Digital decimation filter, data processing method and equipment
By adding the first type of filtering only and not decimating to the digital decimation filter of Sigma-Delta ADC, the high computational complexity and delay problems caused by the multi-stage cascade structure are solved, and more efficient signal conversion and Voice interaction synchronization.
Patent Information
- Application Number
- CN202411189041.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-27
- Publication Date
- 2024-11-22
AI Technical Summary
The digital decimation filter in Sigma-Delta ADC adopts a multi-stage cascade structure, resulting in a large coefficient of finite impulse response, high computational complexity, and large signal conversion delay, which cannot guarantee the real-time nature of voice interaction.
Add a first type filter to the digital decimation filter that only filters but not decimation. Through this filter, the digital input signal is preprocessed to remove noise and improve signal quality, thereby reducing the number of coefficients of the decimation filter and Hardware structure to improve signal conversion rate.
The hardware structure and chip area of the digital decimation filter are reduced, the synchronization of voice interaction and signal conversion rate are improved, and the delay and calculation amount of signal processing are reduced.
Smart Images

Figure CN119010838A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present application relate to the field of digital signal processing technology, and relate to but are not limited to a digital extraction filter, a data processing method and a device. Background Art
[0002] With the rapid development of digital signal processing, people have an increasingly urgent need for high-performance analog-to-digital converters (ADCs) as analog-to-digital interfaces. Sigma-Delta ADCs have been widely used in high-precision analog-to-digital conversion fields such as digital audio and data acquisition due to their unique sampling mechanism and noise shaping methods.
[0003] However, the digital decimation filter in the Sigma-Delta ADC is usually designed with a multi-stage cascade structure, in which the cascaded finite impulse response (FIR) has many coefficients, requires a large memory to store, has high computational complexity, and has a large signal conversion delay. Summary of the invention
[0004] In order to solve the problems existing in the related art, the embodiments of the present application provide a digital extraction filter, a data processing method and a device.
[0005] In a first aspect, the present application provides a digital decimation filter, the digital decimation filter comprising at least a connected receiving component, a first type filter and a decimation filter; wherein:
[0006] The receiving component is used to receive an audio signal and convert the received audio signal into a digital input signal;
[0007] The first type of filter is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the input signal;
[0008] The extraction filter is used to perform filtering and extraction processing on the first digital signal to obtain an output signal with a sampling frequency lower than that of the first digital signal.
[0009] In some embodiments, the decimation filter includes a second type filter and a signal optimizer connected;
[0010] The second type filter is used to downsample the first digital signal to obtain a second digital signal with a sampling frequency lower than that of the first digital signal;
[0011] The signal optimizer is used to perform signal optimization on the second digital signal to obtain the output signal.
[0012] In some embodiments, the signal optimizer includes a compensation filter and a finite impulse response filter;
[0013] The compensation filter is used to compensate for the passband attenuation of the second digital signal to obtain a compensation signal;
[0014] The finite impulse response filter is used to perform low-pass filtering on the compensation signal to obtain the output signal.
[0015] In some embodiments, the first type of filter includes a first integrator unit and a first comb unit;
[0016] The second type filter comprises a second integrator unit, a decimator unit and a second comb unit.
[0017] In some embodiments, the first type of filter is a multi-order filter, and the order of the first type of filter is determined based on a total stopband attenuation of the digital decimation filter and a stopband attenuation of the decimation filter.
[0018] In some embodiments, the finite impulse response filter in the digital decimation filter is a multi-stage cascade filter with a preset number of stages;
[0019] The preset level is determined based on the difference between the total stopband attenuation of the digital extraction filter and the stopband attenuation of the first type of filter, and the stopband attenuation of the first type of filter is determined based on the order of the first type of filter.
[0020] In some embodiments, the stopband ripple of the first type of filter is less than a preset value;
[0021] The total stop-band attenuation and the total stop-band ripple of the digital decimation filter satisfy a preset relationship, and the total stop-band ripple is determined based on the stop-band ripples of the first type filter and the decimation filter.
[0022] In a second aspect, an embodiment of the present application provides a data processing method applied to a digital decimation filter, the method comprising:
[0023] The receiving component receives the audio signal and converts the received audio signal into a digital input signal;
[0024] A first type filter is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the input signal;
[0025] The decimation filter performs filtering and decimation processing on the first digital signal to obtain an output signal with a sampling frequency lower than that of the first digital signal.
