Cancel signal generation device

Abstract

The present invention provides a signal processing technique for noise cancellation in which processing delay is suppressed. This cancel signal generation device includes a first AD conversion unit, a second AD conversion unit, a cancel filter coefficient generation unit, a cancel signal generation unit, and a DA conversion unit. The first AD conversion unit, the second AD conversion unit, the DA conversion unit, and the cancel signal generation unit are configured by an LSI, and the cancel filter coefficient generation unit is configured by a GPU.

Description

Cancellation signal generator

[0001] The present invention relates to active noise control technology.

[0002] For example, Patent Document 1 discloses a system that uses active noise control technology to cancel noise (hereinafter referred to as a noise cancellation system). The AD conversion and DA conversion in the noise cancellation system use, for example, ΔΣ modulation AD conversion and DA conversion. Furthermore, an adaptive algorithm is used to generate cancellation filter coefficients used in filtering by a cancellation filter that generates a signal for canceling noise. The adaptive algorithm in Patent Document 1 reduces processing delay by introducing block processing to reduce the amount of filtering calculations.

[0003] Japanese Patent Application Publication No. 10-247838

[0004] When the sound propagation characteristics in a space where a noise cancellation system is used are complex, it is expected that noise cancellation performance will improve by placing the reference microphone closer to the desired quiet location (i.e., the error microphone location). However, the sound propagation time from the reference microphone to the error microphone location is shorter, so the signal processing for noise cancellation must be performed within a shorter time. For example, if the distance between the reference microphone and the error microphone is 20 cm, the sound speed is approximately 340 m / s, resulting in a propagation time of approximately 0.6 ms. Therefore, the signal processing for noise cancellation must be performed within the time calculated by subtracting the sound propagation time from the speaker location to the error microphone location from the approximately 0.6 ms, which is the sound propagation time from the reference microphone to the error microphone location. Generally, the amount of noise cancellation decreases as the delay in the signal processing for noise cancellation increases. If signal processing is not required to be completed in a short time, there is no problem even if the AD conversion and DA conversion each have a processing delay of about 0.2 ms, and the data transfer has a delay of about 0.1 ms, for a total processing delay of about 0.5 ms. However, as can be seen from the above example, in order to improve cancellation performance, it is necessary to execute signal processing for noise cancellation in a shorter time.

[0005] Therefore, an object of the present invention is to provide a signal processing technique for noise cancellation that suppresses processing delays.

[0006] One aspect of the present invention is a cancellation signal generating device including: a first AD conversion unit that generates, by first AD conversion, a digital reference signal from an analog reference microphone pickup signal output by a reference microphone that picks up noise in a predetermined space; a second AD conversion unit that generates, by second AD conversion, a digital error signal from an analog error microphone pickup signal output by an error microphone that picks up sound at a position where quietness is desired; a cancellation filter coefficient generation unit that generates a cancellation filter coefficient from the digital reference signal and the digital error signal; a cancellation signal generation unit that generates a digital cancellation signal for canceling the noise at the position where quietness is desired from the digital reference signal by filtering with a cancellation filter using the cancellation filter coefficient; and a DA conversion unit that generates an analog cancellation signal from the digital cancellation signal by DA conversion and outputs the analog cancellation signal to a speaker, wherein the first AD conversion unit, the second AD conversion unit, the DA conversion unit, and the cancellation signal generation unit are configured by LSIs, and the cancellation filter coefficient generation unit is configured by a GPU.

[0007] One aspect of the present invention is a cancellation signal generating device including: a first AD conversion unit that generates, by first AD conversion, a digital reference signal from an analog reference microphone pickup signal output by a reference microphone that picks up noise in a predetermined space; a second AD conversion unit that generates, by second AD conversion, a digital error signal from an analog error microphone pickup signal output by an error microphone that picks up sound at a position where quietness is desired; a cancellation filter coefficient generation unit that generates a cancellation filter coefficient from the digital reference signal and the digital error signal; a cancellation signal generation unit that generates a digital cancellation signal from the digital reference signal by filtering with a cancellation filter using the cancellation filter coefficient, for canceling the noise at the position where quietness is desired; and a DA conversion unit that generates an analog cancellation signal from the digital cancellation signal by DA conversion and outputs the analog cancellation signal to a speaker, wherein the first AD conversion unit, the second AD conversion unit, and the DA conversion unit are configured as an automotive audio bus board using LSI, and the cancellation filter coefficient generation unit and the cancellation signal generation unit are configured by a GPU.

