KR101128353B1 - Method of processing a signal and a High efficiency and directivity speaker system using a Block-based Detecting Signal - Google Patents

Abstract

The present invention relates to a high efficiency, highly directional speaker system based on block unit input signal detection and a method for signal processing.

The high-efficiency high-direction speaker system based on the block unit input signal detection includes: an envelope detector configured to calculate an envelope of an audio input signal currently input; A block unit signal detection unit dividing the signal calculated by the envelope calculation unit into blocks having a predetermined length, and calculating a block energy by squaring and adding the signals in the block; A block adaptive filter calculator configured to compare a value obtained by applying the block filter coefficients obtained from the current block input signal and the previous block input signal and calculate a current block filter coefficient accordingly; An SSB modulator for SSB modulating an input signal to which the block adaptive filter coefficient value is applied; A transducer inverse filter unit for modeling and applying an inverse filter corresponding to a frequency characteristic of the ultrasonic transducer; And an ultrasonic conversion unit for converting the modulated signal into ultrasonic waves, which can detect a time block unit signal, thereby reducing unnecessary power consumption used in the silent section, and a very fast convergence speed due to the block unit signal processing technology. It can be applied to the input signal in real time, and the SSB modulation of the modulated signal can minimize the distortion change of the input signal, frequency domain and various other block unit algorithms can be applied, Since the length N of the signal detection block and the length M of the block filter can be set differently, it is possible to design an optimal directivity system having a low cost, high efficiency and high directivity suitable for various application environments.

Block Signal Processing, Signal Detection, High-Directional Speaker, Ultrasound, Ultrasonic Converter, Single Sideband, SSB, Frequency Domain Signal Processing

Description

Method of processing a signal and a High efficiency and directivity speaker system using a Block-based Detecting Signal}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an ultrasonic speaker for reproducing high efficiency high-directional signals, and more particularly, to a directional speaker system and a signal processing method to which a new signal processing method for improving sound quality while improving power efficiency of an ultrasonic speaker system is applied.

Recently, with the development of signal processing technology, electronic devices having a water-directivity are being developed one after another. In the mobile communication field, a directional communication system using an array antenna has been developed, and in the defense and entertainment industries, the use of a directional speaker in the audible frequency band is increasing, and research and development are being actively conducted. A representative technology of directional speakers is parametric speaker technology that uses the nonlinear characteristics of air. Parametric loudspeakers have high directivity but very low efficiency, so it is essential to develop high-efficiency high-directional loudspeaker source technology.

In general, speakers deliver sound in the air in all directions. However, in places such as exhibitions and exhibitions where sound needs to be delivered only to specific users, sound directivity is needed. In addition, the development of high-direction speakers is essential for sound attack or fleet long range voice communication for military purposes.

Traditionally, a parabolic dish is a speaker designed to have a general speaker at the focal point of the dish so that the sound output is reflected on the dish so that it is straight. This method requires a very large diameter plate, has a short range of sound and has a disadvantage in that sound quality is degraded due to mutual interference of reflected signals. Recently, ultrasonic speaker technology using nonlinear propagation characteristics of an air medium has been implemented to compensate for these disadvantages. Techniques have been developed to reduce distortion due to the nonlinear nature of air.

ATC filed "Modulator Processing for a Parametric Speaker System (US 6,584,205)" proposed an iterative error compensation method without an infinite increase in bandwidth.

The proposed method calculates the ideal modulated signal waveform through SSB (Single Side Band) channel model, compares it with the actual modulated signal, calculates the error and compensates this error for the signal before modulation several times. This method compensates for the distortion of sound quality.

Also, conventionally, a method due to the signal processing method as shown in FIG. 1 has been disclosed.

An adaptive filter calculator for comparing the envelope of the current audio input signal with the envelope to which the adaptive filter coefficients obtained from the previous stage audio input signal are applied and calculating the current adaptive filter coefficient accordingly; A VSB modulator for VSB modulating an audio input signal to which the adaptive filter coefficient value is applied; And an ultrasonic converter for converting the modulated signal into ultrasonic waves, wherein the adaptive filter calculator includes a first envelope calculator 10, a square root calculator 20, a second envelope calculator 40, an error operator 50, A pre-distortion adaptive filter unit 30 for applying the adaptive filter coefficients, a VSB modulating unit, a dynamic VSB modulating unit 70, an ultrasonic transducer unit, an ultrasonic transducer model 80, and an ultrasonic amplifier 90, the ultrasonic transducer 100 is configured.

