Speech Decoder Frame Erasure Recovery via Adaptive Gain
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Frame erasure in speech transmission systems, such as those used in cellular telephony and VoIP, occurs due to errors in wireless channels, leading to unrecoverable corruption of speech frames, which existing technologies struggle to address effectively without significant quality loss.
Innovation Solution
The method involves detecting erasure in sustained voiced segments and calculating replacement frames using excitation signals based on gain values, where a higher gain factor is used to generate replacement excitation signals, allowing for the reconstruction of decoded speech frames even in the presence of frame erasure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If speech compression techniques are employed to reduce information transfer, then bandwidth efficiency is improved, but speech quality may deteriorate
Solution Approach 1:
The patent changes the gain parameter of excitation signals to compensate for frame erasures. By adjusting the gain value based on detected erasure conditions and comparing it with threshold values, the system maintains speech quality despite using compressed representations, thus resolving the contradiction between compression efficiency and quality preservation
2Device complexity
If frame erasure recovery is not implemented, then device complexity is reduced, but speech quality deteriorates during transmission errors
Solution Approach 1:
The patent performs preliminary detection of frame erasures using CRC checks before decoding, and prepares compensatory excitation signals in advance. This preliminary action allows the system to quickly respond to erasures without adding complex real-time processing, maintaining speech quality while keeping device complexity manageable
Solution Approach 2:
The patent introduces an intermediary excitation signal generation mechanism that bridges the gap between received encoded frames and reconstructed speech. By generating intermediate excitation signals with adjusted gain values, the system recovers from frame erasures without requiring complex error correction codes, thus maintaining quality with moderate complexity
3Reliability
If higher gain factors are used to generate replacement excitation signals, then frame erasure recovery is improved, but distortion in voiced segments may increase
Solution Approach 1:
The patent dynamically adjusts the gain factor based on the detected frame erasure condition and compares it with threshold values derived from the encoded speech signal. This dynamic adjustment allows the system to use higher gain factors when needed for recovery while preventing excessive distortion, thus resolving the contradiction between recovery reliability and speech fidelity
Solution Approach 2:
The patent implements a feedback mechanism where the gain factor is adjusted based on comparison with threshold values and the detected erasure condition. This feedback control ensures that compensation is applied appropriately - strong enough to recover from erasures but controlled enough to maintain speech quality, resolving the contradiction between recovery effectiveness and distortion prevention
Data Source
AI summary
In one configuration, erasure of a significant frame of a sustained voiced segment is detected. An adaptive codebook gain value for the erased frame is calculated based on the preceding frame. If the calculated value is less than (alternatively, not greater than) a threshold value, a higher adaptive codebook gain value is used for the erased frame. The higher value may be derived from the calculated value or selected from among one or more predefined values.