Adaptive Frame Error Concealment Using Overlapping Signal Synthesis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing time-domain error concealment methods, such as repetition and interpolation, face challenges in effectively concealing frame errors, especially when errors occur over long intervals or with different window types, leading to modulation noise and increased complexity.
Innovation Solution
A frame error concealment method using an overlapping technique that synthesizes similar signals from previous frames to reconstruct error frames, preventing modulation noise by combining signals with high correlation and applying adaptive window functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the repetition method is used in the time domain, then the error frame can be reconstructed by reusing previous frame signals, but modulation noise occurs when errors occur over long intervals
Solution Approach 1:
The patent applies dynamics by making the error concealment method adaptive rather than static. The system dynamically selects between repetition method, interpolation method, or overlapping method based on real-time detection of error patterns, frame types, and correlation values. This dynamic adaptation prevents modulation noise by choosing the appropriate method for each specific error condition rather than consistently using repetition.
Solution Approach 2:
The patent changes parameters by introducing correlation value thresholds and error pattern classifications as decision criteria. The system calculates correlation values between frames and uses these parameters to determine whether to apply repetition, interpolation, or overlapping methods. This parameter-based adaptation allows the system to avoid modulation noise by switching methods when error intervals exceed certain thresholds.
2Reliability
If the interpolation method is used in the time domain, then the error frame can be reconstructed by interpolating previous and next frame signals, but it becomes difficult to use when errors occur over long intervals
Solution Approach 1:
The patent makes the error concealment system dynamic by enabling it to switch between interpolation method and overlapping method based on the duration and pattern of errors. When errors occur over long intervals where interpolation becomes difficult, the system dynamically transitions to the overlapping method, which can handle extended error periods more effectively by combining multiple previous frame signals.
3Reliability
If other time-domain methods are used that analyze previous frame signals, then error frames can be reconstructed, but device complexity increases
Solution Approach 1:
The patent applies partial action by using a simplified version of signal analysis that focuses only on calculating correlation values between frames, rather than performing full spectral analysis. This partial analysis approach provides sufficient information for method selection and signal reconstruction without the computational complexity of complete signal analysis, thus reducing device complexity while maintaining effectiveness.
4Reliability
If frequency-domain error concealment methods are used, then spectrum coefficients can be reused or interpolated, but they fail when window types differ or burst errors occur
Solution Approach 1:
The patent introduces the overlapping method as an intermediary approach that bridges the gap between frequency-domain methods and pure time-domain methods. By combining signals from multiple previous frames using overlapping and addition operations, this intermediary method can handle different window types and burst errors more effectively than traditional frequency-domain approaches, while maintaining computational efficiency.
Data Source
AI summary
Provided are a frame error concealment method and apparatus and an error concealment scheme construction method and apparatus. The frame error concealment method includes generating a new signal by synthesizing a plurality of previous signals that are similar to a signal of an error frame and reconstructing the signal of the error frame using the generated signal.