[0026] In some embodiments, the decimation filter includes a second type filter and a signal optimizer connected to each other; the decimation filter performs filtering and decimation processing on the first digital signal to obtain an output signal having a sampling frequency lower than that of the first digital signal, including:
[0027] The second type filter performs downsampling processing on the first digital signal to obtain a second digital signal with a sampling frequency lower than that of the first digital signal;
[0028] The signal optimizer performs signal optimization on the second digital signal to obtain the output signal.
[0029] In a third aspect, an embodiment of the present application provides a data processing device, comprising the above-mentioned digital decimation filter, wherein the digital decimation filter at least comprises a connected receiving component, a first type filter and a decimation filter; wherein,
[0030] The receiving component is used to receive an audio signal and convert the received audio signal into a digital input signal;
[0031] The first type of filter is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the input signal;
[0032] The extraction filter is used to perform filtering and extraction processing on the first digital signal to obtain an output signal with a sampling frequency lower than that of the first digital signal.
[0033] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 It is a structural schematic diagram of a digital decimation filter in the related art;
[0035] Figure 2 This is an optional structural diagram of a digital decimation filter provided in an embodiment of the present application. Figure 1 ;
[0036] Figure 3 is a schematic diagram of the structure of the first type of filter provided in an embodiment of the present application;
[0037] Figure 4 This is an optional structural diagram of a digital decimation filter provided in an embodiment of the present application. Figure 2 ;
[0038] Figure 5is a schematic diagram of the structure of the second type of filter provided in an embodiment of the present application;
[0039] Figure 6 It is an optional flow chart of the data processing method provided in the embodiment of the present application;
[0040] Figure 7 This is an optional structural diagram of a digital decimation filter provided in an embodiment of the present application. Figure 3 ;
[0041] Figure 8 This is a schematic diagram of the structure of the data processing device provided in the embodiment of the present application. Figure 1 ;
[0042] Fig. 9 This is a schematic diagram of the structure of the data processing device provided in the embodiment of the present application. Figure 2 . DETAILED DESCRIPTION
[0043] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0044] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict. Unless otherwise defined, all technical and scientific terms used in the embodiments of the present application have the same meaning as those commonly understood by those skilled in the art of the technical field of the embodiments of the present application. The terms used in the embodiments of the present application are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.
[0045] In related technologies, a fully functional Sigma-Delta ADC includes modules such as an anti-aliasing filter, a pre-programmable amplifier, a Sigma-Delta modulator, and a digital decimation filter. The digital decimation filter module is usually designed with a multi-stage cascade structure. Figure 1 It is a schematic diagram of the structure of the digital decimation filter in the related technology, such as Figure 1As shown, the digital decimation filter is composed of a cascaded integrator-comb (CIC) filter 101, a compensation filter 102 and an FIR filter 103. In the process of converting a pulse density modulation (PDM) signal to a pulse code modulation (PCM) signal, the CIC filter 101 completes a larger digital decimation to reduce the signal frequency, and the compensation filter is cascaded to compensate for the passband attenuation of the CIC filter. Finally, the FIR filter is cascaded to complete the remaining filtering and decimation tasks.
[0046] Among them, the CIC filter is a linear phase FIR filter. The CIC filter is an efficient implementation method of a narrowband low-pass filter, which saves area and power consumption and is used in the fields of extraction and interpolation. The compensation filter is a CIC compensation filter, which can compensate for the passband ripple of the CIC filter. The FIR filter can be a half-band filter or a FIR universal filter, etc., to complete the digital extraction task.
[0047] However, the sidelobe suppression effect of the CIC filter is not good, and the cascaded FIR filter has many coefficients, which requires a large memory to store the digital extraction filter, and the calculation complexity is high, and the signal conversion delay is large. When it is applied to voice interaction, the real-time interaction cannot be guaranteed.
[0048] Based on the problems existing in the related technology, an embodiment of the present application provides a digital extraction filter, wherein a receiving component is used to receive an audio signal and convert the received audio signal into a digital input signal; a first type of filter is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the digital input signal; and an extraction filter is used to filter and extract the first digital signal to obtain an output signal with a sampling frequency lower than the first digital signal.
[0049] In this way, the digital extraction filter provided in the present application sets a first type filter that only performs filtering but not extraction before the extraction filter, so that the noise of the digital input signal is removed before the extraction operation is performed. The preprocessing of the first type filter can ensure that the subsequent extraction filter obtains a cleaner and higher quality signal; secondly, filtering the digital input signal through the first type filter in advance can reduce the number of coefficients of the extraction filter, reduce the hardware structure, and reduce the chip area occupied by the digital extraction filter. At the same time, the reduction of the hardware structure in the extraction filter can increase the signal conversion rate, which can improve the synchronization of voice interaction in the intelligent voice scenario.
