a block diagram illustrating an embodiment of a configuration of an encoding apparatus according to the present invention. The encoding apparatus illustrated in FIG. 2 includes a bandpass filter 200, a speech encoder 210, an audio encoder 220, and a bandwidth extension encoder. And a bit packing unit 240. A description of the same operations as those described with reference to FIG. 1 among operations of the encoding apparatus illustrated in FIG. 2 will be omitted.

The band pass filter 200 performs band pass filtering on the input signal to be encoded, and divides it into a low frequency band, a medium frequency band, and a high frequency band signal. For example, the low frequency band may be a band of 1 kHz or less, the medium frequency band may be a band of 1 kHz to 8 kHz, and the high frequency band may be a band of 8 kHz or more.

In general, in the case of a signal that is rapidly changing on the time axis, the most psychologically important part of the signal, that is, the part having the most information is the low frequency band.

Accordingly, as shown in FIG. 2, the low frequency band signal is preferably encoded using the speech encoder 210 which performs encoding in the time domain.

In addition, the mid-frequency band is a psychoacoustic band that has an important influence on the perception of presence and brightness attributes, and accurately identifies the harmonic component of a signal having continuous characteristics on the time axis. It is a necessary band to express.

Therefore, as shown in FIG. 2, the high frequency band signal is preferably encoded by using an audio encoder 220 which performs encoding in the frequency domain through domain transformation.

In general, the high frequency band has a psychoacoustic effect on the sound quality of the signal is relatively less than the low frequency band or the high frequency band signal. Therefore, in order to reduce the bit rate, it is preferable to encode the high frequency band signal using the band extension encoder 230.

The band extension encoder 230 generates band extension information necessary to recover the high frequency band signal from the low frequency band and the high frequency band signal.

The bit packing unit 240 uses the low frequency band signal encoded by the speech encoder 210, the high frequency band signal encoded by the audio encoder 220, and the band extension information generated by the band extension encoder 230. Create

Detailed embodiments of the voice encoder 210, the audio encoder 220, and the band encoder 230 will be described with reference to FIGS. 3 to 8.

3 is a block diagram illustrating a first embodiment of the configuration of the speech coder 210 illustrated in FIG. 2. The illustrated speech coder includes a linear prediction analyzer 211, a coefficient quantizer 212, and a linear prediction. The inverse filtering unit 213 and the residual signal quantization unit 214 are included.

The speech signal may be modeled with reference to a human speech instrument. That is, the vibration of the vocal cords can be replaced by an oscillator oscillating at an arbitrary frequency, and the part of the larynx to the mouth changes the sound spectrum generated by the vibration of the vocal cords, thereby changing the signal output from the oscillator. It can be substituted with. Accordingly, the audio signal can be encoded by determining the frequency of the oscillator and the coefficients representing the characteristics of the filter.

In addition, since human speech can be largely divided into voiced sound having periodic characteristics and unvoiced sound having noise characteristics without periodicity, the oscillator can be considered as a combination of a periodic signal oscillator and a noise signal oscillator.

Therefore, the speech signal can be encoded into coefficients representing characteristics in the frequency space, information on pitch, which is a periodic component, and information on noise components.

The linear prediction analyzer 211 obtains linear prediction coefficients by performing linear prediction analysis on the input signal. The obtained linear prediction coefficient may represent a spectral envelope of the input signal, that is, a signal characteristic in frequency space. Therefore, the linear prediction coefficients can express the characteristics of the filter among the substituted components of the speech signal.

As shown in Equation 1 below, the current signal X (n) may be expressed as a linear combination of past input signals.

In Equation 1, p is a linear prediction order, a ₁ to a _p are linear prediction coefficients, and e (n) is a residual signal representing an error of linear prediction. The linear prediction coefficients are determined in units of frames such that the residual signal e (n) is minimized throughout the frame.

The coefficient quantization unit 212 quantizes the obtained linear prediction coefficients. For example, the coefficient quantization unit 212 may convert the linear prediction coefficients into a LAR (Log Area Ratio) or LSP (Line Spectrum Pair), which is a parameter suitable for quantization, and then quantize them, and encode the linear prediction coefficients in an index form of a quantization table. have.

