Audio controller for semi-adaptive active noise cancellation devices

Abstract

An audio controller for an active noise reduction (ANR) device is disclosed for reducing an ambient noise signal. The audio controller includes a processing circuit for providing a fixed ANR filter (520), the fixed ANR filter for generating a first noise reduction signal. In addition, the processing circuit is also used to provide an adaptive ANR filter (521), wherein the adaptive ANR filter (521) includes one or more adjustable filter coefficients for adjusting the adaptive ANR filter, and the adaptive ANR filter is used to generate a second noise reduction signal. The processing circuit is also used to generate a total noise reduction signal as an adjustable weighted linear combination (522, 523, 524) of the first noise reduction signal and the second noise reduction signal.

Description

Technical Field

[0001] The present invention generally relates to audio processing. More specifically, the present invention relates to an audio controller for a semi-adaptive active noise reduction device and a corresponding active noise reduction method. In addition, the present invention relates to a semi-adaptive active noise reduction device, such as a headset, comprising such an audio controller. Background Art

[0002] Headphones are often worn in public environments where there may be loud ambient noise around the users of such headphones. Therefore, active noise cancellation (ANC) / active noise reduction (ANR) headphones are becoming increasingly important as they can effectively attenuate the perceived ambient noise.

[0003] ANC / ANR headphones attempt to reduce unwanted (noise) signals by actively compensating using an additional sound source provided by the headphone speakers. Unwanted signals are usually defined as the dominant sound sources located in the user's environment. The noise signal (primary source) is recorded by a microphone installed in the headphone to calculate a corresponding anti-noise or compensation signal, which is then played back through the headphone speakers (secondary source). Ideally, the anti-noise and unwanted noise should have the same amplitude and opposite phase at the listening position.

[0004] ANC / ANR headphones can generally be divided into three types: feedforward (FF), feedback (FB) and hybrid (FF+FB) ANC / ANR headphones. In addition, each type can be implemented using a fixed filter (offline calculation) or an adaptive filter (online interactive calculation).

[0005] For feedforward (FF) ANC / ANR, a microphone mounted on the outside of the ear cup captures the noise. The noise is then processed and anti-noise is created through ANC / ANR data processing, and the resulting signal is then sent to the headphone speaker. In the case of adaptive FF ANC / ANR devices, another microphone mounted inside the ear cup is used to check the residual error signal and further adjust the FF controller. For feedback (FB) ANC / ANR, a microphone mounted inside the ear cup and in front of the headphone transducer captures the resulting signal in the same way as the listener (at least at low frequencies). The recorded signal is filtered by a designed controller to generate an anti-noise signal. Hybrid ANC / ANR is a hybrid approach that combines FF and FB ANC / ANR filters.

[0006] Fixed ANC / ANR filters are commonly used in commercially available ANC / ANR headphones due to their robustness. However, fixed ANC / ANR filters cannot adapt to changes that occur in dynamic environments (changes in the position of the headphone relative to the head, source direction, source type, etc.), resulting in a degradation of the ANC / ANR performance. In contrast, adaptive ANC / ANR filters can be used to adaptively compensate for noise, even in dynamic environments. However, sudden changes in direction or noise type and the secondary path (transfer function between the headphone transducer to the error microphone) may cause large estimation errors. The robustness of adaptive filters cannot always be guaranteed. Therefore, a combination of fixed and adaptive filters is needed to effectively attenuate noise in dynamic environments while ensuring the robustness of ANC / ANR performance. Summary of the invention

[0007] An object of the present invention is to provide an improved audio controller for combining fixed ANC / ANR filters and adaptive ANC / ANR filters in a semi-adaptive ANC / ANR device, and a corresponding ANC / ANR device and method.

[0008] The above and other objects are achieved by the subject matter claimed in the independent claims. Other implementations are apparent from the dependent claims, the description and the drawings.

[0009] Generally, embodiments disclosed herein provide an audio controller for a semi-adaptive ANC / ANR device (e.g., headphones) for combining a fixed ANC / ANR filter and an adaptive ANC / ANR filter in an adaptive manner. In one embodiment, the fixed ANC / ANR filter and the adaptive ANC / ANR filter are weighted in real time using interactively updated weighting factors to achieve high attenuation performance while maintaining the robustness of the ANC / ANR performance. In one embodiment, the ANC / ANR performance of the adaptive filter is compared with the overall ANC / ANR performance in real time. In one embodiment, the adaptive filter is used only when the estimated ANC / ANR performance of the adaptive filter is better than the performance of the overall system. In one embodiment, the weights of the fixed filter and the adaptive filter are adjusted to output an improved anti-noise signal.

[0010] More specifically, according to a first aspect, an audio controller for an active noise reduction (ANR) device is disclosed for reducing an ambient noise signal. In one embodiment, the ANR device may be an active noise cancellation (ANC) device for cancelling an ambient noise signal.

