Adaptive Beamforming Noise Suppression Steering Vector Errors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing microphone array beamforming technologies are inadequate in noisy environments, such as automotive and reverberant settings, as they fail to effectively suppress isotropic ambient noise while maintaining robustness against beamsteering errors.
Innovation Solution
The adaptive beamforming technique incorporates sound-source presence probability into an adaptive blocking matrix, using instantaneous direction of arrival and voice activity detection to estimate interference and cancel it, thereby enhancing noise suppression without imposing ad hoc constraints on filter coefficients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If robust GSC with adaptive blocking matrix is used to prevent target signal cancellation, then robustness to steering vector errors is improved, but noise suppression capability deteriorates under isotropic ambient noise conditions
Solution Approach 1:
The patent changes the parameter of filter coefficient constraints from fixed ad hoc constraints to dynamically adapted constraints based on the statistical properties of the blocking matrix outputs. This allows the system to maintain robustness while adapting to different noise environments, specifically improving performance under isotropic ambient noise conditions without sacrificing steering vector error robustness
Solution Approach 2:
The patent introduces a feedback mechanism where the statistical properties of the adaptive blocking matrix outputs are continuously monitored and used to adaptively adjust the filter coefficients of the adaptive interference canceller. This feedback loop enables the system to optimize noise suppression while maintaining target signal integrity in varying environmental conditions
2Object-affected harmful factors
If adaptive interference canceller is used to cancel interfering signals, then directional interference rejection is improved, but target signal cancellation increases when DOA estimation is inaccurate
Solution Approach 1:
The patent applies partial action by using only the necessary portion of the blocking matrix output that corresponds to actual interference, rather than treating all blocking matrix outputs as interference. By adaptively determining which components represent true interference versus target signal leakage, the system cancels directional interference effectively while avoiding excessive cancellation of the target signal
Solution Approach 2:
The patent changes the filter coefficient parameters of the adaptive interference canceller based on the statistical properties of the blocking matrix outputs. This dynamic parameter adjustment allows the system to differentiate between actual interference and target signal leakage, improving directional interference rejection while minimizing target signal cancellation
Data Source
AI summary
A novel adaptive beamforming technique with enhanced noise suppression capability. The technique incorporates the sound-source presence probability into an adaptive blocking matrix. In one embodiment the sound-source presence probability is estimated based on the instantaneous direction of arrival of the input signals and voice activity detection. The technique guarantees robustness to steering vector errors without imposing ad hoc constraints on the adaptive filter coefficients. It can provide good suppression performance for both directional interference signals as well as isotropic ambient noise.