The linear predictive inverse filtering unit 213 performs linear prediction inverse filtering based on the obtained linear prediction coefficients for the speech signal to be calculated and encoded as in Equation 2 below.

In Equation 2, X (n) is an input signal, p is a linear prediction order, a ₁ to a _p are the obtained linear prediction coefficients, and e (n) is a prediction residual signal which is an output of linear prediction inverse filtering. .

Since the predicted residual signal e (n) obtained as described above is obtained by removing the spectral characteristics due to the linear prediction coefficients from the input signal to be encoded, information about the frequency of the oscillator among the substituted components of the speech signal may be expressed.

The residual signal quantization unit 214 performs quantization on the obtained prediction residual signal e (n).

The speech encoder 210 may encode the speech signal into quantized linear prediction coefficients and a prediction residual signal obtained by the above method.

FIG. 4 is a block diagram illustrating a second embodiment of the configuration of the speech coder illustrated in FIG. 2. The illustrated speech coder may include a pitch detector 300 and a linear predictive analyzer 310.

Referring to FIG. 4, the pitch detector 300 detects a pitch of a signal to be encoded. In the case of an audio signal, only one pitch may be included, but in the case of an audio signal, two or more pitches may be included.

As shown in FIG. 4, the pitch detector 300 may be represented by Equation 3 below.

In Equation 3, T denotes a period of a pitch, and g _p denotes a gain of the pitch.

The pitch detector 300 may encode a periodic component of the input signal by detecting a period and a gain of the pitch of the input signal in units of frames. In order to be applicable to not only an audio signal but also an audio signal, the encoding apparatus of the present invention preferably detects two or more pitch periods and gains in units of frames.

As described above, the linear prediction analyzer 310 performs a linear prediction analysis on the input signal to obtain a linear prediction coefficient. The linear prediction analyzer 310 may perform linear prediction analysis on a frame-by-frame basis for the input signal using the autocorrelation coefficient obtained by using an asymmetric window. In obtaining the autocorrelation coefficient, the linear predictive analysis unit 200 may perform a predictive interval, for example, when the asymmetric window has a length of 30 ms, a linear predictive analysis may be performed with a predictive interval having a length of 5 ms. The autocorrelation coefficients are converted to linear prediction coefficients using the Levinson-Durbin algorithm.

As described above, the obtained linear prediction coefficient may be transformed into a Log Area Ratio (LAR) or a Line Spectrum Pair (LSP), which is a parameter suitable for quantization, and then quantized and encoded in an index form of a quantization table.

For the periodic sound source signal represented by the input excitation signal and the detected pitch, LPC synthesis filtering may be performed using the obtained linear prediction coefficients to obtain a synthesized signal.

The excitation signal represents a noise component of an input signal and is determined to minimize an error between the synthesized signal and the input signal that is an original signal.

As shown in Figure 4, this signal has the gain (g _c) is multiplied makin, there is no periodicity in a strong periodic signal as the control gain (g _c) and the gain (g _p) of the pitch of the excitation signal It is possible to generate various signals up to noise signals. The gains g _c and g _p may be quantized and encoded in an index form of a quantization table.

FIG. 5 is a block diagram illustrating a third embodiment of the configuration of the speech encoder illustrated in FIG. 2.

The linear prediction unit 400 performs a linear prediction analysis on the input signal to obtain a linear prediction coefficient, and the quantization unit 410 performs a log area ratio (LAR) or LSP (Line) which is a parameter suitable for quantization of the obtained linear prediction coefficient. Spectrum Pair) and then quantize.

Pitch estimator 420 estimates the pitch period of the input signal. The adaptive codebook 430 stores sound source signals that are periodic components among the input signals of the past in the form of a codebook. The noise generator 440 generates the aperiodic noise signal, and may store the noise signals in the form of a codebook.