[0011] The audio controller includes a processing circuit, which is used to provide (i.e., implement) at least one fixed ANR filter, wherein the fixed ANR filter includes multiple fixed filter coefficients and is used to generate a first noise reduction signal. In addition, the processing circuit is used to provide (i.e., implement) at least one adaptive ANR filter, wherein the adaptive ANR filter includes one or more adjustable filter coefficients for adjusting the adaptive ANR filter, and the adaptive ANR filter is used to generate a second noise reduction signal. The processing circuit is also used to generate a total noise reduction signal as an adjustable weighted linear combination (i.e., sum) of the first noise reduction signal and the second noise reduction signal. Therefore, an improved audio controller for combining a fixed ANC / ANR filter and an adaptive ANC / ANR filter in a semi-adaptive ANC / ANR device is provided, which supports achieving an optimal balance between high attenuation performance and robustness / stability of the ANC / ANR device.

[0012] In another possible implementation of the first aspect, the processing circuit is further configured to determine a noise reduction estimate of a reduction in the ambient noise signal caused by the second noise reduction signal, and determine the adjustable weighted linear combination based on the noise reduction estimate.

[0013] In another possible implementation of the first aspect, the processing circuit is used to generate the total noise reduction signal based on the adjustable weighting (i.e., gain) coefficient a as the adjustable weighted linear combination (i.e., sum) of the first noise reduction signal and the second noise reduction signal.

[0014] In another possible implementation of the first aspect, the processing circuit is used to generate the total noise reduction signal as the adjustable weighted linear combination (i.e., the sum) of the first noise reduction signal and the second noise reduction signal based on the following equation:

[0015] y=ay fixed +(1-a)y adap ,

[0016] Among them, y fixed represents the first noise reduction signal provided by the fixed ANR filter, y adap represents the second noise reduction signal provided by the adaptive ANR filter, and y represents the total noise reduction signal.

[0017] In another possible implementation manner of the first aspect, the processing circuit is used to determine the adjustable weighting (ie, gain) coefficient a based on the following equation:

[0018]

[0019] Where MM(x) represents the magnitude measure of the parameter vector x, n represents the time sample index, e(n) represents the total residual noise signal, and e ′ (n) represents the fractional residual noise signal.

[0020] In another possible implementation manner of the first aspect, the processing circuit is configured to determine the adjustable weighting (ie, gain) coefficient a using a root mean square as the amplitude metric MM(x).

[0021] In another possible implementation of the first aspect, the processing circuit is used to estimate a secondary path transfer function G'(z), wherein the secondary path transfer function G'(z) describes a modification of the total residual noise signal e(n) that results in the generation of the fractional residual noise signal e'(n), and the processing circuit is further used to estimate the secondary path transfer function G'(z) based on the second noise reduction signal y adap , the total residual noise signal e(n) and the secondary path transfer function G'(z) determine the fractional residual noise signal e'(n).

[0022] In another possible implementation of the first aspect, the processing circuit is also used to continuously adjust the adjustable weighted linear combination (i.e., the sum) of the first noise reduction signal and the second noise reduction signal to generate the total noise reduction signal, wherein initially the total noise reduction signal is equal to the first noise reduction signal provided by the fixed ANR filter.

[0023] In another possible implementation of the first aspect, the processing circuit is further used to adjust the one or more adjustable filter coefficients of the adaptive ANR filter based on the total residual noise signal.

[0024] According to a second aspect, an active noise reduction (ANR) device, in particular an active noise cancellation (ANC) device, is provided, wherein the ANR device comprises an audio controller and a speaker according to the first aspect, wherein the speaker is driven based on the total noise reduction signal generated by the audio controller. For example, the ANR device may be a headphone device, an earphone device, etc.

[0025] In another possible implementation of the second aspect, the ANR device includes a feedforward (FF) microphone for detecting the ambient noise signal, wherein the fixed ANR filter is used to generate the first noise reduction signal based on the ambient noise signal, and the adaptive ANR filter is used to generate the second noise reduction signal based on the ambient noise signal. Therefore, the ANR device according to the second aspect can be implemented as a fixed FF ANR device.

[0026] In another possible implementation of the second aspect, the ANR device includes a feedback (FB) microphone for detecting a total residual noise signal, wherein the fixed ANR filter is used to generate the first noise reduction signal based on the total residual noise signal, and the adaptive ANR filter is used to generate the second noise reduction signal based on the total residual noise signal. Therefore, the ANR device according to the second aspect can be implemented as a fixed FB ANR device or an adaptive FB ANR device.

[0027] In another possible implementation of the second aspect, the ANR device includes a feedforward (FF) microphone for detecting the ambient noise signal and a feedback (FB) microphone for detecting the total residual noise signal, wherein the fixed ANR filter is used to generate the first noise reduction signal based on the ambient noise signal and the total residual noise signal, and the adaptive ANR filter is used to generate the second noise reduction signal based on the ambient noise signal and the total residual noise signal. Therefore, the ANR device according to the second aspect can be implemented as a fixed or adaptive hybrid ANR device or an adaptive FF ANR device.