[0008] According to the present invention, it is possible to suppress processing delays in signal processing for noise cancellation.

[0009] Fig. 1 is a block diagram showing an example of the configuration of noise cancellation system 1000. Fig. 2 is a flowchart showing an example of the operation of noise cancellation system 1000 / 2000. Fig. 3 is a block diagram showing an example of the configuration of noise cancellation system 1010. Fig. 4 is a flowchart showing an example of the operation of noise cancellation system 1010. Fig. 5 is a block diagram showing an example of the configuration of noise cancellation system 2000. Fig. 6 is a diagram showing an example of the functional configuration of a computer that realizes each device in an embodiment of the present invention.

[0010] Hereinafter, embodiments of the present invention will be described in detail. Components having the same functions are given the same numbers, and duplicated explanations will be omitted.

[0011] Before describing each embodiment, the notation used in this specification will be explained.

[0012] ^ (caret) represents a superscript, e.g., xy^z Yes z is a superscript to x, and x y^z Yes z is a subscript to x. Also, _ (underscore) represents a subscript. For example, x y_z Yes z is a superscript to x, and x y_z Yes z is a subscript to x.

[0013] The superscripts "^" and "~" such as ^x and ~x for a certain letter x should be written directly above the "x", but due to restrictions on the notation in the specification, they are written as ^x and ~x.

[0014] First Embodiment A noise cancellation system 1000 will be described below with reference to FIGS. 1 and 2. FIG. 1 is a block diagram showing the configuration of the noise cancellation system 1000. FIG. 2 is a flowchart showing the operation of the noise cancellation system 1000. As shown in FIG. 1, the noise cancellation system 1000 includes a reference microphone 910, an error microphone 920, a speaker 930, and a cancellation signal generator 100. The cancellation signal generator 100 includes a first AD converter 110, a second AD converter 120, a cancellation filter coefficient generator 130, a cancellation signal generator 140, a DA converter 150, and a recording unit 190. The recording unit 190 is a component that appropriately records information necessary for the processing of the cancellation signal generator 100. The cancellation filter coefficient generator 130 is configured by a GPU 600, which includes a first GPU memory 601, a second GPU memory 602, and a third GPU memory 603. The first AD conversion unit 110, the second AD conversion unit 120, the DA conversion unit 150, and the erasure signal generation unit 140 are configured by an LSI 700, and the LSI 700 includes a first RDMA buffer 701, a second RDMA buffer 702, and a third RDMA buffer 703. Here, the LSI 700 is an LSI that allows changes to be made to the AD conversion settings in the first AD conversion unit 110 and the second AD conversion unit 120, and the DA conversion settings in the DA conversion unit 150, and can be, for example, an FPGA (Field Programmable Gate Array).

[0015] Note that AD is an acronym for Analog Digital, DA is an acronym for Digital Analog, GPU is an acronym for Graphics Processing Unit, LSI is an acronym for Large Scale Integration, and RDMA is an acronym for Remote Direct Memory Access.

[0016] The operation of the noise cancellation system 1000 will now be described with reference to FIG.

[0017] In step S910, the reference microphone 910 picks up noise in a predetermined space and outputs an analog reference microphone pickup signal. Here, the predetermined space is a space where noise is present. The reference microphone 910 is a microphone for picking up sound from a noise source.

[0018] In S920, error microphone 920 picks up sound at a position where quietness is desired and outputs an analog error microphone pickup signal. Here, error microphone 920 is a microphone for picking up sound from a noise source and sound from speaker 930, which is a secondary sound source, and is installed at a position where quietness is desired.

[0019] In S110, the first AD conversion unit 110 receives the analog reference microphone pickup signal output in S910 as input, generates a digital reference signal from the analog reference microphone pickup signal through first AD conversion, and outputs the digital reference signal. Here, the first AD conversion is a process of converting an analog signal into a digital signal. The output digital reference signal is recorded in the first RDMA buffer 701 and then transferred to the first GPU memory 601. The digital reference signal is then recorded in the first GPU memory 601. Note that the transfer of the digital reference signal between the first RDMA buffer 701 and the first GPU memory 601 is performed by RDMA.

