Audible Acoustic Synchronization Using MCLT for Low-Load Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In acoustic communication, existing methods require a large amount of calculation for synchronization in the audible frequency range, leading to performance deterioration and increased computational load, especially when calculating correlation values for each sample.
Innovation Solution
A synchronization system and method that transforms audio signals using a first-type transform like MCLT for inserting synchronization data and a second-type transform like FFT for receiver processing, normalizing coefficients, and determining synchronization locations based on peak correlation values, reducing the need for per-sample calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If correlation is calculated for each sample to achieve precise synchronization detection, then measurement precision is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments the continuous signal processing into discrete frame-based operations. Instead of calculating correlation for every single sample point, the system divides the audio signal into frames and performs synchronization detection at frame boundaries using MCLT coefficients. This segmentation reduces the number of correlation calculations from potentially thousands per second to a manageable number of frame-level operations, directly resolving the contradiction between precision and computational complexity.
Solution Approach 2:
The patent applies preliminary action by pre-calculating and storing MCLT coefficients for each frame before performing correlation detection. The synchronization data is embedded in the MCLT coefficients during encoding, and the receiver prepares reference MCLT coefficients in advance. When synchronization is needed, the pre-computed coefficients are directly compared rather than calculating correlations from raw samples, significantly reducing real-time computational burden while maintaining detection precision.
2Measurement precision
If MCLT coefficients are calculated for each sample to ensure accurate synchronization, then measurement precision is improved, but processing time increases
Solution Approach 1:
The patent segments the signal processing into frame-based operations where MCLT coefficients are calculated once per frame rather than for each individual sample. This frame-based segmentation reduces the frequency of MCLT calculations from sample-rate to frame-rate, dramatically reducing processing time while maintaining synchronization accuracy through the use of phase information in the coefficients.
Solution Approach 2:
The patent changes the parameter being analyzed from time-domain samples to frequency-domain MCLT coefficients. By transforming the synchronization detection problem from the time domain to the frequency domain using MCLT, the system can extract synchronization information from phase characteristics of coefficients rather than computing correlations across all time samples, significantly reducing processing time while preserving accuracy.
3Device complexity
If fast Fourier transform is used to reduce calculation amount, then computational complexity is reduced, but synchronization performance deteriorates
Solution Approach 1:
The patent changes the transform parameter from standard FFT to MCLT (Modified Complex Lapped Transform). This parameter change in the transform type preserves the computational efficiency benefits of transform-based methods while maintaining the time-domain overlap characteristics necessary for accurate synchronization detection. The MCLT's specific properties allow it to maintain synchronization performance that FFT cannot achieve, while still providing computational reduction compared to time-domain correlation.
Solution Approach 2:
The patent substitutes the mechanical time-domain correlation process with a frequency-domain MCLT-based approach. Instead of mechanically sliding one signal over another and computing point-by-point correlations in the time domain, the system uses MCLT to transform signals into the frequency domain where correlation can be computed more efficiently through coefficient comparison, reducing computational load while maintaining synchronization accuracy.
Data Source
AI summary
Provided is a synchronization system and method for acoustic communication in audible frequency range, and an apparatus applied thereto. The synchronization system for acoustic communication in audible frequency range is configured to prevent deterioration of a synchronization performance and to reduce an amount of calculation by calculating a correlation based on a few samples as opposed to calculating a correlation for each sample when a receiver of the acoustic communication performs synchronization while the acoustic communication is performed in the audible frequency range through modification of an audio signal or adding of a predetermined signal to an audio signal.