[0050] In some embodiments, the digital extraction filter provided by the present application can be set in a data processing device, and voice human-computer interaction can be realized based on the data processing device. The data processing device can be an independent physical device, placed in a home, office, vehicle or other place, and connected to the terminal through short-distance communication methods such as Bluetooth, infrared, and wired. The terminal performs human-computer interaction and performs corresponding operations through the digital signal output by the data processing device; the data processing device can also be any smart device such as a mobile phone, a laptop, a tablet computer, a TV, etc.
[0051] Here, the data processing device can be a mobile phone terminal. The mobile phone can collect the user's voice signal, convert and recognize the collected voice signal, and perform corresponding operations according to the recognition result, such as playing music, reading short messages and recording.
[0052] Figure 2 This is an optional structural diagram of a digital decimation filter provided in an embodiment of the present application. Figure 1 ,like Figure 2 As shown, the digital extraction filter at least includes a receiving component 201, a first type filter 202 and an extraction filter 203, wherein the receiving component 201 is used to receive an audio signal and convert the received audio signal into a digital input signal; the first type filter 202 is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the digital input signal; the extraction filter 203 is used to filter and extract the first digital signal to obtain an output signal with a sampling frequency lower than the first digital signal.
[0053] Here, the receiving component 201 can be a microphone equipped with a processing chip, and after receiving the audio signal, the audio signal is converted into a digital input signal, and the digital input signal can be a pulse density modulation signal (i.e., a PDM signal). The receiving component 201 can continuously receive the audio signal, and convert it into a digital input signal and send it to the first type filter.
[0054] In some embodiments, the first type filter 202 may be any filter that implements filtering without extraction, for example, a CIC filter, which performs filtering processing on the digital input signal, the first digital signal, and the signal-to-noise ratio of the first digital signal is higher than the digital input signal. Here, the CIC filter has a simple structure, only requires an adder and a delay unit, does not require a multiplier and a storage unit, occupies less resources, has a small amount of calculation, can achieve high speed, low power consumption and anti-aliasing low-pass requirements, and improves the quality and reliability of the sampled data.
[0055] Figure 3 is a schematic diagram of the structure of the first type of filter provided in an embodiment of the present application, such as Figure 3As shown, the first type filter may include a first integration unit 301 and a first comb unit 302. -1 Indicates a delay of 1 sample point, z -D represents a delay of D sampling points. The first integration unit 301 is responsible for the accumulation processing of the audio signal, and performs the integration operation on the signal by accumulating each sampling point one by one. Here, the differential equation of the first integration unit 301 can be expressed as formula (1):
[0056] y(n)=x(n)+y(n-1) (1);
[0057] Among them, y(n) is the output of the first integration unit 301 at the current moment, x(n) is the input at the current moment, and y(n-1) is the output at the previous moment.
[0058] The first comb unit 302 performs filtering but not extraction, and is composed of a plurality of delay units, adders and subtractors. The equation of the first comb unit 302 can be expressed as formula (2):
[0059] x(z)=y(n)+x(nD) (2);
[0060] Where D is the delay factor of the comb filter.
[0061] In some embodiments, the decimation filter 203 may be any filter that realizes PDM conversion to PCM, such as a digital decimation filter, and can realize filtering and decimation during the conversion process. In some embodiments, the decimation filter 203 may include a CIC filter, a CIC compensator, and a FIR filter, wherein the CIC filter is used to filter and extract the PDM signal, while reducing the sampling rate while maintaining the spectrum information of the input PDM signal; the CIC compensator is used to adjust the frequency response distortion introduced by the CIC filter, and compensates for the frequency response drop caused by the CIC filter; the FIR filter filters the impulse response value of the output signal of the CIC compensator to obtain an output signal, which may be a PCM signal.
[0062] Here, since a first type filter that only performs filtering but not extraction is set before the decimation filter, the stopband attenuation of the digital decimation filter can be increased without affecting the passband ripple of the digital decimation filter, and the number of coefficients of the decimation filter FIR filter can be reduced, thereby saving chip area and power consumption and reducing delay.
[0063] The digital decimation filter provided in the present application sets a first type filter that only performs filtering but not decimation before the decimation filter, so that the noise of the digital input signal is removed before the decimation operation is performed. The preprocessing of the first type filter can ensure that the subsequent decimation filter obtains a cleaner and higher quality signal; secondly, filtering the digital input signal through the first type filter in advance can reduce the number of coefficients of the decimation filter, reduce the hardware structure, and also reduce the chip area occupied by the digital decimation filter. At the same time, the reduction of the hardware structure in the decimation filter can increase the signal conversion rate, and can improve the synchronization of voice interaction in the scenario of intelligent voice.