[0020] In S120, the second AD conversion unit 120 receives the analog error microphone pickup signal output in S920 as input, generates a digital error signal from the analog error microphone pickup signal by second AD conversion, and outputs the digital error signal. Here, the second AD conversion is a process of converting an analog signal into a digital signal. The output digital error signal is recorded in the second RDMA buffer 702 and then transferred to the second GPU memory 602. The digital error signal is then recorded in the second GPU memory 602. Note that the transfer of the digital error signal between the second RDMA buffer 702 and the second GPU memory 602 is performed by RDMA.

[0021] In S130, the cancellation filter coefficient generation unit 130 receives the digital reference signal output in S110 and the digital error signal output in S120 as input, generates and outputs cancellation filter coefficients from the digital reference signal and the digital error signal. Here, the cancellation filter coefficients are used in filtering by a cancellation filter to generate a digital signal (hereinafter referred to as a digital cancellation signal) for canceling noise at a position to be quieted from the digital reference signal. The cancellation filter coefficient generation unit 130 reads the digital reference signal from the first GPU memory 601 and the digital error signal from the second GPU memory 602. The output cancellation filter coefficients are then recorded in the third GPU memory 603 and then transferred to the third RDMA buffer 703. The cancellation filter coefficients are then recorded in the third RDMA buffer 703. The transfer of the cancellation filter coefficients between the third RDMA buffer 703 and the third GPU memory 603 is performed by RDMA.

[0022] In S140, the cancellation signal generator 140 receives the digital reference signal output in S110 and the cancellation filter coefficients output in S130, and generates and outputs a digital cancellation signal from the digital reference signal by filtering with a cancellation filter using the cancellation filter coefficients. The cancellation signal generator 140 reads the cancellation filter coefficients from the third RDMA buffer 703.

[0023] In S150, the DA conversion unit 150 receives the digital cancellation signal output in S140, generates an analog cancellation signal from the digital cancellation signal by DA conversion, and outputs the analog cancellation signal. DA conversion is a process of converting a digital signal into an analog signal.

[0024] In S930, the speaker 930 receives the analog cancellation signal output in S150 and emits a sound based on the analog cancellation signal. The sound based on the analog cancellation signal is a sound that is in antiphase with the noise at the location where quietness is desired.

[0025] When the noise cancellation system 1000 is applied to a location where the volume of noise is high, such as a public transportation facility, there is no practical problem even if low-volume sounds such as floor noise are not handled. In such a case, the low-pass filters in the first AD conversion unit 110, the second AD conversion unit 120, and the DA conversion unit 150 can be low-pass filters designed to reduce processing delay by lowering accuracy. For example, the low-pass filters can be designed to reduce processing delay by adjusting parameters including the number of stages of a Cascaded Integrator-Comb (CIC) filter, the thinning coefficient, the interpolation coefficient, and the number of taps of the compensation filter of the CIC filter.

[0026] <<Modifications>> In order to compensate for the processing delay (AD conversion processing delay) in the first AD conversion unit 110 and the second AD conversion unit 120 and the processing delay (DA conversion processing delay) in the DA conversion unit 150, the cancellation signal generating device 100 may include a component that changes the frequency characteristic, amplitude characteristic, or phase characteristic, such as an analog equalizer or a phase characteristic changing circuit. An example of an application of active noise control technology using analog circuits is the system described in Reference Non-Patent Document 1.

[0027] (Reference non-patent document 1: Ohira Ikuo, "Active noise control headphones using analog circuits," Journal of the Institute of Electrical Engineers of Japan, vol. 123, no. 8, pp. 516-518, 2003.) A noise cancellation system including a cancellation signal generating device including a component that executes processing to change at least one of the frequency characteristics, amplitude characteristics, and phase characteristics of a signal picked up by an analog reference microphone will be described as a modified example of the first embodiment.