[0028] According to a third aspect, an active noise reduction (ANR) method for reducing an ambient noise signal is provided. The ANR method comprises the following steps:

[0029] generating a first noise reduction signal using a fixed ANR filter;

[0030] generating a second noise-reduced signal using an adaptive ANR filter, wherein the adaptive ANR filter comprises one or more adjustable filter coefficients for adjusting the adaptive ANR filter;

[0031] An overall noise-reduced signal is generated as an adjustable weighted linear combination (ie, sum) of the first noise-reduced signal and the second noise-reduced signal.

[0032] The ANR method according to the third aspect of the present invention can be performed by the audio controller according to the first aspect or the ANR device according to the second aspect of the present invention. Therefore, other features of the ANR method according to the third aspect of the present invention come directly from the functions of the audio controller according to the first aspect and the ANR device according to the second aspect of the present invention, as well as their different implementation methods described above and below.

[0033] According to a fourth aspect, a computer program product is provided, comprising a non-transitory computer-readable storage medium for storing program code, wherein when the program code is executed by the computer or the processor, the program code causes the computer or the processor to perform the ANR method according to the third aspect.

[0034] The following drawings and description set forth one or more embodiments in detail. Other features, objects, and advantages are apparent in the description, drawings, and claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the accompanying drawings:

[0036] Figure 1 is a schematic diagram of a signal processing architecture including an audio controller for a fixed FF ANR / ANC device;

[0037] Figure 2 is a schematic diagram of a signal processing architecture including an audio controller for an adaptive FF ANR / ANC device;

[0038] Figure 3 is a schematic diagram of a signal processing architecture including an audio controller for a fixed FF ANR / ANC device;

[0039] Figure 4 is a schematic diagram of a signal processing architecture including an audio controller for an adaptive FB ANR / ANC device;

[0040] Figure 5 is a schematic diagram of a signal processing architecture including an audio controller for a semi-adaptive FF ANR / ANC device according to one embodiment;

[0041] Figure 6 is a schematic diagram of a signal processing architecture including an audio controller for a semi-adaptive FB ANR / ANC device according to one embodiment;

[0042] Figure 7 is a schematic diagram of a signal processing architecture including an audio controller for a semi-adaptive hybrid ANR / ANC device according to one embodiment;

[0043] Figure 8 is a flow chart of an ANR / ANC method provided by an embodiment;

[0044] Fig. 9 shows the magnitude and phase of the secondary path transfer function as a function of frequency for various embodiments;

[0045] Fig.10aA graph showing the performance of a semi-adaptive FF ANR / ANC device provided by one embodiment;

[0046] Fig.10b A graph showing the performance of a semi-adaptive FB ANR / ANC device provided by one embodiment;

[0047] Fig.10c A graph showing the performance of a semi-adaptive hybrid ANR / ANC device provided by one embodiment.

[0048] In the following, identical reference signs refer to identical or at least functionally equivalent features. DETAILED DESCRIPTION

[0049] In the following description, reference is made to the accompanying drawings that form a part of the present invention, which illustrate specific aspects of embodiments of the present invention or specific aspects in which embodiments of the present invention may be used by way of illustration. It should be understood that embodiments of the present invention may be used in other aspects and include structural or logical changes that are not depicted in the accompanying drawings. Therefore, the following detailed description should not be understood in a restrictive sense, and the scope of the present invention is defined by the appended claims.

[0050] For example, it should be understood that the disclosure related to describing a method may also be applicable to a corresponding device or system for performing the method, and vice versa. For example, if one or more specific method steps are described, the corresponding device may include one or more units (e.g., functional units) to perform the one or more method steps described (e.g., one unit performs one or more steps, or multiple units perform one or more of the multiple steps respectively), even if the one or more units are not explicitly described or illustrated in the drawings. On the other hand, for example, if a specific device is described based on one or more units (e.g., functional units), the corresponding method may include a step to perform the function of one or more units (e.g., one step performs the function of one or more units, or multiple steps perform the function of one or more of the multiple units respectively), even if the one or more units are not explicitly described or illustrated in the drawings. In addition, it should be understood that, unless otherwise explicitly stated, the features of the various exemplary embodiments and / or aspects described herein may be combined with each other.