[0064] In some embodiments, Figure 4 This is an optional structural diagram of a digital decimation filter provided in an embodiment of the present application. Figure 2 ,like Figure 4 As shown, the decimation filter 203 includes a second type filter 401 and a signal optimizer 402 connected to each other. The second type filter 401 is used to downsample the first digital signal to obtain a second digital signal with a sampling frequency lower than that of the first digital signal; the signal optimizer 402 is used to optimize the second digital signal to obtain an output signal.
[0065] In some embodiments, the second type filter 401 can be a CIC filter for filtering and extraction, which is used to extract data from the first digital signal sent by the first type filter 202, that is, down-sampling, to obtain a second digital signal with a sampling frequency lower than the first digital signal. The signal optimizer 402 is used to optimize the second digital signal to obtain an output signal.
[0066] Figure 5 is a schematic diagram of the structure of the second type of filter provided in an embodiment of the present application, such as Figure 5 As shown, the second type filter 401 includes a second integration unit 501 , a decimation unit 502 and a second comb unit 503 .
[0067] Here, the second integration unit 501 may be the same as the first integration unit 301 to implement an integration operation on the signal.
[0068] The decimation unit 502 and the second comb unit 503 are used to reduce the sampling rate, that is, to perform data decimation. Here, D of the decimation unit 502 is a decimation factor, which determines the decimation multiple of the second type of filter, that is, only one sample is retained for output in every D samples in the first digital signal, thereby achieving a reduction in the data rate.
[0069] Here, the second type filter 401 can perform multi-stage data extraction on the first digital signal to reduce the sampling rate of the first digital signal. Each stage of the multi-stage extraction will reduce the amount of data of the first digital signal, and multi-stage filtering can be implemented, thereby improving the quality of the output first digital signal.
[0070] After completing the CIC filtering, the second type filter 401 may also perform data type conversion on the first digital signal, for example, converting the processed signal from a single-precision floating point number to a double-precision floating point number or other required data types to ensure data accuracy and compatibility.
[0071] Here, the multi-stage extraction of the second type filter 401 has the problem of passband attenuation, so that the signal amplitude-frequency response in the passband does not meet the requirements. Therefore, the first digital signal output by the second type filter 401 can be optimized by a signal optimizer to compensate for the passband attenuation caused by the CIC filter to achieve a smooth low-ripple passband. The signal optimizer can be any compensator that can achieve compensation for passband attenuation.
[0072] The embodiment of the present application realizes data extraction and avoids passband attenuation caused by data extraction through the first digital signal and signal optimization. While realizing downsampling, it makes the passband response smoother, reduces the frequency response distortion caused by the first digital signal, and improves the signal quality of the first digital signal.
[0073] In some embodiments, the signal optimizer may include a compensation filter and a finite impulse response filter; the compensation filter is used to compensate for the signal loss in the first digital signal output by the second type filter 401 to obtain a first compensation signal, and the finite impulse response filter is used to perform low-pass filtering on the first compensation signal sent by the compensation filter to obtain an output signal.
[0074] Here, the compensation filter in the signal optimizer may be a CIC compensation filter, and the finite impulse response filter may be a FIR filter. It may also be other filters and compensators that can achieve compensation and filtering, and this application does not limit this.
[0075] Here, the CIC compensation filter can compensate for the passband attenuation caused by the CIC filter and compensate for the signal loss in the first digital signal sent by the CIC filter to achieve a relatively smooth low-ripple passband; the FIR filter can perform low-pass filtering on the signal output by the compensation filter, remove noise, and extract frequency components of interest, thereby achieving signal enhancement and reconstruction.
[0076] The embodiment of the present application cascades the CIC compensation filter and the FIR filter, which can effectively compensate for the passband attenuation problem of the second type of filter, and at the same time utilize the high precision and stability of the FIR filter to improve the filtering effect of the entire digital extraction filter.
[0077] In some embodiments, Figure 3 and Figure 5 As shown, the first type of filter includes a first integration unit and a first comb unit; the second type of filter includes a second integration unit, an extraction unit and a second comb unit.
[0078] That is to say, the first type of filter provided in the embodiment of the present application reduces the extraction unit compared to the second type of filter. Since the CIC filter does not require a multiplier and a storage unit, it can meet the requirements of high speed, low power consumption and anti-aliasing low-pass. By adding a stage of CIC filter that only filters but does not extract before the second type of filter, the stopband attenuation of the digital decimation filter can be increased without affecting the passband ripple of the digital decimation filter.