[0028] Noise cancellation system 1010 will be described below with reference to FIGS. 3 and 4. FIG. 3 is a block diagram showing the configuration of noise cancellation system 1010. FIG. 4 is a flowchart showing the operation of noise cancellation system 1010. As shown in FIG. 3, noise cancellation system 1010 includes a reference microphone 910, an error microphone 920, a speaker 930, and a cancellation signal generator 101. Cancellation signal generator 101 includes a first AD converter 110, a second AD converter 120, a cancellation filter coefficient generator 130, a cancellation signal generator 140, a DA converter 150, a characteristics changer 160, and a recorder 190. Recorder 190 is a component that appropriately records information necessary for the processing of cancellation signal generator 101. The cancellation filter coefficient generation unit 130 is configured by a GPU 600, which includes a first GPU memory 601, a second GPU memory 602, and a third GPU memory 603. The first AD conversion unit 110, the second AD conversion unit 120, the DA conversion unit 150, and the cancellation signal generation unit 140 are configured by an LSI 700, which includes a first RDMA buffer 701, a second RDMA buffer 702, and a third RDMA buffer 703. The characteristic change unit 160 is configured by an analog LSI 800, which may be, for example, an analog equalizer or a phase characteristic change circuit. In other words, the cancellation signal generation device 101 differs from the cancellation signal generation device 100 only in that it further includes a characteristic change unit 160.

[0029] The operation of noise cancellation system 1010 will be described with reference to Fig. 4. The processing from S910 to S150 is the same as that of noise cancellation system 1000, so only the processing from S160 onwards will be described here.

[0030] In S160, the characteristic modification unit 160 receives the analog reference microphone pickup signal output in S910 as input, and generates and outputs a modified reference microphone pickup signal from the analog reference microphone pickup signal by modifying at least one of the frequency characteristics, amplitude characteristics, and phase characteristics of the analog reference microphone pickup signal.

[0031] In S930, the speaker 930 receives the analog cancellation signal output in S150 and the modified reference microphone signal output in S160, and emits a sound based on the analog cancellation signal and the modified reference microphone signal. Here, the sound based on the analog cancellation signal and the modified reference microphone signal is a sound that is in an antiphase relationship with the noise at the position where quietness is desired.

[0032] According to the embodiment of the present invention, it is possible to suppress processing delay in signal processing for noise cancellation.

[0033] Since the erasure filter coefficient generation algorithm generally requires a large amount of calculation and is desirably easy to improve, the erasure filter coefficient generation unit is configured as a GPU capable of executing a large amount of calculation at high speed, thereby making it possible to suppress processing delays even when the erasure filter coefficient generation algorithm becomes complex.

[0034] Furthermore, signal and data transfer between the LSI and GPU is performed directly via RDMA without going through the CPU, which allows signal processing for noise cancellation to be performed without being affected by CPU performance constraints or other processing being performed by the CPU, thereby reducing processing delays.

[0035] In signal processing for noise cancellation, the processing delay due to AD conversion and DA conversion is dominant. However, as mentioned above, in places with high noise levels, the noise tolerance in AD and DA conversion filter processing is high. Therefore, AD conversion filters and DA conversion filters can be designed to shorten the processing time. Therefore, even when the sound propagation time from the reference microphone to the error microphone is short, such as 0.1 ms to 1 ms, a highly accurate analog cancellation signal can be generated. Furthermore, by using an analog LSI that combines an analog equalizer and a phase characteristic change circuit, a highly accurate analog cancellation signal can be generated even when the sound propagation time is short, such as 0.1 ms or less.

[0036] By using an analog equalizer or phase characteristic modification circuit, it is possible to correct the direct wave component of noise from the position of a reference microphone, which can be measured in advance, to the desired quiet position, as well as the acoustic reproduction characteristics of the speaker, so as to suppress processing delay. This effect is particularly noticeable in the frequency range below 200 Hz. Furthermore, by combining an analog equalizer or phase characteristic modification circuit with a component that generates a digital cancellation signal, the number of filter taps in that component can be reduced, thereby suppressing processing delay in that component.

[0037] <Second embodiment> When a noise cancellation system is applied to a vehicle, in order to reduce the weight, cost and wiring of the audio cable, an automotive audio bus (ABA) microphone is used. 2 B) The compatible microphone may be used as the reference microphone 910 or the error microphone 920. Here, a noise cancellation system configured using an automotive audio bus compatible microphone will be described.