The adder 450 gains gain _gp , at each of the periodic sound source signal extracted from the adaptive codebook 430 and the noise signal output from the noise generator 440 according to the pitch period estimated by the pitch estimation unit 420. It added after multiplied by g _c) to generate a residual signal. In general, the residual signal refers to a signal obtained by performing linear prediction inverse filtering on an input signal using a linear prediction coefficient, and the residual signal output from the adder 450 is exactly a signal close to the residual signal.

The synthesis filter 460 generates a synthesis signal by performing synthesis filtering based on the quantized linear prediction coefficients on the residual signal output from the adder 450.

The error calculator 470 calculates an error between the input signal, which is the original signal, and the synthesized signal, and the auditory weight filter 480 gives an acoustic weight to the calculated error.

The error minimizing unit 490 determines a pitch period T _i , a gain g _pi , g _ci , and a noise signal to minimize the error in consideration of the auditory characteristic. The excitation signal refers to the noise signal determined to minimize the error.

The voice encoder described with reference to FIGS. 2 to 5 is only one embodiment of the configuration of the encoding apparatus according to the present invention, and the voice encoder 210 illustrated in FIG. 2 performs various speech encodings for encoding in the time domain. Ways can be used.

FIG. 6 is a block diagram illustrating a first embodiment of the configuration of the audio encoder 220 illustrated in FIG. 2. The illustrated audio encoder includes a domain converter 221, a psychoacoustic modeling unit 222, and a quantization unit. 223.

The domain converter 221 converts an input signal into a frequency domain using a filter bank. For example, the domain transform unit 221 may perform a cosine transform, for example, a modified disc transform (MDCT) on the input signal.

The psychoacoustic modeling unit 222 calculates a masking threshold of the input signal or a signal-to-mask ratio (SMR). The quantizer 223 performs quantization on the MDCT coefficients output from the domain transform unit 400 using the masking threshold. In addition, the quantization unit 223 may use the signal-to-mask ratio SMR to minimize audible distortion of the quantized signal within a given bit rate.

FIG. 7 is a block diagram illustrating a second embodiment of the configuration of the audio encoder illustrated in FIG. 2. The illustrated audio encoder includes a preprocessor 500, a domain converter 510, a vector quantizer 520, Domain inverse transform unit 530 and gain optimization / quantization unit 540.

The preprocessor 500 performs filtering and windowing on the input signal to convert the input signal into a signal efficient for encoding.

The domain transform unit 510 performs a frequency domain transform, for example, a Fast Fourier Transform (FFT), on the input signal on which the preprocessing has been performed, and the vector quantizer 520 performs vector quantization to perform a codebook index. Outputs the code vector index and the Voronoi extention index.

In order to improve the encoding of low energy blocks before the first spectral peak, spectral pre-emphasis and de-emphasis may be performed before and after the vector quantization. have.

The domain inverse transformer 530 performs an inverse transform, for example, an inverse FFT, on the output signal of the vector quantizer 520.

The gain optimization / quantization unit 540 calculates a gain difference between the signal before the domain transform and the signal after the domain inverse transform, and outputs an optimum gain index.

The indices output from the vector quantizer 520 and the optimum gain indices output from the gain optimizer / quantizer 540 may be included in the bitstream as data for restoring a signal.

The audio coder described with reference to FIGS. 6 to 7 is only an embodiment of a configuration of an encoding apparatus according to the present invention, and the audio coder 220 illustrated in FIG. 2 is an AAC (Advanced Audio Coding) or a TCX (Transform). Various coding schemes that perform encoding using domain transformation such as Coded Excitation may be used.

FIG. 8 is a block diagram illustrating an embodiment of the configuration of the extended coder 230 shown in FIG. 2. The extended band coder includes a linear predictive inverse filtering unit 231 and a high frequency signal synthesis unit 232. , A gain vector calculator 233, a linear predictive analyzer 234, a predictive gain vector calculator 235, and a gain correction information generator 236.

The linear prediction band filtering unit 231 performs a linear prediction band filtering as shown in Equation 2 on the low frequency / medium frequency band signal to obtain a residual signal. The high frequency signal synthesis unit 232 generates a synthesized high frequency band signal by performing high frequency synthesis filtering on the obtained residual signal. When the synthesized high frequency band signal is in a down-sampling state, the synthesized high frequency band signal may be restored after being up-sampled.