[0079] Here, the system function H(z) of the multi-stage cascaded digital decimation filter can be expressed by formula (3):
[0080] H(z)=H1(z)*H2(z)...H N (z)*H N+1 (z)(3);
[0081] Among them, H i (z) is the transfer function of each filter in the digital decimation filter, H N+ 1(z) is the transfer function of the first type of filter.
[0082] Assume that the total stopband ripple of the digital decimation filter is δ s , the stopband ripples of the first, second, and Nth order filters are δ s1 , δ s2 , δ sN Since the digital decimation filter is a series structure, the total stopband ripple δ of the digital decimation filter is s The relationship with each level can be expressed by formula (4):
[0083] δ s =δ s1 *δ s2 *...δ sN *δ s(N+1) (4);
[0084] Among them, δ s(N+1) is the stopband ripple of the first type of filter and can be a decimal less than 1.
[0085] In some embodiments, the stopband ripple of the first type of filter is less than a preset value; the total stopband attenuation and the total stopband ripple of the digital decimation filter satisfy a preset relationship, and the total stopband ripple is determined based on the stopband ripples of the first type of filter and the decimation filter. Therefore, after the first type of filter that only filters but does not decimate is added in the embodiment of the present application, since the stopband ripple of the first type of filter is less than 1, the total stopband ripple δ of the multi-stage cascade filter is s becomes smaller, that is, the stopband attenuation As increases. Here, the stopband attenuation As can be expressed by formula (5):
[0086] A s =20*log 10 δ s =20*log 10 δ s1 +20*log 10 δ s2 +…+20*log 10 δ sN +20*log 10 δ s(N+1) (5);
[0087] In some embodiments, the first type filter is a multi-order filter, and the order of the first type filter is determined based on a total stopband attenuation of the digital decimation filter and a stopband attenuation of the decimation filter.
[0088] For example, when designing a filter, the total stopband attenuation As is required to be no less than -200 decibels (dB, Decibel), such as Figure 1 The stopband attenuation provided by each stage of the digital decimation filter is as follows: total stopband attenuation As = -66dB (5th order CIC filter) + (-134dB (FIR filter)). In the embodiment of the present application, the addition of a 4th order CIC filter that only filters but does not extract can bring a stopband attenuation of -53dB. Therefore, the total stopband attenuation As of the digital decimation filter in the embodiment of the present application = (-53dB (4th order CIC filter)) + (-66dB (5th order CIC filter)) + (-81dB (FIR filter)).
[0089] Therefore, the related technology requires 50 filter coefficients to implement a -134dB FIR filter, while the embodiment of the present application requires 35 filter coefficients to implement a -81dB FIR filter. The embodiment of the present application reduces the FIR filter coefficients by adding a 4th-order CIC filter that only filters but does not extract, thereby reducing the delay and calculation amount of signal processing.
[0090] In some embodiments, the finite impulse response filter in the digital extraction filter is a multi-stage cascade filter with a preset number of stages; wherein the preset number of stages is determined based on the difference between the total stopband attenuation of the digital extraction filter and the stopband attenuation of the first type of filter, and the stopband attenuation of the first type of filter is determined based on the order of the first type of filter.
[0091] That is to say, when designing the digital decimation filter, the embodiment of the present application can calculate the order of the finite impulse response filter (FIR filter) based on the total stopband attenuation of the digital decimation filter that has been determined and the stopband attenuation of the first type of filter that has been determined to be added for filtering without decimation. That is, the order of the FIR filter is determined based on the difference between the total stopband attenuation and the stopband attenuation of the first type of filter. Since the total stopband attenuation of the digital decimation filter is determined, the order of the FIR filter can be reduced after adding the first type of filter that only filters without decimation. The FIR filter usually has a greater delay than the infinite impulse response (IIR) filter, but reducing the order of the FIR filter can reduce the delay, thereby improving the real-time processing capability of the signal; at the same time, reducing the order of the FIR filter can better balance the filtering effect and the computational complexity, thereby improving the efficiency while ensuring the performance.
[0092] Figure 6 is an optional flow chart of the data processing method provided in the embodiment of the present application. The execution subject of the data processing method provided in the embodiment of the present application is a digital extraction filter, such as Figure 6 As shown, the data processing method provided in the embodiment of the present application can be implemented through steps S601 to S603:
[0093] Step S601: The receiving component receives an audio signal and converts the received audio signal into a digital input signal.
