Loop Start Trunk Echo Cancellation via Impulse Response Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for echo cancellation in Loop-Start (LS) trunk lines, particularly in PBX to CO connections, are ineffective due to nonlinear effects caused by power amplifiers, D/A and A/D converters, and noise, leading to intrusive echoes and residue echoes that are difficult to manage with linear filters, requiring manual and time-consuming impedance matching.
Innovation Solution
An offline method using a Digital Signal Processor (DSP) to estimate the impulse response of the LS trunk circuit, convolve a speech-like signal with it, and measure residue echo power levels to derive distortion characteristics, allowing for automatic adjustment of echo cancellation settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual impedance matching is performed to reduce echo, then echo cancellation effectiveness is improved, but time consumption and human effort increase significantly
Solution Approach 1:
The system automatically performs impedance matching and echo cancellation parameter optimization by having the LS trunk circuit itself provide test signals and measurements, eliminating the need for manual operator intervention while achieving optimal echo cancellation settings
Solution Approach 2:
The patent performs offline measurements and impulse response estimation before actual voice communication occurs, pre-determining the optimal echo cancellation parameters so that when live calls occur, the system is already optimized and ready to cancel echoes effectively without real-time adjustment delays
2Device complexity
If linear FIR filter is used for echo cancellation, then implementation simplicity is improved, but effectiveness deteriorates due to nonlinear effects in the circuit
Solution Approach 1:
The patent performs preliminary offline measurements to estimate the impulse response and characterize nonlinear distortion of the echo path before actual echo cancellation. This pre-characterization allows the system to understand both linear and nonlinear properties of the circuit, enabling more accurate cancellation without requiring complex real-time processing
Solution Approach 2:
The system uses feedback from the measured impulse response and distortion characteristics to adaptively adjust echo cancellation parameters. By continuously monitoring the echo path properties and adjusting the cancellation algorithm accordingly, the system maintains high effectiveness despite nonlinear effects while keeping the implementation manageable
3Speed
If online adaptive procedure is used to reduce residue echo, then real-time echo cancellation is improved, but ability to reduce residue echo below threshold deteriorates due to buried nonlinearity
Solution Approach 1:
The patent performs comprehensive offline measurements to pre-estimate the impulse response and characterize nonlinear distortion before real-time operation. This preliminary characterization captures the buried nonlinearity that would be difficult to detect during online adaptive procedures, providing a foundation for more effective residue echo reduction while maintaining real-time performance
Solution Approach 2:
The patent introduces an intermediary offline measurement and analysis stage that acts as a mediator between the physical circuit and the real-time echo cancellation algorithm. This intermediary process extracts and characterizes the nonlinear properties that are otherwise hidden during normal operation, enabling the real-time system to compensate for these effects more effectively
Data Source
AI summary
A method of determining residue-echo threshold for the trans-hybrid echo path of a loop start (LS) trunk line, comprising the steps of estimating the impulse response of the trans-hybrid echo path, passing a speech-like signal through the LS trunk circuit and collecting the echo signal. The collected echo signal is compared to the convolution of the speech-like signal with the estimated impulse response. The difference between the convolution result and the collected echo signal is the residue echo, which represents the effect of nonlinearity. With different levels of excitation signal, a curve of residue echo power level against input signal level is obtained, and the distortion characteristics are derived from this curve.


