Signal Detection Using Median Absolute Deviation for DC Offset Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting received signals in wireless communication systems, particularly in low-powered systems, face inefficiencies and inaccuracies due to noise correlation, signal distortion, and DC offset, making it challenging to achieve reliable signal detection.
Innovation Solution
A method and device that utilize correlation between received signals and training symbols, along with median absolute deviation (MAD) analysis, to detect signal presence, effectively mitigating noise and DC offset influences by using MAD as a representative value for the reference interval, which is less complex and more accurate than covariance calculation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If correlation function is used for signal detection, then detection capability is improved, but false alarms increase due to noise correlation
Solution Approach 1:
The patent introduces MAD (Median Absolute Deviation) as an intermediary metric to detect DC offset and noise characteristics separately from the correlation function. By using MAD to identify and compensate for DC offset effects, the system maintains the correlation function's detection capability while eliminating its susceptibility to false alarms caused by noise correlation and DC offset.
Solution Approach 2:
The patent changes the detection parameter from relying solely on correlation magnitude to using a combination of correlation and MAD-based DC offset detection. By monitoring the MAD value across different time intervals and comparing it to threshold values, the system dynamically adjusts its detection criteria to distinguish between actual signal presence and noise/DC offset effects.
2Measurement precision
If covariance calculation is used to mitigate DC offset, then detection accuracy is improved, but hardware complexity increases
Solution Approach 1:
The patent replaces the computationally expensive covariance calculation with the simpler MAD (Median Absolute Deviation) calculation. MAD requires only absolute value operations and median finding, which are much less complex to implement in hardware than covariance calculation. This substitution maintains the ability to detect and compensate for DC offset while significantly reducing hardware complexity and computational burden.
3Device complexity
If energy-based detection is used, then hardware complexity is reduced, but detection accuracy deteriorates due to noise influence
Solution Approach 1:
The patent merges the advantages of both correlation-based and energy-based detection methods. It uses the correlation function to detect signal presence while incorporating MAD-based DC offset detection to mitigate noise influence. This combination allows the system to maintain high detection accuracy like correlation methods while avoiding their susceptibility to false alarms, and simultaneously avoids the DC offset sensitivity of pure energy-based methods.
Data Source
AI summary
A method of detecting a received signal, includes determining correlation between a received signal in a predetermined signal interval and training symbols, and determining a median absolute deviation (MAD) of the received signal in a predetermined reference interval. The method further includes detecting a presence of the received signal based on the correlation and the MAD.