[0051] Before describing different embodiments in detail, some terminology and technical background about active noise reduction (ANR) / active noise cancellation (ANC) will be introduced below. Figure 11 is a schematic diagram of a signal processing architecture 100 including an audio controller for a fixed FF ANR / ANC device (i.e., an ANR / ANC device with a fixed FF filter). In the case of fixed FF ANR / ANC, a noise signal is recorded by an external microphone (reference microphone), a noise cancellation signal is generated by the audio controller, and the calculated anti-noise signal is played through a speaker. Figure 1 , P(z) 110 represents the primary path, which represents the acoustic transfer function between the external (reference) microphone and the internal (error) microphone. G(z) 130 represents the secondary path, which consists of the transfer function between the loudspeaker and the internal (error) microphone. W(z) 120 represents the FF controller 120 for generating the acoustic cancellation signal. The purpose is to suppress the internal noise signal d(n), that is, the external noise signal x(n), which has been modified by the acoustic transfer function of the primary path 110, for example, transmitted and attenuated by the earphone housing. The combination 140a of the internal noise signal d(n) and the acoustic cancellation signal (modified by the acoustic transfer function of the secondary path 130) produces a residual noise signal or error signal e(n). In general, the causal approximation of the FF controller 120 can be obtained by solving the Wiener-Hopf equation:

[0052]

[0053] Among them, gg The autocorrelation matrix describing the impulse response of G(z)130, φ pg represents the cross-correlation vector between the impulse responses of P(z) 110 and G(z) 130.

[0054] Figure 22 is a schematic diagram of a signal processing architecture 200, which includes an audio controller for an adaptive FF ANR / ANC device, i.e., an ANR / ANC device with an adaptive FF filter, such as based on a filtered-x least mean square (FxLMS) algorithm. The audio controller 220 is updated in real time based on the recorded noise signal x(n) and the error signal e(n). The recorded noise signal is first filtered by an estimated secondary path G'(z)230, which is an approximation of the acoustic transmission path (i.e., secondary path G(z)230) between the earphone speaker and the internal microphone. Afterwards, the filtered input signal and the error signal are fed into the LMS processing block 250 for determining the transfer function of the audio controller W(z)220. The recorded input signal is filtered by the estimated controller and reproduced by the speaker. The combination 240a of the internal noise signal d(n) (generated by the acoustic transfer function of the primary path 210 modifying the noise signal x(n)) and the acoustic cancellation signal (modified by the acoustic transfer function of the secondary path 230) produces a residual noise signal or error signal e(n). The FxLMS algorithm implemented by the LMS processing block 250 can be expressed as:

[0055] e(n)=d(n)-g T (n)[w T (n)x(n)]

[0056] w(n+1)=w(n)-μ[g ′T (n)x(n)]e(n)

[0057] Where n represents the time index, g(n) and g'(n) are the actual impulse response and measured impulse response of the secondary path, respectively, and w(n) = [w0(n), w1(n), …, w L-1 (n)] is the coefficient of the controller W(z) 220 with a filter order of L, and x(n) = [x(n), x(n-1), x(n-2), ... x(n-L+1)] is a recorded signal vector consisting of the last L samples. Instead of the FxLMS algorithm, the LMS processing block 250 can implement other adaptive algorithms based on the FxLMS algorithm, such as leakage FxLMS, ​​FxNLMS, band-limited FxLMS, ​​adaptive algorithms based on Kalman filtering, etc.

[0058] Figure 33 is a schematic diagram of a signal processing architecture 300 including an audio controller for a fixed FB ANR / ANC device (i.e., an ANR / ANC device with a fixed FB filter). In a FB ANR / ANC device, the error signal e(n) is used to synthesize a reference signal and further generate an anti-noise signal. Compared to a FF ANC device, an external microphone is no longer required to obtain a reference noise signal. Figure 3 In the signal processing architecture 300 shown, the FB controller W(z) 320 is designed based on the secondary path G(z) 330. The combination 340a of the internal noise signal d(n) (generated by the acoustic transfer function of the main path 310 modifying the noise signal x(n)) and the acoustic cancellation signal y(n) (modified by the acoustic transfer function of the secondary path 330) produces a residual noise signal or error signal e(n). The transfer function / sensitivity function S(z) between the desired signal d(n) and the error signal e(n) is minimized to achieve high attenuation performance. On the other hand, the complementary sensitivity function T(z)=1-S(z) represents the robustness of the entire system and should be small compared to the measurement noise. Since it is usually impossible to achieve both criteria at the same time, a trade-off is usually made between performance and robustness. To this end, the fixed controller 320 may be optimized taking both criteria into account using a mixed sensitivity H-Inf synthesis method (e.g., as described in Skogestad, S. and Postlethwaite, I. (2007), “Multivariable feedback control: analysis and design”, Vol. 2, Wiley New York).

[0059] Figure 4 4 is a schematic diagram of a signal processing architecture 400 including an audio controller for an adaptive FB ANR / ANC device, i.e., an ANR / ANC device with an adaptive FB filter, such as based on a filtered-x least mean square (FxLMS) algorithm. The basic idea of ​​the adaptive FB ANR / ANC device is to predict a reference signal of the audio controller 420. The synthesized reference signal x syn It can be calculated as:

[0060] x syn =e(n)+[g ′T (n)y(n)]

[0061] Under ideal conditions, ie, g'(n) = g(n), the adaptive FB ANR / ANC device is equivalent to the adaptive FF ANR / ANC device. The FxLMS-based adaptive filtering algorithm can be used to obtain the FB audio controller 420, as described above in Figure 2 The present invention is described in the context of adaptive FF ANR / ANC devices.