In other words, the super-directional speaker system of the present invention calculates an envelope of the currently input audio input signal x (t) to generate a first envelope signal E (t) with respect to the envelope; Square root calculator 20 for calculating the ideal envelope signal E (t) 0.5 using the first envelope signal E (t) generated by the first envelope calculator 10, the previous stage audio input signal x (t− A pre-distortion adaptive filter unit 30 performing pre-distortion compensation of the current audio input signal x (t) by applying the adaptive filter coefficient update term calculated from the envelope of 1) to generate the distortion compensation signal x (t) ', The second envelope calculator 40 and the square root calculator 40 which calculate the envelope of the distortion-compensated compensation signal x (t) 'output from the pre-distortion adaptive filter unit 30 to generate the second envelope signal E (t)'. 20) compares the output signal E (t) 0.5 and the second envelope signal E (t) ' The error calculation unit 50 for outputting the difference signal e (t) and the adaptive filter coefficient update for calculating the coefficient update terms of the predistortion adaptive filter corresponding to the error signal e (t) and supplying the predistortion adaptive filter unit 30. The dynamic VSB modulator 70 generates a modulated signal x (t) " by dynamically modulating the distortion-compensated compensation signal x (t) 'output from the pre-distortion adaptive filter unit 30 into an ultrasonic band. ), An ultrasonic transducer that generates a converted signal x (t) '' 'by modeling an inverse filter h (t) corresponding to a frequency-specific characteristic of the ultrasonic transducer 100 and applying the modulated signal x (t)' '. The model 80, the ultrasonic amplifier 90 which amplifies the converted signal x (t) '' 'output from the ultrasonic transducer model 80 to generate the amplified signal x (t) "", and the ultrasonic amplified amplified signal x ( and an ultrasonic transducer 100 for converting the output of "t)" "into an ultrasonic signal.

This method improves the sound quality by minimizing the distortion of the real-time reproduction signal by using the pre-distortion adaptive filter and by removing the imperfection of the single-side band filter by using the residual sideband modulation.

Since the first method is not a method based on input signal detection, a power amplifier consumes power at all times, and thus, power efficiency is inferior.

In addition, the second method described above calculates an adaptive filter using every input signal, and the convergence time may vary according to the situation, the amount of distortion error may vary, and the performance of the input signal may vary depending on the environment of the input signal. The disadvantage is that change can occur badly.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and an object of the present invention is to provide a directional speaker system and a signal processing method with high power efficiency by using block unit signal detection.

Another object of the present invention is to compare a signal of a block unit input signal and a signal to which a block filter coefficient of a previous input signal is applied to each other, and calculate and apply a block filter coefficient value of a current input signal accordingly, thereby applying high-speed high-performance ultrasonic speaker sound quality. It is to provide a directional speaker system and a signal processing method that can be improved.

Another object of the present invention is to provide various frequency domain signal processing methods and block filter signal processing methods such as FFT, Discrete Cosine Transform (DCT), and the like due to the use of block filters in the directional system configuration. .

Still another object of the present invention is that the block length (N) and the block filter length (M) for the input signal detection can be combined in different lengths, the directivity that can ensure signal detection and system performance in a variety of input situations The present invention provides a speaker system and a signal processing method.

Still another object of the present invention is to generate an inverse filter model of an ultrasonic transducer applied to a block-based signal processing system and apply it to an SSB modulated signal, thereby minimizing distortion during ultrasonic conversion of the modulated signal, thereby improving sound quality. To provide a directional speaker system and a signal processing method.

According to an aspect of the present invention, there is provided a high-efficiency high-direction speaker system based on block unit signal detection, including: an envelope detector configured to calculate an envelope of an audio input signal currently input; A block unit signal detection unit dividing the signal calculated by the envelope calculation unit into blocks having a predetermined length, and calculating a block energy by squaring and adding the signals in the block; A block adaptive filter calculator configured to compare a value obtained by applying the block filter coefficients obtained from the current block input signal and the previous block input signal and calculate a current block filter coefficient accordingly; An SSB modulator for SSB modulating an input signal to which the block adaptive filter coefficient value is applied; A transducer inverse filter unit for modeling and applying an inverse filter corresponding to a frequency characteristic of the ultrasonic transducer; And an ultrasonic converter for converting the modulated signal into ultrasonic waves.

