Decision Feedback Equalization for Unequal Bit Pattern Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Decision Feedback Equalization (DFE) technologies are less accurate for colored signals, which are common in high-data-transmission rates, due to their unequal distribution of bit patterns, leading to inefficiencies in signal equalization.
Innovation Solution
A method and system that adapt equalizing coefficients by disregarding, averaging, or weighting signals with unequally distributed bit patterns, using a record to track bit pattern occurrences and adjust coefficients based on a more-equal distribution, effectively de-colorizing signals and improving equalization accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If DFE uses standard equalization for colored signals, then the system can process high-data-transmission signals, but the equalization accuracy deteriorates due to unequal bit pattern distribution
Solution Approach 1:
The patent changes the parameter distribution by deliberately generating training sequences with equal distribution of bit patterns (0s and 1s), contrasting with the unequal distribution in actual colored signals. This parameter change in the training data enables accurate DFE coefficient adaptation, which then improves equalization accuracy for the actual colored signals during data transmission.
Solution Approach 2:
The patent performs preliminary equalization by using specially designed training sequences before actual data transmission. The DFE coefficients are adapted in advance using these balanced training sequences, allowing the system to prepare optimal equalization parameters that compensate for channel effects and colored signal characteristics, thereby improving subsequent equalization accuracy.
2Adaptability or versatility
If DFE adapts coefficients using all received signals, then the adaptation process is continuous, but the equalization accuracy deteriorates due to dominance of frequently occurring bit patterns
Solution Approach 1:
The patent changes the weighting parameter in coefficient adaptation by assigning equal weight to all bit patterns in the training sequence, regardless of their occurrence frequency in actual signals. This parameter change prevents dominant bit patterns from skewing the adaptation process, ensuring balanced and accurate coefficient updates.
Solution Approach 2:
The patent applies different quality treatment to different bit patterns by using weighted adaptation where each bit pattern contributes equally to coefficient updates. This local quality adjustment ensures that rare bit patterns receive the same attention as common patterns, preventing adaptation bias and improving overall equalization accuracy for colored signals.
Data Source
AI summary
Tools capable of improving the accuracy of decision feedback equalization (DFE) are described. The tools may adapt a DFE using a more-equal distribution of signals than those actually received. The tools may do so by disregarding, averaging, or weighting certain signals when adapting the DFE when those signals represent an unequal distribution of bit patterns. In one example, the tools detect and disregard some of the signals representing idle bit patterns that are received more often than other bit patterns. The tools may also or instead compensate for a bit pattern that is never or rarely received.