The gain vector calculator 233 calculates a gain vector having information on gain between the two signals after performing perceptual filtering on the actual high frequency band signal and the synthesized high frequency band signal.

The linear prediction analyzer 234 performs linear prediction analysis on the low / medium frequency band signal and the high frequency band signal, and the predictive gain vector calculator 235 predicts the gain having information about the gain between the two signals. Calculate the vector

The correction information generator 236 calculates a difference between the gain vector calculated by the gain vector calculator 233 and the predicted gain vector calculated by the predicted gain vector calculator 235 to correct the gain of the high frequency band signal. Generate calibration information.

The decoding apparatus according to the present invention may reconstruct a signal from an input bitstream by performing an inverse process of the encoding process of the encoding apparatus described with reference to FIGS. 1 to 8.

9 is a block diagram illustrating an embodiment of a configuration of a decoding apparatus according to the present invention. The decoding apparatus illustrated in FIG. 9 includes a bit unpacking unit 600, a voice decoder 610, an audio decoder 620, A bandwidth extension decoder 630 and a signal synthesizer 640 are included.

The bit unpacking unit 600 extracts band extension information for reconstructing the encoded low frequency band signal, the high frequency band signal, and the high frequency band signal from the input bitstream.

The speech decoder 610 decodes the encoded low frequency band signal using a speech decoding scheme performed in the time domain.

The audio decoder 620 performs decoding using the audio decoding scheme performed on the encoded high frequency band signal in the frequency domain.

The decoding schemes performed by the speech decoder 610 and the audio decoder 620 may perform an inverse process of the speech or audio encoding scheme described with reference to FIGS. 3 to 7.

For example, the speech decoder 610 may reconstruct a signal by performing linear prediction synthesis filtering using the linear prediction coefficients extracted from the bitstream. More specifically, the encoded low frequency band signal extracted from the bitstream includes pitch information and excitation signal information of a signal to be restored, and the speech decoding unit 610 uses the pitch information and the excitation signal information to generate a residual signal. The signal is reconstructed by performing linear prediction synthesis filtering based on the extracted linear prediction coefficients on the reconstructed residual signal.

In addition, the audio decoder 620 may restore the signal by performing inverse quantization of the encoded data using the scale factor extracted from the bitstream. The encoded high frequency band signal extracted from the bitstream may include a Huffman codebook index or an MDCT coefficient normalized by the scale factor, and the audio decoder 620 performs lossless decoding using the Huffman codebook index. The normalized MDCT coefficients may be inversely quantized using the scale factor. The dequantized coefficients are domain transformed onto the time domain for reproduction of the signal.

The band extension decoder 630 generates a high frequency band signal using the reconstructed low frequency / medium frequency band signal and the extracted band extension information. The bandwidth extension decoder 630 may generate the high frequency band signal by performing an inverse process of the operation of the bandwidth extension encoder described with reference to FIG. 8.

The encoding / decoding apparatus according to the present invention described above is provided in a multimedia broadcasting transmission / reception apparatus such as digital audio broadcasting (DAB) or digital multimedia broadcasting (DMB), and may be used to encode / decode audio signals or audio signals. . In addition, the multimedia broadcasting transmission / reception apparatus may include a mobile communication terminal.

The encoding / decoding method according to the present invention described above may be stored in a computer-readable recording medium that is produced as a program for execution on a computer, and multimedia data having a data structure according to the present invention may also be read by a computer. Can be stored in a recording medium. The computer readable recording medium includes all kinds of storage devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like, and may also be implemented in the form of a carrier wave (for example, transmission over the Internet). Include. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for implementing the user tracking method can be easily inferred by programmers in the art to which the present invention belongs. In addition, the bitstream generated by the above-described encoding method may be stored in the computer-readable recording medium or transmitted using a wired / wireless communication network.