As an example, the block adaptive filter calculator may include a square root calculator configured to generate a square root signal by calculating a square root of a currently input block signal; A block filter unit for performing signal compensation on the currently input block signal by applying a block filter coefficient calculated by a previous block signal; An error calculator configured to calculate an error by comparing and calculating the block filter output signal and the square root signal of the block signal; And a block adaptive filter coefficient calculator for calculating block adaptive filter coefficients from the error signal.

As an example, the block adaptive filter coefficient calculating unit may apply at least one of a frequency domain method, a discrete fourier transform (DFT) method, and a discrete cosine transform (DCT) method.

As an example, the block adaptive filter coefficient calculator may compare the square root of the current block input signal with the output of the block filter unit obtained from the previous block input signal, and calculate the current block adaptive filter coefficient accordingly.

As an example, the block filter unit may include a linear block FIR filter.

A signal processing method of a high efficiency high-direction speaker based on block unit signal detection for achieving the above object of the present invention, the first step of calculating and generating the envelope of the currently input signal; Dividing the signal calculated by the envelope detector into blocks having a predetermined length, and squaring and adding signals in the block to generate block energy; A third step of generating a square root signal by calculating a square root of the block signal calculated by the block unit signal detection unit; A fourth step of generating an output signal by applying the block filter coefficient determined in the previous stage to the input block signal; Generating an error signal between the block filter output signal and the square root signal of the block signal; A sixth step of producing a block adaptive filter value for calculating a block filter coefficient from the error signal; A seventh step of modulating the block filter output signal into an ultrasonic band to generate an SSB modulated signal; And an eighth step of modeling and applying an inverse filter corresponding to the frequency characteristic of the ultrasonic transducer. And a ninth step of amplifying the filtering signal and converting the filtered signal into an ultrasonic wave for converting the signal into an ultrasonic signal.

As an example, the method may further include a tenth step in which the ultrasonic signal is nonlinearly demodulated in air and converted into an output.

As an example, the inverse filter is characterized in that the ultrasonic transducer used in the eighth step is modeled as a specific filter and calculated and applied using frequency characteristics.

The high efficiency high directional speaker system and signal processing method based on block unit signal detection according to an exemplary embodiment of the present invention can implement an economical high oriented speaker system by increasing power efficiency of a power amplifier because signals are detected and processed on a block basis. It has an effect.

In addition, by comparing the signal of the block unit input signal and the signal to which the block filter coefficients of the previous input signal is applied to each other and calculates and applies the block filter coefficient value of the current input signal according to the high speed of the high-directional speaker due to the fast convergence speed The sound quality can be improved and various frequency domain signal processing methods such as FFT, Discrete Cosine Transform (DCT), and the like can be used due to the use of block filters in directional speaker system construction.

In addition, the block length (N) and the block filter length (M) for input signal detection can be used in combination with each other to implement a directional speaker system that can ensure signal detection and system performance in various input situations. By generating an inverse filter model of the applied ultrasonic transducer and applying it to the SSB modulated signal, there is an effect of improving the sound quality by minimizing distortion during ultrasonic conversion of the modulated signal.

Hereinafter, embodiments of the present invention will be described in detail with reference to FIGS.

FIG. 2 is a diagram illustrating a high efficiency high-direction speaker system based on block unit signal detection according to the present invention.

Referring to FIG. 2, the high-directional speaker system divides the signal calculated by the envelope detector 101 and the envelope detector 101 that calculate the envelope of the currently input audio input signal into blocks having a predetermined length, and divides the signal in the block. The block unit signal detector 102, which calculates block energy by squares and sums, and the block adaptive filter calculator that compares a value obtained by applying the block filter coefficients obtained from the current block input signal and the previous block input signal and calculates the current block filter coefficient accordingly. 110, an SSB modulator 120 for SSB modulating the input signal to which the block adaptive filter coefficient value is applied, and an ultrasonic converter 140 for converting the modulated signal into ultrasonic waves.