[0094] Here, the receiving component may be a microphone equipped with a processing chip, which, after receiving the audio signal, converts the audio signal into a digital input signal, which may be a pulse density modulation signal (ie, a PDM signal).
[0095] Step S602: A first type filter is used to filter the digital input signal to obtain a first digital signal having a higher signal-to-noise ratio than the input signal.
[0096] In some embodiments, the first type filter may be any filter that only filters but does not extract, for example, may be a CIC filter, which filters the digital input signal, the first digital signal, and the signal-to-noise ratio of the first digital signal is higher than that of the digital input signal.
[0097] The first type filter may include a first integration unit and a first comb unit, and is configured to perform filtering processing on the digital input signal.
[0098] Step S603: A decimation filter performs filtering and decimation processing on the first digital signal to obtain an output signal whose sampling frequency is lower than that of the first digital signal.
[0099] In some embodiments, the decimation filter may be any filter such as a digital decimation filter that implements PDM to PCM conversion and can perform filtering and decimation during the conversion process.
[0100] In some embodiments, the extraction filter may include a CIC filter, a CIC compensator and a FIR filter, wherein the CIC filter is used to filter and extract the PDM signal, maintaining the frequency spectrum information of the input PDM signal while reducing the sampling rate; the CIC compensator is used to adjust the frequency response distortion introduced by the CIC filter, compensating for the frequency response drop caused by the CIC filter; the FIR filter filters the impulse response value of the output signal of the CIC compensator to obtain an output signal, which may be a PCM signal.
[0101] In some embodiments, the extraction filter includes a second type filter and a signal optimizer connected to each other; the extraction filter performs filtering and extraction processing on the first digital signal to obtain an output signal with a sampling frequency lower than the first digital signal, including: the second type filter performs downsampling processing on the first digital signal to obtain a second digital signal with a sampling frequency lower than the first digital signal; the signal optimizer performs signal optimization on the second digital signal to obtain the output signal.
[0102] In some embodiments, the signal optimizer includes a compensation filter and a finite impulse response filter; the signal optimizer performs signal optimization on the second digital signal to obtain the output signal, including: the compensation filter compensates for the passband attenuation of the second digital signal to obtain the compensation signal; the finite impulse response filter performs low-pass filtering on the compensation signal to obtain the output signal.
[0103] In some embodiments, the first type filter includes a first integration unit and a first comb unit; the second type filter includes a second integration unit, a decimation unit, and a second comb unit.
[0104] In some embodiments, the first type of filter is a multi-order filter, and the order of the first type of filter is determined based on a total stopband attenuation of the digital decimation filter and a stopband attenuation of the decimation filter.
[0105] In some embodiments, the finite impulse response filter in the digital extraction filter is a multi-stage cascade filter with a preset number of stages; wherein the preset number of stages is determined based on the difference between the total stopband attenuation of the digital extraction filter and the stopband attenuation of the first type of filter, and the stopband attenuation of the first type of filter is determined based on the order of the first type of filter.
[0106] In some embodiments, the stopband ripple of the first type filter is less than a preset value; the total stopband attenuation and the total stopband ripple of the digital extraction filter satisfy a preset relationship, and the total stopband ripple is determined based on the stopband ripples of the first type filter and the extraction filter.
[0107] It should be noted that the description of the data processing method in the embodiment of the present application is similar to the description of the digital decimation filter described above, and has similar beneficial effects as the digital decimation filter, so it will not be described in detail. For technical details not disclosed in the embodiment of the data processing method, please refer to the description of the digital decimation filter of the present application for understanding.
[0108] Next, an application of a digital decimation filter in a practical scenario is provided.
[0109] In order to solve the problem that the sidelobe suppression effect of the CIC filter in the related technology is poor, the cascade FIR filter has more coefficients, which makes the digital extraction filter require a larger memory to complete the storage, and the calculation complexity is high, and the signal conversion delay is large. When it is applied to voice interaction, the real-time interaction cannot be guaranteed.
[0110] An embodiment of the present application provides a digital decimation filter, in which a first-stage CIC filter is added to the digital decimation filter, which only filters but does not extract. Under the premise that the passband attenuation of the digital decimation filter is not affected, the stopband attenuation of the digital decimation filter is increased, and the number of coefficients of the FIR filter is reduced. The hardware implementation is simple, and chip area and power consumption can be saved, and delay can be reduced.
[0111] Figure 7 This is an optional structural diagram of a digital decimation filter provided in an embodiment of the present application. Figure 3 ,like Figure 7 As shown, the digital decimation filter includes a first CIC filter 701 (ie, a first type filter), a second CIC filter 702 (ie, a second type filter), a compensation filter 703 (ie, a compensation filter) and a FIR filter 704 (ie, a finite impulse response filter).