[0062] In the following, reference will be made to Figure 5 , Figure 6 , Figure 7 and Figure 8 Several embodiments of the present invention are described in more detail. As will be appreciated below, the embodiments disclosed herein combine fixed ANR / ANC filters and adaptive ANR / ANC filters using adjustable weighting factors to generate an improved anti-noise signal. This can improve ANR / ANC performance while maintaining the robustness and stability of semi-adaptive ANR / ANC devices. As described in detail below, this approach can be implemented in FF, FB, and even hybrid ANR / ANC devices.

[0063] Figure 5 FIG. 5 is a schematic diagram of a signal processing architecture 500 including an audio controller for a semi-adaptive FF ANR / ANC device according to one embodiment. Figure 5 In an embodiment of the present invention, the audio controller includes a processing circuit for providing (ie, implementing) a fixed ANR / ANC filter (W fixed (z)) 520, wherein the fixed ANR / ANC filter 520 includes a plurality of fixed filter coefficients and is used to generate a first noise reduction signal y fixed (n) as an output signal. In addition, the processing circuit of the audio controller is used to provide (ie, implement) an adaptive ANR / ANC filter (W adap (z)) 521, wherein the adaptive ANR / ANC filter 521 includes one or more adjustable filter coefficients for adjusting the adaptive ANR / ANC filter 521, and the adaptive ANR / ANC filter 521 is used to generate a second noise reduction signal y adap The processing circuit of the audio controller is also used to generate a total noise reduction signal y(n) as the first noise reduction signal y fixed (n) and the second noise reduction signal y adap In one embodiment, the processing circuit is used to calculate the linear combination (i.e., sum) of (n) based on the adjustable weighting factor a using the following equation (given by Figure 5 The processing block 524 shown in FIG. 5 is implemented to generate the total noise reduction signal y(n):

[0064] y(n)=ay fixed (n)+(1-a)yadap (n).

[0065] like Figure 5 As shown, the audio controller may implement weighting blocks 522 and 523 for weighting the output signal y of the fixed ANR / ANC filter 520 by factor a. fixed (n) is weighted and the output signal y of the adaptive ANR / ANC filter 521 is adjusted by a factor (1-a) adap (n) Weighting.

[0066] exist Figure 5 In the embodiment of FIG. 5 , the combination 540a of the internal noise signal d(n) (generated by modifying the noise signal x(n) by the acoustic transfer function P(z) of the primary path 510) and the total noise reduction signal, i.e., the acoustic cancellation signal y(n) (modified by the acoustic transfer function G(z) of the secondary path 230) produces a residual noise signal or error signal e(n). The recorded error signal e(n) is used to estimate (i.e., simulate) the desired error signal (d'(n)) using the total noise reduction signal y(n) and the average (simulated) secondary path (G'(z)). In FIG. Figure 5 In the embodiment shown, the output signal of the adaptive filter 521, that is, the second noise reduction signal y adap (n) is calculated by applying the FxLMS algorithm (implemented by LMS processing block 550) based on the simulated desired error signal (d'(n)), the simulated secondary path (G'(z)), and the input signal (x(n)). The simulated residual error signal (e'(n)) is calculated based on y adap (n), d'(n) and G'(z) are obtained.

[0067] exist Figure 5 In the illustrated embodiment, the weighting factor a can be determined based on the RMS of the simulated error signal and the actual measured RMS to adjust the weighting of the outputs of the fixed filter 520 and the adaptive filter 521. In one embodiment, the processing circuit can be used to determine the adjustable weighting factor a based on the following equation:

[0068]

[0069] Wherein, RMS(x) represents the root mean square measure of the parameter vector x, n represents the time sample index, e(n) represents the total residual noise signal, and e′(n) represents the fractional residual noise signal. It should be understood that other amplitude measures may be used instead of the root mean square measure RMS(x).

[0070] It should be understood that, based on the above equation, the processing circuit adjusts the weighting factor a, in particular decreases it, only when the ANR / ANC performance of the adaptive filter 521 is higher than the current overall ANR / ANC performance. If the performance of the adaptive filter 521 is worse than the overall ANR / ANC performance (i.e., the current combination of the fixed ANC filter and the adaptive ANC filter), the weighting factor a is reset to an initial value of 1 to ensure the stability / robustness of the ANR / ANC device. Therefore, in one embodiment, the processing circuit is configured to continuously adjust the first noise reduction signal y fixed (n) and the second noise reduction signal y adap (n) to generate a total noise reduction signal y(n), wherein the initial total noise reduction signal y(n) is equal to the first noise reduction signal y fixed (n).