[0038] Noise cancellation system 2000 will be described below with reference to FIGS. 5 and 2. FIG. 5 is a block diagram showing the configuration of noise cancellation system 2000. FIG. 2 is a flowchart showing the operation of noise cancellation system 2000. As shown in FIG. 5, noise cancellation system 2000 includes a reference microphone 910, an error microphone 920, a speaker 930, and cancellation signal generator 200. Cancellation signal generator 200 includes a first AD converter 110, a second AD converter 120, a cancellation filter coefficient generator 130, a cancellation signal generator 140, a DA converter 150, and a recording unit 190. Recording unit 190 is a component that appropriately records information necessary for the processing of cancellation signal generator 200. The cancellation filter coefficient generation unit 130 and the cancellation signal generation unit 140 are configured by a GPU 650, and the GPU 650 includes a first GPU memory 651, a second GPU memory 652, and a third GPU memory 653. The first AD conversion unit 110, the second AD conversion unit 120, and the DA conversion unit 150 are configured by an LSI 750, and the LSI 750 includes a first RDMA buffer 751, a second RDMA buffer 752, and a third RDMA buffer 753. Here, the LSI 750 is an LSI that allows changes to be made to the AD conversion settings in the first AD conversion unit 110 and the second AD conversion unit 120, and the DA conversion settings in the DA conversion unit 150, and can be, for example, an FPGA (Field Programmable Gate Array). The first AD conversion unit 110, the second AD conversion unit 120, and the DA conversion unit 150 are configured as an automotive audio bus board 850 using an LSI 750.

[0039] The operation of noise cancellation system 2000 will now be described with reference to FIG.

[0040] In step S910, the reference microphone 910 picks up noise in a predetermined space and outputs an analog reference microphone pickup signal. Here, the predetermined space is a space where noise is present. The reference microphone 910 is a microphone for picking up sound from a noise source.

[0041] In S920, error microphone 920 picks up sound at a position where quietness is desired and outputs an analog error microphone pickup signal. Here, error microphone 920 is a microphone for picking up sound from a noise source and sound from speaker 930, which is a secondary sound source, and is installed at a position where quietness is desired.

[0042] In S110, the first AD conversion unit 110 receives the analog reference microphone pickup signal output in S910 as input, generates a digital reference signal from the analog reference microphone pickup signal through first AD conversion, and outputs the digital reference signal. Here, the first AD conversion is a process of converting an analog signal into a digital signal. The output digital reference signal is recorded in the first RDMA buffer 751 and then transferred to the first GPU memory 651. The digital reference signal is then recorded in the first GPU memory 651. Note that the transfer of the digital reference signal between the first RDMA buffer 751 and the first GPU memory 651 is performed by RDMA.

[0043] In S120, the second AD conversion unit 120 receives the analog error microphone pickup signal output in S920 as input, generates a digital error signal from the analog error microphone pickup signal through second AD conversion, and outputs the digital error signal. Here, the second AD conversion is a process of converting an analog signal into a digital signal. The output digital error signal is recorded in the second RDMA buffer 752 and then transferred to the second GPU memory 652. The digital error signal is then recorded in the second GPU memory 652. Note that the transfer of the digital error signal between the second RDMA buffer 752 and the second GPU memory 652 is performed by RDMA.

[0044] In S130, the cancellation filter coefficient generation unit 130 receives the digital reference signal output in S110 and the digital error signal output in S120 as input, generates and outputs cancellation filter coefficients from the digital reference signal and the digital error signal. Here, the cancellation filter coefficients are used in filtering by the cancellation filter to generate a digital signal (hereinafter referred to as a digital cancellation signal) for canceling noise at a position that is to be made quieter from the digital reference signal. The cancellation filter coefficient generation unit 130 reads the digital reference signal from the first GPU memory 651 and reads the digital error signal from the second GPU memory 652.

[0045] In S140, the cancellation signal generator 140 receives the digital reference signal output in S110 and the cancellation filter coefficients output in S130 as input, generates a digital cancellation signal from the digital reference signal by filtering with a cancellation filter using the cancellation filter coefficients, and outputs the generated digital cancellation signal. The output digital cancellation signal is recorded in the third GPU memory 653 and then transferred to the third RDMA buffer 753. The digital cancellation signal is then recorded in the third RDMA buffer 753. Note that the transfer of the digital cancellation signal between the third RDMA buffer 753 and the third GPU memory 653 is performed by RDMA.

[0046] In S150, the DA converter 150 receives the digital erasure signal output in S140, and generates and outputs an analog erasure signal from the digital erasure signal through DA conversion. DA conversion is a process of converting a digital signal into an analog signal. The DA converter 150 reads the digital erasure signal from the third RDMA buffer 753.

