Audio Signal Aliasing for MEMS Eavesdropping Protection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for protecting speech signals from side-channel eavesdropping by limiting MEMS sensor sampling rates restrict hardware and software, limiting the operational scope of applications and failing to ensure data security effectively.
Innovation Solution
The method involves frame blocking, frequency domain conversion, aliasing processing of low-frequency components, and frame fusion to create a second speech signal that is not restorable by MEMS sensors sensitive to low frequencies, leveraging the Haas effect to ensure data security without hardware or software restrictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the sampling rate of MEMS sensors is limited to prevent eavesdropping, then speech signal security is improved, but the operational capability of applications requiring high sampling rates deteriorates
Solution Approach 1:
The patent changes the frequency domain characteristics of the speech signal by introducing aliasing in the low-frequency range (below Nyquist frequency) while preserving high-frequency components. This parameter transformation allows the signal to resist MEMS sensor eavesdropping (which is sensitive to low frequencies) while maintaining compatibility with applications requiring high sampling rates, as the high-frequency information remains intact
Solution Approach 2:
The patent converts the harmful effect of low-frequency aliasing (which would normally degrade signal quality) into a security benefit. By deliberately introducing controlled aliasing in the low-frequency range, the system makes the speech signal resistant to MEMS sensor capture, while the high-frequency components remain preserved for application functionality
2Reliability
If authorization requirements are imposed on MEMS sensor access, then eavesdropping prevention is improved, but system complexity and user convenience deteriorate
Solution Approach 1:
The patent replaces the mechanical/administrative control mechanism (authorization systems, software restrictions, hardware limitations) with a signal processing mechanism. By transforming the speech signal in the frequency domain, security is achieved through mathematical transformation rather than through complex authorization management systems, thereby reducing system complexity
3Reliability
If low-frequency components are removed from the speech signal, then resistance to MEMS sensor eavesdropping is improved, but speech quality and naturalness deteriorate
Solution Approach 1:
Instead of removing low-frequency components (which would degrade speech quality), the patent inverts the approach by adding controlled aliasing to the low-frequency range. This inversion allows low-frequency energy to be present in the signal (maintaining speech quality) while being structured in a way that prevents MEMS sensor reconstruction of the original speech, thus achieving both goals
Data Source
AI summary
This application discloses a method and an apparatus for processing an audio signal. The method includes obtaining a first speech signal acquired by a first device; performing frame blocking on the first speech signal, to obtain multiple speech signal frames; converting the multiple speech signal frames into multiple first frequency domain signal frames; performing aliasing processing on a first sub-frequency domain signal frame among the multiple first frequency domain signal frames with a frequency lower than or equal to a target frequency threshold, and retaining a second sub-frequency domain signal frame among the multiple first frequency domain signal frames with a frequency higher than the target frequency threshold, to obtain multiple second frequency domain signal frames, the target frequency threshold being related to a sampling frequency of a second device; and performing frame fusion on the multiple second frequency domain signal frames, to obtain a second speech signal.