[0112] In the embodiment of the present application, a first CIC filter 701 for filtering but not decimating is added in front of the digital decimation filter provided by the related art. Since the CIC filter does not require a multiplier and a storage unit, it can meet the requirements of high speed, low power consumption and anti-aliasing low-pass. Therefore, a first CIC filter is added to the digital decimation filter for filtering but not decimating. Under the premise that the passband ripple of the system is not affected, the stopband attenuation of the system can be increased.
[0113] The system function H(z) of the multi-stage cascaded digital decimation filter provided in the embodiment of the present application can be expressed by formula (6):
[0114] H(z)=H1(z)*H2(z)…H N (z)*H N+1 (z) (6);
[0115] Among them, H i (z) is the transfer function of each filter in the digital decimation filter, H N+1 (z) is the transfer function of the first CIC filter 701 .
[0116] Assume that the total stopband ripple of the digital decimation filter is δ s , the stopband ripples of the first, second, and Nth order filters are δ s1 , δ s2 , δ sN Since the digital decimation filter is a series structure, the total stopband ripple δ of the digital decimation filter is s The relationship with each level can be expressed by formula (7):
[0117] δ s =δ s1 *δ s2 *...δ sN *δ s(N+1) (7);
[0118] Among them, δ s(N+1) is the stop-band ripple of the first CIC filter 701 and can be a decimal less than 1.
[0119] Therefore, after the first CIC filter 701 which only filters but does not extract is added in the embodiment of the present application, since the stopband ripple of the first CIC filter 701 is less than 1, the total stopband ripple δ of the multi-stage cascaded digital decimation filter is s becomes smaller, that is, the stop-band attenuation As increases.
[0120] In some embodiments, under the premise that the total passband ripple of the digital decimation filter meets the performance requirements, the stopband ripple δ of the N+1th stage is increased. s(N+1) , the total stopband ripple δ of the multi-stage cascade filter sbecomes smaller, that is, the stopband attenuation As increases. Or the total stopband ripple δ s Keep unchanged and increase the stopband ripple δ of the N+1th level s(N+1) , the stopband ripple of the cascaded FIR filter can be appropriately relaxed, and the FIR filter coefficients can be reduced, thereby reducing the delay and calculation amount.
[0121] Here, the stopband attenuation As can be expressed by formula (8):
[0122] A s =20*log 10 δ s =20*log 10 δ s1 +20*log 10 δ s2 +…+20*log 10 δ sN +20*log 10 δ s(N+1) (8);
[0123] In some embodiments, the first type filter is a multi-order filter, and the order of the first type filter is determined based on a total stopband attenuation of the digital decimation filter and a stopband attenuation of the decimation filter.
[0124] In some embodiments, when designing the filter, the total stopband attenuation As is required to be no less than -200 dB. Figure 1 The stopband attenuation provided by each stage of the digital decimation filter in the related art shown is as follows: total stopband attenuation As = -66dB (5th order CIC filter) + (-134dB (FIR filter)). However, the addition of a 4th order CIC filter that only filters but does not extract in the embodiment of the present application can bring about a stopband attenuation of -53dB. Therefore, the total stopband attenuation As of the digital decimation filter in the embodiment of the present application = (-53dB (4th order CIC filter)) + (-66dB (5th order CIC filter)) + (-81dB (FIR filter)).
[0125] Therefore, the related technology requires 50 filter coefficients to implement a -134dB FIR filter, while the embodiment of the present application requires 35 filter coefficients to implement a -81dB FIR filter. The embodiment of the present application reduces the FIR filter coefficients by adding a 4th-order CIC filter that only filters but does not extract, thereby reducing the delay and calculation amount of signal processing.
[0126] Figure 8 This is a schematic diagram of the structure of the data processing device provided in the embodiment of the present application. Figure 1 , Figure 8 The data processing device 800 shown includes a digital decimation filter 801, which includes the above-mentioned structure.
[0127] Fig. 9 This is a schematic diagram of the structure of the data processing device provided in the embodiment of the present application. Figure 2 , Fig. 9 The data processing device 900 shown may also include: at least one processor 901, a digital extraction filter 902 and a user interface 903. The various components in the data processing device are coupled together via a bus system 904. It is understood that the bus system 904 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 904 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, the bus system 904 is not described in detail. Fig. 9 Various buses are labeled as bus system 904.