[0047] In S930, the speaker 930 receives the analog cancellation signal output in S150 and emits a sound based on the analog cancellation signal. The sound based on the analog cancellation signal is a sound that is in antiphase with the noise at the location where quietness is desired.

[0048] When the noise cancellation system 2000 is applied to a location where the noise level is high, such as a public transportation facility, there is no practical problem even if low-volume sounds such as floor noise are not handled. In such a case, the low-pass filters in the first AD conversion unit 110, the second AD conversion unit 120, and the DA conversion unit 150 can be low-pass filters designed to reduce processing delay by lowering accuracy. For example, the low-pass filters can be designed to reduce processing delay by adjusting parameters including the number of stages of a Cascaded Integrator-Comb (CIC) filter, the thinning coefficient, the interpolation coefficient, and the number of taps of the compensation filter of the CIC filter.

[0049] According to the embodiment of the present invention, it is possible to suppress processing delay in signal processing for noise cancellation.

[0050] Since the erasure filter coefficient generation algorithm generally requires a large amount of calculation, and it is desirable that it can be easily improved, the erasure filter coefficient generation unit is configured as a GPU that can execute a large amount of calculation at high speed. This makes it possible to suppress processing delays even when the erasure filter coefficient generation algorithm becomes complex. Furthermore, since the digital erasure signal generation process can also be a heavy process, the erasure signal generation unit is also configured as a GPU. This makes it possible to suppress processing delays.

[0051] Furthermore, signals are transferred directly between the LSI and the GPU using RDMA without going through the CPU, which allows signal processing for noise cancellation to be performed without being affected by CPU performance constraints or other processing being performed by the CPU, thereby reducing processing delays.

[0052] In signal processing for noise cancellation, the processing delay due to AD conversion and DA conversion is dominant, but as mentioned above, in places with high noise levels, the noise tolerance in filter processing for AD conversion and DA conversion is high, so AD conversion filters and DA conversion filters can be designed to shorten the processing time.As a result, even if the sound propagation time from the installation position of the reference microphone to the installation position of the error microphone is short, at 0.1 ms to 1 ms, a highly accurate analog cancellation signal can be generated.

[0053] <Additional Notes> The functions performed by the components described herein may be implemented in circuitry or processing circuitry, including general-purpose processors, application-specific processors, integrated circuits, ASICs (Application Specific Integrated Circuits), CPUs (Central Processing Units), conventional circuits, and / or combinations thereof, programmed to perform the described functions. A processor includes transistors and other circuits and is considered to be circuitry or processing circuitry. A processor may also be a programmed processor that executes a program stored in a memory.

[0054] In this specification, a circuitry, unit, or means is hardware that is programmed to realize or performs the described functions, which may be any hardware disclosed herein or any hardware known to be programmed to realize or perform the described functions.

[0055] If the hardware is a processor considered to be a type of circuitry, the circuitry, means, or unit is a combination of the hardware and software used to configure the hardware and / or processor.

[0056] The various processes described above can be implemented by loading a program that executes each step of the above method into the recording unit 2020 of the computer 2000 shown in Figure 6, and operating the control unit 2010, input unit 2030, output unit 2040, display unit 2050, etc.

[0057] The program describing the processing contents can be recorded on a computer-readable recording medium, which may be, for example, a magnetic recording device, an optical disk, a magneto-optical recording medium, a semiconductor memory, or any other suitable recording medium.

[0058] The program may be distributed by, for example, selling, transferring, lending, etc. portable recording media such as DVDs and CD-ROMs on which the program is recorded. Furthermore, the program may be stored in a storage device of a server computer, and then transferred from the server computer to other computers via a network, thereby distributing the program.

[0059] A computer that executes such a program may first temporarily store the program recorded on a portable recording medium or transferred from a server computer in its own storage device. Then, when executing a process, the computer reads the program stored in its storage device and executes the process in accordance with the read program. Alternatively, the computer may read the program directly from a portable recording medium and execute the process in accordance with the program. Furthermore, the computer may execute the process in accordance with the program each time a program is transferred from a server computer to the computer. Alternatively, the server computer may not transfer the program to the computer, but may instead execute the process through a so-called ASP (Application Service Provider) service, which realizes the processing function by issuing an execution instruction and obtaining the results. Furthermore, the server computer may execute the process at the terminal using a so-called SaaS (Software as a Service) service, which allows users to use part of the server computer along with the program. In this embodiment, the program includes information used for processing by an electronic computer that is equivalent to a program (such as data that is not a direct instruction to a computer but has properties that dictate computer processing).