[0071] Figure 6 FIG. 1 shows a method for semi-adaptive FB ANR / ANC device according to one embodiment. Figure 5 A variation of the signal processing architecture 500. Figure 6 It can be seen that the recorded error signal e(n) is used to generate the output signal (i.e., the total noise reduction signal y(n) = ay fixed (n)+(1-a)y adap (n)) and the averaged or simulated secondary path (G'(z) 530') simulates the desired error signal (d'(n)) and the reference signal (x syn For example, by using a simulated desired error signal (d'(n)), a simulated secondary path (G'(z) 530'), and a synthesized input signal (x syn (n)) Apply the FxLMS algorithm to calculate the output signal y adap (n). The simulated residual error signal (e'(n)) is based on y adap (n), d'(n) and G'(z) are obtained. The weighting factor a can be obtained by the processing circuit in the same way as described above for Figure 5 The FF ANR / ANC embodiment is determined in the same manner as described.

[0072] Figure 7 A method for a semi-adaptive hybrid ANR / ANC device is shown according to one embodiment. Figure 5 and Figure 6 Another variation of the signal processing architecture 500. Figure 7 It can be seen that in this embodiment, the signal processing architecture 500 includes a fixed FF filter (W fixed,FF (z) 520a), fixed FB filter (W fixed,FB (z) 520b), adaptive FF filter (W adap,FF(z) 521a) and adaptive FB filter (W adap,FB (z) 521b). In this embodiment, the output signals of the two fixed filters 520a and 520b (fixed FF and FB) are summed by processing block 526 and weighted by processing block 522 using the adjusted weighting factor a. For example, the output signals of the two adaptive filters 521a, 521b based on the FxLMS algorithm implemented by processing block 550 are also summed by processing block 528 and weighted by processing block 523 using the factor 1-a. The recorded error signal e(n) is used to generate the output signal (y(n) = ay fixed (n)+(1-a)y adap (n)) and the average secondary path (G'(z) 530) simulates the desired error signal (d'(n)) and the reference signal (x syn The output signals of the FF adaptive filter 521a and the FB adaptive filter 521b are processed by the processing circuit based on the simulated desired error signal (d'(n)), the simulated secondary path (G'(z) 530') and the synthesized (x syn (n), for FB filter 521b) or the recorded input signal (x(n), for FF filter 521a). The simulated residual error signal (e'(n)) is based on y adap (n), d'(n) and G'(z) are obtained. The weighting factor a can be obtained by the processing circuit in the same way as described above for Figure 5 The FF ANR / ANC embodiment is determined in the same manner as described.

[0073] In one embodiment, the processing circuitry of the audio controller may be configured to iteratively adjust the weighting factor a using the following processing stages:

[0074] (1) Initially, a is set to 1 to ensure the robustness / stability of ANR / ANC processing.

[0075] (2) The recorded error signal e(n) is used to simulate the desired error signal (d'(n)) using the output signal (ie, the total noise-reduced signal y(n)) and the average (simulated) secondary path (G'(z) 530').

[0076] (3) The output signal is calculated by applying an adaptive filtering method based on the simulated desired error signal (d'(n)), the simulated secondary path (G'(z) 530') and the input signal (x(n), for FF ANR / ANC).

[0077] (4) Therefore, a simulated residual error signal (e′(n)) is obtained based on the outputs of the adaptive filters 521 , 521 a , 521 b , d′(n) and G′(z) 530 .

[0078] (5) Calculate the root mean square error (RMS) of the simulated error signal (RMS(e'(n))) and compare it with the actual measured root mean square error (RMS(e(n))).

[0079] (6) Calculate the weighting factor a to adjust the weighting of one or more fixed filters 520, 520a, 520b and one or more adaptive filters 521, 521a, 521b.

[0080] In one embodiment, Figure 5 , Figure 6 and Figure 7 The signal processing architecture 500 shown in FIG. 5 may be implemented by an ANR / ANC device such as a headset, wherein the ANR / ANC device further includes a speaker for generating a total noise reduction signal y(n)=ay based on the signal processing architecture 500. fixed (n)+(1-a)y adap (n) Drive.

[0081] In one embodiment, the ANR device may be a fixed FF ANR device including a feed-forward microphone for detecting an ambient noise signal, wherein a fixed ANR filter 520 is used to generate a first noise reduction signal y based on the ambient noise signal x(n) fixed (n), and the adaptive ANR filter 521 is used to generate a second noise reduction signal y based on the environmental noise signal x(n). adap (n).

[0082] In one embodiment, the ANR device may be a fixed or adaptive FB ANR device including a feedback microphone for detecting a total residual noise signal, wherein the fixed ANR filter 520 is used to generate a first noise reduction signal y based on the total residual noise signal. fixed (n), and the adaptive ANR filter 521 is used to generate a second noise reduction signal y based on the total residual noise signal adap (n).