The block adaptive filter calculator 110 includes a square root calculator 112, a block adaptive filter coefficient calculator 118, and a block filter 114 for applying block filter coefficients.

를 계산하는 제곱근 연산부(112), 이전 블록 입력신호 b(t-1)에 사용된 블록 적응 필터 계수를 현재 블록 입력신호 b(t)에 적용하여 블록 필터 출력 y(t)를 생성하는 블록 필터부(114), 상기 블록 필터부(114) 에서 출력되는 신호 y(t)와 제곱근 연산부(112)의 출력신호인

를 비교 계산하여 그 오차 신호 e(t)를 출력하는 오차계산부(116), 오차 신호 e(t)에 따라 블록 적응 필터의 계수를 계산하고 수정하여 블록 필터부 (114)에 공급하는 블록 적응 필터 계수 계산부(118), 블록 필터부(114)에서 출력되는 신호 y(t)를 초음파 대역으로 변조시켜 변조 신호 z(t)를 생성하는 SSB 변조부(120), 초음파 변환부(140)가 가지는 주파수 응답 특성을 보상하는 역필터를 추정하여 변조된 신호 z(t)에 적용함으로써 변환 신호 x(t)를 생성하는 초음파 변환기 역필터(130), 상기 변환기 역필터(130)로부터 출력되는 변환 신호 x(t)를 증폭하고 출력을 초음파 신호 o(t) 로 변환하는 초음파 변환부(140)로 구성되어 있다. In more detail, the high-efficiency high-directional speaker system of the present invention includes an envelope detector 101 and an envelope detector 101 that detect an envelope of an input signal s (t) that is currently input to generate an envelope signal E (t). The block signal b generated by the block unit signal detector 102 and the block unit signal detector 102 that divide the generated envelope signal E (t) into blocks of a predetermined length, square and sum the signals in the block to calculate the block energy. (t) is an ideal envelope signal

Square root computing unit 112 for calculating a block filter for applying a block adaptive filter coefficient used in the previous block input signal b (t-1) to the current block input signal b (t) to generate a block filter output y (t). And a signal y (t) output from the block filter unit 114 and an output signal of the square root calculating unit 112.

Is calculated by comparing and calculating the error signal e (t) and outputting the error signal e (t), and calculating and modifying the coefficients of the block adaptive filter according to the error signal e (t). SSB modulator 120 and ultrasonic converter 140 which modulate signal y (t) output from filter coefficient calculator 118 and block filter 114 into an ultrasonic band to generate a modulated signal z (t). Ultrasonic transducer inverse filter 130 for generating a converted signal x (t) by estimating an inverse filter that compensates the frequency response characteristic of the modulated signal z (t) and outputting from the transducer inverse filter 130 And an ultrasonic converter 140 for amplifying the converted signal x (t) and converting the output into an ultrasonic signal o (t).

Detailed explanation of each step is as follows.

The envelope detector 101 calculates an envelope signal E (t) for the current input signal s (t), and a generally known envelope signal is calculated as follows.

Equation 1

E (t) '= 1 + ms (t)

Where m is the modulation index.

The block unit signal detector 102 divides the envelope signal E (t) generated by the envelope detector 101 into blocks having a predetermined length, squares and sums the signals in the block, and calculates block energy. Only more than one signal block b (to) can be processed in the next step.

If the length of the block is N, b (to) is expressed as

Equation 2

b (to) = E (to), if || E (to) ||> threshold

Where N is the length of the block for signal detection and E (to) is the block signal vector of length N and the starting point to.

The starting point of the next block is updated by the following equation to perform a new block signal detection.

Equation 3

to = to + N

를 계산한다. The square root calculator 112 is an ideal square root envelope signal of the envelope signal b (t1) detected by the block unit signal detector 102 at the present time t1.

Calculate

The block filter 114 applies the block adaptive filter coefficient w (t1) calculated by the previous stage input signal b (t1-M) to the block input signal b (t1) input at the present time t1, The signal y (t) is output in the same way.

Equation 4

y (t1) = diag (b (t1)) w (t1)

Here diag means changing a diagonal matrix of a vector, and w (t1) means a block filter coefficient vector having a length M at a time point t1.

