Post-Equalization Impulse Noise Removal in MCM Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multiple carrier modulation (MCM) techniques, such as OFDM, fail to detect and remove impulse noise with magnitudes comparable to or less than the largest expected signal magnitudes, causing masquerading impulse noise to go undetected and unaddressed during pre-equalization noise removal.
Innovation Solution
Implementing a post-equalization impulse noise removal method that estimates and reduces impulse noise in the equalized signal, using a combined FFT, equalization, and impulse-noise-compensation unit to detect and subtract estimated impulse noise from the signal, based on an approximation of total noise and the estimated channel transfer function.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pre-equalization impulse noise removal is performed using traditional clipping methods, then large magnitude impulse noise is reduced, but masquerading impulse noise with magnitudes comparable to or less than expected signal magnitudes remains undetected and unremoved
Solution Approach 1:
The patent performs impulse noise estimation and removal after equalization (post-equalization) rather than before, allowing the equalizer to first compensate for channel effects. This preliminary equalization action enables more accurate subsequent impulse noise detection by working with a cleaner signal baseline, thereby resolving the contradiction between removing large noise and detecting subtle masquerading noise.
Solution Approach 2:
The patent replaces traditional mechanical clipping-based noise removal with a statistical estimation approach. Instead of simply clipping signals above a threshold, the system estimates impulse noise characteristics using signal statistics and subtracts the estimated noise component, enabling detection and removal of masquerading impulse noise that traditional mechanical methods miss.
2Device complexity
If traditional pre-equalization noise removal methods are used, then processing simplicity is maintained, but symbol error rates remain high due to undetected masquerading impulse noise
Solution Approach 1:
The patent merges the equalization process with impulse noise compensation by performing impulse noise estimation using the equalized signal and then subtracting the estimated noise. This combined approach integrates two previously separate functions (equalization and noise removal) into a unified process, achieving better symbol error rates without proportionally increasing system complexity.
Solution Approach 2:
The patent employs feedback mechanisms where the equalized signal is used to estimate impulse noise characteristics, which are then subtracted to produce a cleaned signal. This feedback loop allows the system to adaptively remove masquerading impulse noise based on actual signal conditions, improving reliability while maintaining manageable complexity through iterative refinement.
Data Source
AI summary
A method reduces noise in a multiple carrier modulated (MCM) signal. Such a method may include: estimating impulse noise based in the equalized signal; and removing a portion of the noise in the equalized signal as a function of the estimated impulse noise. An apparatus reduces noise in the MCM. Such an apparatus may include: a down-converter; an analog to digital converter to digitize the output of the down-converter; a guard-interval removing unit operable upon the digitized output of the down-converter; and a combined FFT, equalization and impulse-noise-compensation unit operable upon a signal from the guard-interval-removing unit.