[0083] In one embodiment, the ANR device may be a fixed or adaptive hybrid ANR device or an adaptive FF ANR device, comprising a feedforward microphone for detecting an ambient noise signal and a feedback microphone for detecting a total residual noise signal, wherein a fixed ANR filter (520; 520a, 520b) is used to generate a first noise reduction signal y based on the ambient noise signal and the total residual noise signal fixed (n), and the adaptive ANR filter (521; 521a, 521b) is used to generate a second noise reduction signal y based on the environmental noise signal and the total residual noise signal adap (n).

[0084] Figure 8 is a flow chart of an ANR / ANC method 800 for reducing an ambient noise signal provided by an embodiment. The method 800 comprises a first step 801: using one or more fixed ANR filters 520, 520a, 520b to generate a first noise reduction signal y fixed (n). In addition, the method 800 comprises another step 803: using one or more adaptive ANR filters 521, 521a, 521b to generate a second noise reduction signal y adap (n), wherein the one or more adaptive ANR filters 521, 521a, 521b include one or more adjustable filter coefficients for adjusting the one or more adaptive ANR filters 521, 521a, 521b. The method 800 includes another step 805: generating a total noise reduction signal y(n) as a first noise reduction signal y, for example based on the following equation fixed (n) and the second noise reduction signal y adap Adjustable weighted linear combination 522, 523, 524 of (n):

[0085] y(n)=ay fixed (n)+(1-a)y adap (n).

[0086] In order to test the improved ANR / ANC performance of the embodiments disclosed in this article, the following exemplary scenarios were selected. The ambient noise (x(n)) is a 10s long mixed signal consisting of broadband white noise and sinusoidal signals at 100Hz, 500Hz and 900Hz. The main path (P(z)) is measured for a 90° relative angle between the earphone and the sound source. On the dummy head KU-100, 10 different secondary paths (G(z)) are measured with different mounting positions (seats), including two special cases, namely half on the ear and half on the table. Fig. 9 The magnitude and phase response of the measured secondary path transfer function G(z) as a function of frequency are shown (dashed line). The results are different due to different headphone mounting positions. The corresponding solid line represents the average value of the measured results and can be further used to design the FF filter and the FB filter. As described above, the embodiments disclosed herein can provide high performance of an adaptive ANC system while maintaining the robustness / stability of a fixed ANC system in dynamic scenarios. Fig. 9 In the simulation shown, the dynamic scenario is considered as different headphone sockets, i.e. the time-varying secondary path: G(z) changes every 2 seconds during the 10-second long simulation. The overall time-varying attenuated noise level is calculated based on the RMS values ​​of the residual error (e(n)) and the expected error (d(n)).

[0087] Fig.10a , Fig.10b and Fig.10c A graph showing the noise level reduction achieved by a semi-adaptive FF ANR / ANC device, a semi-adaptive FB ANR / ANC device provided by an embodiment, and an embodiment of a semi-adaptive hybrid ANR / ANC device is shown. For each device, the results of a fixed filter (dashed line), an adaptive filter (dashed line), and a semi-adaptive filter (solid line) are calculated. G1-G5 represent five different secondary paths, where G1, G3, and G5 are measured under normal assembly conditions (the earphones are placed on the head), and G2 and G4 are measured when the earphones are placed on the table or half on the ears (special cases). It can be seen that the fixed ANR / ANC device provides stable attenuation performance under different secondary paths. In addition, there is no large estimation error when changing the secondary path. In contrast, the adaptive ANR / ANC device exhibits better ANR / ANC performance than the fixed ANR / ANC device after the required adaptation time, but a large estimation error caused by the sudden change of the secondary path can be observed. Compared with fixed ANR / ANC devices and adaptive ANR / ANC devices, it can be clearly seen that the semi-adaptive ANR / ANC device embodiments disclosed in this article exhibit better ANR / ANC performance than fixed ANR / ANC devices, and more stable performance than adaptive ANR / ANC devices.

[0088] Those skilled in the art will understand that the "blocks" ("units") in the various figures (methods and devices) represent or describe the functions of an embodiment of the present invention (and not necessarily independent "units" in hardware or software), thereby equally describing the functions or features of the device embodiments and the method embodiments (unit = step).

[0089] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the described embodiments of the device are merely exemplary. For example, unit division is only a logical function division, and other division methods may be used in actual implementation. For example, multiple units or components may be merged or integrated into another system, or some features may be ignored or not performed. In addition, the mutual coupling or direct coupling or communication connection shown or described may be implemented through some interfaces. The direct coupling or communication connection between devices or units may be implemented in electronic, mechanical or other forms.

[0090] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one location, or may be distributed over multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiment.