The starting point of the next block for the block adaptive filter calculation is updated by the following equation.

Equation 5

t1 = t1 + M

According to the above equation, it can be seen that various combinations of N and M are possible, and to and t1 change accordingly.

For example, if the block length N = 512 and the block filter length M = 64 for the signal detection, the energy of the input signal is calculated every 512 sample periods, and the detected signal block is N / M = 8 block adaptation. Updated by the calculator and filtered by the block filter.

The error calculating unit 116 calculates an error signal vector e (t1) of the output signal vector y (t1) of the block filter unit 114 and the signal vector b '(t1) calculated by the square root calculating unit 112 and an error. E (t) calculated by the calculator 116 is expressed by Equation 5 below.

Equation 6

e (t1) = b '(t1)-y (t1)

The block adaptive filter coefficient calculator 118 applies a Discrete Fourier Transform (DFT), a Discrete Cosine Transform (DCT), a Frequency Domain (Frequency Domain) scheme, and the like to the error signal e (t) calculated by the error calculator 116. The block adaptive filter coefficient vector is calculated. The method of calculating block filter coefficients from the error signal vector e (t) of the present invention can be applied to various other methods. In order to help the understanding of the present invention, the following description will focus on the FFT method. The newly calculated w (t1 + 1) by the adaptive filter coefficient calculator 118 may be expressed as shown in Equation (7).

Equation 7

w (t1 + 1) = w (t1) + β * FFT [Q (t1) 0] T

Where β is an adaptive coefficient, 0 is a vector of all zero factors, Mx1, and β is a positive constant of small magnitude in time.

And Q (t1) is as follows.

Equation 8

Q (t1) = first M factor of IFFT ([D (t1) UT (t1) Er (t1)])

In the above equation, Er (t1) is e (t1) as the FFT vector.

Equation 9

Er (t1) = FFT ([0 e (t1)] T)

In the above equation, e (t1) is equal to Equation 6.

e (t1) = b '(t1)-y (t1)

Where y (t1) is expressed as

Equation 10

yT (t1) = last M factor of IFFT ([B (t1) w (t1)])

B (t1) is composed of a diagonal matrix by FFT of b '(t1) and can be written as the following equation.

Equation 11

B (t1) = diag (FFT (b '(t1)))

The block filter 114 is applied in real time to the block input signal b (t1) input to the next stage by using the update term w (t1 + 1) obtained by the block adaptive filter coefficient calculator 118. .

The block filter 114 may be implemented using a finite impulse response (FIR), a finite impulse response (IIR) lattice structure, or the like.

The SSB modulator 120 modulates the output block signal y (t) into the ultrasonic band through the upper or lower band SSB modulation by the block filter 114.

The converter inverse filter 130 estimates an inverse filter by grasping the frequency characteristics of the ultrasonic transducer 140 and applies the inverse filter to the modulated signal z (t) generated by the SSB modulator 120 to filter the inverse filter. Generate the signal x (t).

The ultrasonic converter 140 generates the final signal o (t) by radiating ultrasonic waves using the ultrasonic vibrating element filtered by the converter inverse filter x (t). The ultrasonic signal converted by the ultrasonic transducer 140 is radiated into the air to be demodulated by the nonlinear medium characteristic of the air.

A high-efficiency high-directional speaker system and signal processing method based on block unit signal detection according to an embodiment of the present invention having such a structure will be described with reference to FIG. 3 as follows.

The signal processing method of the block-based signal detection-based speaker system of the present invention calculates an envelope of the currently input signal s (t) (S300), and calculates a block unit signal b (t) of the calculated envelope signal E (t). Calculate the energy and compare it with the threshold. If the value is small, the next signal block is taken and the same comparison process is performed. This process is repeated until the block signal is detected (S302).

를 생성한다(S304).If the energy of the current signal block is greater than the threshold, the square root operation is performed to perform the signal

To generate (S304).

Meanwhile, while the block unit signal detection step 300 and step 302 continue, the current input signal b (t) applies the block filter vector calculated from the input signal b (t-1) of the previous stage to output the output signal '. It generates (S312).

의 차이를 계산하여, 그 블록 오차 신호 e(t)를 생성한다(S308).Next, the block filter output signal and the ideal

Is calculated, and the block error signal e (t) is generated (S308).