[0060] Furthermore, in this embodiment, the device is configured by executing a predetermined program on a computer, but at least a part of the processing contents may be realized by hardware.

[0061] The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention. Furthermore, the processes described in the above embodiments may not only be executed in chronological order according to the order described, but may also be executed in parallel or individually depending on the processing capacity of the device that executes the processes or as needed.

Claims

1. A noise cancellation signal generation device including: a first AD conversion unit that generates a digital reference signal from an analog reference microphone pickup signal output by a reference microphone that picks up noise in a predetermined space through first AD conversion; a second AD conversion unit that generates a digital error signal from an analog error microphone pickup signal output by an error microphone that picks up sound at a position where noise is to be canceled through second AD conversion; a cancellation filter coefficient generation unit that generates a cancellation filter coefficient from the digital reference signal and the digital error signal; a cancellation signal generation unit that generates a digital cancellation signal for canceling the noise at the position where noise is to be canceled from the digital reference signal through filtering by a cancellation filter using the cancellation filter coefficient; and a DA conversion unit that generates an analog cancellation signal from the digital cancellation signal and outputs the analog cancellation signal to a speaker, wherein the first AD conversion unit, the second AD conversion unit, the DA conversion unit, and the cancellation signal generation unit are configured by an LSI, and the cancellation filter coefficient generation unit is configured by a GPU.

2. The noise cancellation signal generation device according to claim 1, wherein the LSI includes a first RDMA buffer, a second RDMA buffer, and a third RDMA buffer, the GPU includes a first GPU memory, a second GPU memory, and a third GPU memory, the first RDMA buffer and the first GPU memory are used for transferring the digital reference signal by RDMA, the second RDMA buffer and the second GPU memory are used for transferring the digital error signal by RDMA, and the third RDMA buffer and the third GPU memory are used for transferring the cancellation filter coefficient by RDMA.

3. The noise cancellation signal generation device according to claim 1, further including a characteristic change unit that generates a modified reference microphone pickup signal from the analog reference microphone pickup signal by changing at least one of the frequency characteristic, amplitude characteristic, and phase characteristic of the analog reference microphone pickup signal and outputs the modified reference microphone pickup signal to the speaker, wherein the characteristic change unit is configured by an analog LSI.

4. A cancellation signal generation device including: a first AD conversion unit that generates a digital reference signal from an analog reference microphone pickup signal output by a reference microphone that picks up noise in a predetermined space through first AD conversion; a second AD conversion unit that generates a digital error signal from an analog error microphone pickup signal output by an error microphone that picks up sound at a position where noise is to be reduced through second AD conversion; a cancellation filter coefficient generation unit that generates a cancellation filter coefficient from the digital reference signal and the digital error signal; a cancellation signal generation unit that generates a digital cancellation signal for canceling the noise at the position where noise is to be reduced from the digital reference signal through filtering by a cancellation filter using the cancellation filter coefficient; and a DA conversion unit that generates an analog cancellation signal from the digital cancellation signal and outputs the analog cancellation signal to a speaker. The first AD conversion unit, the second AD conversion unit, and the DA conversion unit are configured as an automotive audio board by an LSI, and the cancellation filter coefficient generation unit and the cancellation signal generation unit are configured by a GPU.

5. The cancellation signal generation device according to claim 4, wherein the LSI includes a first RDMA buffer, a second RDMA buffer, and a third RDMA buffer, the GPU includes a first GPU memory, a second GPU memory, and a third GPU memory, the first RDMA buffer and the first GPU memory are used for transferring the digital reference signal by RDMA, the second RDMA buffer and the second GPU memory are used for transferring the digital error signal by RDMA, and the third RDMA buffer and the third GPU memory are used for transferring the digital cancellation signal by RDMA.

6. The cancellation signal generation device according to claim 1 or 4, wherein the low-pass filters in the first AD conversion unit, the second AD conversion unit, and the DA conversion unit are designed such that the processing delay is reduced by reducing the accuracy.

Images