DTX Decision Using Sub-band Signal Variation for Wideband Speech
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current DTX strategies, such as those in AMR and G.729, fail to adaptively transmit SID frames based on actual noise characteristics, leading to delayed noise information transmission and unnecessary bandwidth usage, especially in wideband speech applications where information from both narrowband and wideband components is not fully utilized.
Innovation Solution
A method and device for DTX decision that splits input signals into sub-band signals, obtains characteristic information, and performs a combined decision on the variation of this information to determine whether to transmit SID frames, ensuring timely adaptation to noise changes and efficient bandwidth use by utilizing band-splitting and layered processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If DTX decision is made based on fixed interval or simple energy threshold as in AMR/G.729, then device complexity is reduced, but adaptability to actual noise characteristics deteriorates
Solution Approach 1:
The patent divides the wideband signal into multiple sub-band signals (e.g., low-band and high-band) and performs separate DTX decisions for each sub-band. This segmentation allows the system to capture noise characteristics across different frequency regions, improving adaptability while maintaining manageable complexity through modular processing.
Solution Approach 2:
The patent extends the DTX decision from a single-dimensional energy threshold check to a multi-dimensional analysis by considering characteristic information (such as spectral flatness, zero-crossing rate, or other noise features) across multiple sub-bands. This dimensional expansion enables better adaptation to diverse noise characteristics.
2Loss of time
If SID frames are transmitted continuously, then noise information transmission timeliness is improved, but bandwidth usage increases
Solution Approach 1:
The patent implements dynamic DTX decision-making by continuously monitoring characteristic information variations in each sub-band and adjusting SID frame transmission accordingly. When noise characteristics change significantly in any sub-band, the system promptly transmits updated SID frames, achieving timely adaptation without continuous transmission.
Solution Approach 2:
The system uses feedback from characteristic information analysis to control SID frame transmission. By monitoring whether the variation of noise characteristics exceeds a threshold, the system determines when transmission is necessary, creating a feedback-driven adaptive transmission mechanism that balances timeliness and bandwidth efficiency.
3Device complexity
If only narrowband DTX decision is used, then device complexity is reduced, but measurement precision of wideband noise characteristics deteriorates
Solution Approach 1:
The patent segments the wideband signal into multiple sub-band signals and performs characteristic information extraction and DTX decision for each sub-band separately. This segmentation approach enables accurate wideband noise characterization by combining precise measurements from individual sub-bands, achieving high measurement precision without excessive complexity.
Solution Approach 2:
The patent merges the DTX decision results from multiple sub-bands to form the final wideband DTX decision. By combining the characteristic information variations from low-band and high-band sub-signals, the system achieves comprehensive wideband noise assessment while maintaining modular processing architecture.
Data Source
AI summary
A DTX decision method includes: obtaining sub-band signal(s) according to an input signal; obtaining a variation of characteristic information of each of the sub-band signals; and performing DTX decision according to the variation of the characteristic information of each of the sub-band signals. With the invention, a complete and appreciate DTX decision result is obtained by making full use of the noise characteristic in the speech encoding/decoding bandwidth and using band-splitting and layered processing. As a result, the SID encoding/CNG decoding may closely follow the characteristic variation of the actual noise.