[0128] The digital extraction filter 902 is used to receive an audio signal and convert the received audio signal into a digital input signal; filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the digital input signal; filter and extract the first digital signal to obtain an output signal with a sampling frequency lower than the first digital signal.
[0129] The processor 901 may be an integrated circuit chip with signal processing capability, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., wherein the general-purpose processor may be a microprocessor or any conventional processor, etc. The processor 901 is used to perform speech recognition based on the output signal output by the digital extraction filter 902, obtain the speech instruction corresponding to the audio signal, and perform speech interaction control. For example, the instruction obtained by recognizing the output signal output by the digital extraction filter 902 is to turn on the player. At this time, the processor 901 can turn on the player in the data processing device based on the instruction.
[0130] The user interface 903 comprises one or more output devices 9031 enabling presentation of media content, and one or more input devices 9032, which may be microphones for receiving audio signals.
[0131] The above are only embodiments of the present application and are not intended to limit the protection scope of the present application. Any modifications, equivalent substitutions and improvements made within the spirit and scope of the present application are included in the protection scope of the present application.
[0132] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the size of the sequence number of the above-mentioned processes does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The above-mentioned sequence numbers of the embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.
[0133] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of further restrictions, an element defined by the statement "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed.
[0134] The above are only implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A digital decimation filter, the digital decimation filter comprising at least a connected receiving component, a first type filter and a decimation filter; wherein, The receiving component is used to receive an audio signal and convert the received audio signal into a digital input signal; The first type of filter is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the input signal; The extraction filter is used to perform filtering and extraction processing on the first digital signal to obtain an output signal with a sampling frequency lower than that of the first digital signal.
2. The digital decimation filter of claim 1 , comprising a second type filter and a signal optimizer connected; The second type filter is used to downsample the first digital signal to obtain a second digital signal with a sampling frequency lower than that of the first digital signal; The signal optimizer is used to perform signal optimization on the second digital signal to obtain the output signal.
3. The digital decimation filter of claim 2, wherein the signal optimizer comprises a compensation filter and a finite impulse response filter; The compensation filter is used to compensate for the passband attenuation of the second digital signal to obtain a compensation signal; The finite impulse response filter is used to perform low-pass filtering on the compensation signal to obtain the output signal.
4. The digital decimation filter according to claim 2 or 3, wherein the first type filter comprises a first integration unit and a first comb unit; The second type filter comprises a second integrator unit, a decimator unit and a second comb unit.
5. The digital decimation filter according to claim 2 or 3, wherein the first type filter is a multi-order filter, and the order of the first type filter is determined based on the total stopband attenuation of the digital decimation filter and the stopband attenuation of the decimation filter.
6. The digital decimation filter according to claim 5, wherein the finite impulse response filter in the digital decimation filter is a multi-stage cascade filter with a preset number of stages; in, The preset number of stages is determined based on a difference between a total stopband attenuation of the digital decimation filter and a stopband attenuation of the first type of filter, wherein the stopband attenuation of the first type of filter is determined based on an order of the first type of filter.
7. The digital decimation filter according to claim 6, wherein the stopband ripple of the first type of filter is smaller than a preset value; The total stop-band attenuation and the total stop-band ripple of the digital decimation filter satisfy a preset relationship, and the total stop-band ripple is determined based on the stop-band ripples of the first type filter and the decimation filter.
8. A data processing method, applied to a digital decimation filter, the method comprising: The receiving component receives the audio signal and converts the received audio signal into a digital input signal; A first type filter is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the input signal; The decimation filter performs filtering and decimation processing on the first digital signal to obtain an output signal with a sampling frequency lower than that of the first digital signal.
9. The method according to claim 8, wherein the decimation filter comprises a second type filter and a signal optimizer connected to each other; wherein the decimation filter performs filtering and decimation processing on the first digital signal to obtain an output signal having a sampling frequency lower than that of the first digital signal, comprising: The second type filter performs downsampling processing on the first digital signal to obtain a second digital signal with a sampling frequency lower than that of the first digital signal; The signal optimizer performs signal optimization on the second digital signal to obtain the output signal.
10. A data processing device, comprising a digital decimation filter, the digital decimation filter comprising at least a connected receiving component, a first type filter and a decimation filter; wherein, The receiving component is used to receive an audio signal and convert the received audio signal into a digital input signal; The first type of filter is used to filter the digital input signal to obtain a first digital signal with a higher signal-to-noise ratio than the input signal; The extraction filter is used to perform filtering and extraction processing on the first digital signal to obtain an output signal with a sampling frequency lower than that of the first digital signal.