Next, the block weight vector amount is calculated according to the e (t) signal calculated in step s4 (S5). At this time, a Discrete Fourier Transform (DFT), a Discrete Cosine Transform (DCT), and a Frequency Domain (Frequency Domain) adaptation scheme are used.

Next, the output signal to which the block adaptive filter coefficient calculated in step 312 is applied is SSB modulated to generate a modulated signal x (t) "(S314).

Next, the ultrasonic transducer inverse filter is applied to the SSB modulated signal y (t) (S316).

In this case, the applied inverse filter may be calculated through analysis of frequency response characteristics of the ultrasonic transducer 140 used in the system.

Next, the ultrasonic transducer 140 amplifies the output signal x (t) of the transducer inverse filter 130 and ultrasonically converts the amplified signal (S318).

Finally, the ultrasonic signal is nonlinearly demodulated in air and output as an output signal o (t).

Since the high-efficiency high-directional speaker system based on block unit signal detection according to an embodiment of the present invention uses a block adaptive filter, distortion compensation may be processed in block signal units in real time.

 Therefore, the high-efficiency high-directional speaker system based on the block unit signal detection according to the present invention has a fast convergence characteristic, thereby reducing system delay and making it easy to construct a highly efficient system. In addition, since the block-based signal processing of the present invention, it is possible to build a variety of systems by adjusting the size of the block.

The above description has been described as an example by the embodiment of the present invention, but is not limited to the above-described embodiment and those skilled in the art will appreciate that various changes and modifications can be made without departing from the spirit of the present invention. . Therefore, the technical scope of the present invention should not be limited to the contents described in the detailed description in the specification but should be defined by the claims.

1 is a view showing an audio input signal processing method using a pre-distortion less filter method in a conventional high-directional speaker system.

2 is a diagram illustrating a high efficiency high-direction speaker system based on block unit signal detection according to an embodiment of the present invention.

3 is a flowchart illustrating a signal processing method of a high efficiency high-direction speaker system based on block unit signal detection according to an embodiment of the present invention.

<Description of Symbols for Main Parts of Drawings>

101: envelope detection unit

102: block unit signal detection unit

112: square root calculator

114: block filter unit

116: error calculation unit

118: block adaptive filter coefficient calculation unit

120: SSB modulator

130: converter reverse filter

140: ultrasonic transducer

Claims (8)

An envelope detector configured to calculate an envelope of an audio input signal currently input; A block unit signal detector for dividing the signal calculated by the envelope detector into blocks of a predetermined length, calculating a block energy by squaring and summing signals in the block; A block adaptive filter calculator configured to compare a value obtained by applying the block filter coefficients obtained from the current block input signal and the previous block input signal and calculate a current block filter coefficient accordingly; An SSB modulator for SSB modulating the input signal to which the coefficient calculated by the block adaptive filter calculator is applied; A transducer inverse filter unit for modeling and applying an inverse filter corresponding to a frequency characteristic of the ultrasonic transducer; And Including an ultrasonic converter for converting the modulated signal to ultrasonic, The block adaptive filter calculator A square root calculator configured to generate a square root signal by calculating a square root of a currently input block signal; A block filter unit generating a block filter output signal by applying a block filter coefficient calculated by a previous block signal to the currently input block signal; An error calculator configured to calculate an error signal by comparing and calculating a signal generated by the block filter unit and a square root signal generated by the square root calculator; And a block adaptive filter coefficient calculator for calculating block adaptive filter coefficients from the error signal calculated by the error calculator. delete The method of claim 1, The block adaptive filter coefficient calculation unit is a high-efficiency high-direction speaker system based on block unit signal detection, characterized in that at least one of a frequency domain method, a Discrete Fourier Transform (DFT) method, a Discrete Cosine Transform (DCT) method is applied. . The method of claim 1, The block adaptive filter coefficient calculating unit compares the square root value of the current block input signal with the output of the block filter unit obtained from the previous block input signal and calculates the current block adaptive filter coefficient according to the high efficiency based on block unit signal detection. High directional speaker system. The method of claim 1, The block filter unit is a high efficiency high-direction speaker system based on block-by-block signal detection, characterized in that it comprises a linear block FIR filter. delete delete delete

Images