Active Noise Reduction Using Frequency-Domain Secondary Path Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing active noise reduction systems face high computation demands due to long secondary path estimation, leading to reduced operating speed and poorer noise reduction performance.

Innovation Solution

A method that converts convolution calculations into simple operations using real and imaginary part values of target estimation coefficients, reducing calculation complexity and improving system speed and noise reduction effectiveness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If long secondary path estimation is used for signal convolution calculation and filter coefficient updating, then the noise reduction accuracy is improved, but the computation amount increases and the operating speed decreases

Engineering Contradiction:
Improvenoise reduction accuracyVSAvoidoperating speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the long secondary path estimation into multiple short secondary path estimations corresponding to different frequency bands. Instead of performing one long convolution operation, the system divides the frequency spectrum into multiple bands, estimates secondary paths for each band separately, and processes them independently. This segmentation reduces the computational burden of each individual estimation while maintaining overall noise reduction accuracy through comprehensive frequency coverage.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the time-domain convolution operation into frequency-domain processing by applying Fast Fourier Transform (FFT). This dimensionality change from time domain to frequency domain allows the system to perform filter coefficient updating and signal processing more efficiently. The long secondary path estimation is converted into multiple short estimations in the frequency domain, significantly reducing computation amount while preserving the essential noise cancellation functionality.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If long secondary path estimation is used for signal convolution calculation and filter coefficient updating, then the noise reduction accuracy is improved, but the requirement on chip processing power increases

Engineering Contradiction:
Improvenoise reduction accuracyVSAvoidchip processing power
Core Design Contradiction:
Measurement precisionVSPower

Solution Approach 1:

The patent segments the computationally intensive long secondary path estimation into multiple shorter estimations across different frequency bands. Each short estimation requires less processing power individually, allowing the system to distribute the computational load across multiple smaller operations rather than one large operation. This segmentation enables the system to achieve accurate noise reduction while staying within the processing power constraints of mobile device chips.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent replaces the computationally heavy time-domain convolution mechanism with frequency-domain processing using FFT and inverse FFT transformations. This substitution changes the mathematical approach from direct time-domain multiplication to efficient frequency-domain operations, dramatically reducing the processing power requirement while maintaining the ability to perform accurate filter coefficient updating and noise cancellation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Manufacturing precision

If long secondary path estimation is used, then the filter coefficient updating accuracy is improved, but the computation amount increases

Engineering Contradiction:
Improvefilter coefficient updating accuracyVSAvoidcomputation amount
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent segments the filter coefficient updating process into multiple independent updates corresponding to different frequency bands. Instead of performing one large-scale coefficient update based on long secondary path estimation, the system performs multiple smaller updates based on short secondary path estimations for each frequency band. This segmentation maintains the accuracy of coefficient updating by addressing each frequency band's specific characteristics while significantly reducing the total computation amount required.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the filter coefficient updating from time-domain convolution to frequency-domain processing. By applying FFT to convert the secondary path estimation and reference signal into frequency domain representations, the system can perform coefficient updates more efficiently. The long estimation is broken into multiple short estimations in the frequency domain, reducing computation while preserving the accuracy needed for effective active noise cancellation.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP4723098A1Noise reduction method and apparatus, terminal device, and storage medium
Publication Date: 2026.04.08 BYD CO LTD
  • EP4723098A1 patent drawingFigure 1~2
  • EP4723098A1 patent drawingFigure 3
  • EP4723098A1 patent drawingFigure 4~5

AI summary

Provided are a noise reduction method and apparatus, a terminal device, and a storage medium. The method includes: obtaining a target estimation coefficient, where the target estimation coefficient includes a real part value and an imaginary part value; then, generating, based on a target noise frequency, an original filtered signal whose frequency is the target noise frequency; and obtaining a target filtered signal based on the original filtered signal and the real part value and the imaginary part value in the target estimation coefficient, where the target filtered signal is used to control a speaker to output a filtered acoustic signal, to perform noise reduction on the target noise.