The decoding apparatus according to the present invention described above may be provided in a mobile communication terminal, a multimedia player such as a portable multimedia player (PMP), and a playback device such as a personal digital assistant (PDA). In addition, the above-described decoding apparatus may be implemented in hardware and included in a reproduction apparatus or the like, or as described above, may be implemented in a reproduction apparatus in software.

While the above has been shown and described with respect to preferred embodiments of the present invention, the present invention is not limited to the specific embodiments described above, it is usually in the technical field to which the invention belongs without departing from the spirit of the invention claimed in the claims. Various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical spirit or the prospect of the present invention.

According to the encoding / decoding method and apparatus according to the present invention as described above, the signals to be encoded are divided into a plurality of frequency bands, and low-frequency band signals are encoded using different coding schemes according to characteristics of each frequency band, The band signal can be encoded and decoded by using a band extension coding method, at low bit rates, signals having various characteristics.

Claims (17)

Extracting the encoded low frequency band signal, the encoded medium frequency band signal, and band extension information from the input bitstream; Decoding the encoded low frequency band signal using a first decoding method; Decoding the encoded high frequency band signal using a second decoding method; Restoring a high frequency band signal from the decoded low frequency band signal and the high frequency band signal by using the extracted band extension information; And And synthesizing the decoded low frequency band signal and the high frequency band signal and the reconstructed high frequency band signal. The method of claim 1, The first decoding method is performed in the time domain. The method of claim 1, And the first decoding method uses linear prediction coefficients extracted from the bitstream. The method of claim 1, The second decoding method is performed in the frequency domain. The method of claim 1, wherein the band extension information is And information on a spectral envelope of the high frequency band signal. The method of claim 5, The information on the spectral envelope is obtained by linear prediction analysis. The method of claim 7, wherein the band extension information is And information about a gain between the low frequency and high frequency band signals and the high frequency band signal. The method of claim 7, wherein the band extension information is And information for correcting the gain. Dividing an input signal into a low frequency band, a medium frequency band, and a high frequency band; Encoding the low frequency band signal using a first encoding scheme performed on a time domain; Encoding the high frequency band signal using a second coding scheme performed on a frequency domain; Generating band extension information for reconstructing the signal of the high frequency band by using the low frequency band signal and the high frequency band signal; And And generating a bitstream including the encoded low frequency band signal and the high frequency band signal and the band extension information. The method of claim 9, The first encoding method comprises the step of performing a linear prediction analysis. The method of claim 9, wherein the band extension information is And at least one of information on a spectral envelope of the high frequency band signal and information on a gain between the low frequency and high frequency band signals and the high frequency band signal. A bit unpacking unit which extracts an encoded low frequency band signal, an encoded medium frequency band signal, and band extension information from an input bitstream; A first decoder which decodes the encoded low frequency band signal using a first decoding method performed in a time domain; A second decoder which decodes the encoded high frequency band signal using a second decoding method performed in a frequency domain; A band extension decoding unit which restores a high frequency band signal from the decoded low frequency band signal and the high frequency band signal by using the extracted band extension information; And And a synthesizer for synthesizing the decoded low frequency band signal and the high frequency band signal and the restored high frequency band signal. The method of claim 12, wherein the first decoding unit And decoding the low frequency band signal using the linear prediction coefficients extracted from the bitstream. The method of claim 12, wherein the band extension information is And at least one of information on a spectral envelope of the high frequency band signal and information on a gain between the low frequency and high frequency band signals and the high frequency band signal. A signal splitter for dividing an input signal into a low frequency band, a medium frequency band, and a high frequency band; A first encoder which encodes the low frequency band signal using a first encoding scheme performed in a time domain; A second encoder which encodes the high-frequency band signal using a second encoding scheme performed on a frequency domain; A band extension encoder for generating band extension information for restoring a signal of the high frequency band by using the low frequency band signal and the high frequency band signal; And And a bit packing unit configured to generate a bitstream including the encoded low frequency band signal, the high frequency band signal, and the band extension information. A computer-readable recording medium having recorded thereon a program for executing the method according to any one of claims 1 to 11. A multimedia broadcasting apparatus comprising the apparatus according to any one of claims 12 to 15.

Images