[0091] In addition, the functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

Claims

1. An audio controller for an active noise reduction (ANR) device, characterized in that: For reducing an ambient noise signal, wherein the audio controller comprises a processing circuit, the processing circuit being used for: Providing a fixed ANR filter (520; 520a, 520b), wherein the fixed ANR filter (520; 520a, 520b) is used to generate a first noise reduction signal; Providing an adaptive ANR filter (521; 521a, 521b), wherein the adaptive ANR filter (521; 521a, 521b) comprises one or more adjustable filter coefficients for adjusting the adaptive ANR filter (521; 521a, 521b), and the adaptive ANR filter (521; 521a, 521b) is used to generate a second noise reduction signal; generating a total noise reduction signal based on an adjustable weighting coefficient a as an adjustable weighted linear combination of the first noise reduction signal and the second noise reduction signal (522, 523, 524); The processing circuit is used to determine the adjustable weighting coefficient a based on the following equation: Where MM(x) represents the magnitude measure of the parameter vector x, n represents the time sample index, e(n) represents the total residual noise signal, and e'(n) represents the fractional residual noise signal.

2. The audio controller according to claim 1, characterized in that: The processing circuit is further configured to determine a noise reduction estimate of a reduction in the ambient noise signal caused by the second noise reduction signal, and to determine the adjustable weighted linear combination based on the noise reduction estimate (522, 523, 524).

3. The audio controller according to claim 1, characterized in that: The processing circuit is configured to generate the total noise reduction signal as the adjustable weighted linear combination (522, 523, 524) of the first noise reduction signal and the second noise reduction signal based on the following equation: y=yes fixed +(1-a)y adap , Among them, y fixEd represents the first noise reduction signal, y ADap represents the second noise reduction signal, and y represents the total noise reduction signal.

4. The audio controller according to claim 1, characterized in that: The processing circuit is configured to determine the adjustable weighting coefficient a using a root mean square as the amplitude metric MM(x).

5. The audio controller according to claim 1, characterized in that: The processing circuit is configured to estimate a secondary path transfer function, wherein the secondary path transfer function describes a modification of the total residual noise signal that results in the fractional residual noise signal, and the processing circuit is further configured to determine the fractional residual noise signal based on the second noise reduction signal, the total residual noise signal, and the secondary path transfer function.

6. The audio controller according to any one of the preceding claims, characterized in that The processing circuit is configured to continuously adjust the adjustable weighted linear combination (522, 523, 524) of the first noise reduction signal and the second noise reduction signal to generate the total noise reduction signal, wherein initially the total noise reduction signal is equal to the first noise reduction signal.

7. The audio controller according to any one of the preceding claims, characterized in that The processing circuit is also used to adjust the one or more adjustable filter coefficients of the adaptive ANR filter based on the total residual noise signal.

8. An active noise reduction (ANR) device, characterized in that: The ANR device comprises an audio controller according to any one of the above claims and a speaker, wherein the speaker is configured to be driven based on a total noise reduction signal generated by the audio controller.

9. The ANR device according to claim 8, characterized in that: The ANR device includes a feed-forward microphone for detecting the ambient noise signal, wherein the fixed ANR filter (520) is used to generate the first noise reduction signal based on the ambient noise signal, and the adaptive ANR filter (521) is used to generate the second noise reduction signal based on the ambient noise signal.

10. The ANR device according to claim 8, characterized in that: The ANR device includes a feedback microphone for detecting a total residual noise signal, wherein the fixed ANR filter (520) is used to generate the first noise reduction signal based on the total residual noise signal, and the adaptive ANR filter (521) is used to generate the second noise reduction signal based on the total residual noise signal.

11. The ANR device according to claim 8, characterized in that: The ANR device includes a feedforward microphone for detecting the ambient noise signal and a feedback microphone for detecting a total residual noise signal, wherein the fixed ANR filter (520; 520a, 520b) is used to generate the first noise reduction signal based on the ambient noise signal and the total residual noise signal, and the adaptive ANR filter (521; 521a, 521b) is used to generate the second noise reduction signal based on the ambient noise signal and the total residual noise signal.

12. An active noise reduction (ANR) method (800) for reducing an ambient noise signal, characterized in that: The method (800) comprises: generating a first noise-reduced signal using a fixed ANR filter (520; 520a, 520b); generating a second noise reduction signal using an adaptive ANR filter (521; 521a, 521b), wherein the adaptive ANR filter (521; 521a, 521b) comprises one or more adjustable filter coefficients for adjusting the adaptive ANR filter (521; 521a, 521b); generating a total noise reduction signal based on an adjustable weighting coefficient a as an adjustable weighted linear combination of the first noise reduction signal and the second noise reduction signal (522, 523, 524); The adjustable weighting coefficient a is determined based on the following equation: Where MM(x) represents the magnitude measure of the parameter vector x, n represents the time sample index, e(n) represents the total residual noise signal, and e'(n) represents the fractional residual noise signal.

13. A computer program product, characterized in that A non-transitory computer-readable storage medium is included for storing program code, which, when executed by a computer or a processor, causes the computer or the processor to perform the method (800) according to claim 12.

Images