Multi-Tap Decision Feed-Forward Equalizer for ISI and Crosstalk
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional SerDes architectures face limitations in coupling between clock data recovery (CDR) and equalization adaptation, leading to sub-optimal CDR locking points, sensitivity to transmitter settings, and increased complexity due to the dominance of Feedforward Equalizers (FFE) in noise and crosstalk correction, while Decision Feedback Equalizers (DFE) are limited to postcursor ISI correction and suffer from error propagation.
Innovation Solution
A digital signal processing (DSP) SerDes receiver architecture integrating a Decision Feedforward Equalizer (DFFE) with joint auto-adaptation for optimal signal shaping, decoupling CDR and equalization adaptations, and utilizing a multi-tap DFFE for both precursor and postcursor ISI correction to achieve symmetric pulse responses and improved noise resilience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If FFE is used for both precursor and postcursor ISI correction, then ISI correction capability is improved, but noise and crosstalk amplification increases
Solution Approach 1:
The patent divides the equalizer into two distinct segments: FFE for precursor ISI correction and DFE for postcursor ISI correction. This segmentation allows each equalizer type to operate in its optimal performance range, with FFE handling precursor interference without excessive noise amplification and DFE handling postcursor interference without error propagation, thereby resolving the contradiction between ISI correction capability and noise/crosstalk amplification
Solution Approach 2:
The patent changes the operational parameters by limiting FFE to precursor taps only and DFE to postcursor taps only, rather than allowing both to handle both types of ISI. This parameter change optimizes the noise performance while maintaining comprehensive ISI correction capability across both precursor and postcursor intervals
2Object-affected harmful factors
If DFE is used for postcursor ISI correction, then noise amplification is avoided, but error propagation occurs
Solution Approach 1:
The patent extracts the error propagation issue from the overall equalization system by isolating DFE's function to postcursor ISI correction only, where its decision-based approach is most effective. By taking out DFE from precursor ISI correction, the patent eliminates the source of error propagation while maintaining noise performance benefits in the postcursor domain
3Reliability
If multiple taps of DFE are used for postcursor ISI correction, then ISI correction capability is improved, but implementation complexity increases exponentially
Solution Approach 1:
The patent segments the equalization function so that DFE handles only postcursor ISI correction with limited taps, while FFE handles precursor ISI correction. This segmentation prevents the exponential complexity growth that would occur if DFE were used for multiple taps of both precursor and postcursor correction, as FFE's linear complexity structure is used for the majority of the equalization taps
4Reliability
If FFE dominant architecture is used, then precursor and postcursor ISI correction is achieved, but coupling between CDR and equalization adaptation increases
Solution Approach 1:
The patent segments the equalization adaptation into two independent processes: FFE adaptation for precursor ISI and DFE adaptation for postcursor ISI. This segmentation decouples the adaptation mechanisms, allowing each equalizer to be optimized independently without the complex interactions that occur in FFE-dominant architectures where a single adaptation process must handle both precursor and postcursor correction
Data Source
AI summary
A multi-tap Differential Feedforward Equalizer (DFFE) configuration with both precursor and postcursor taps is provided. The DFFE has reduced noise and/or crosstalk characteristics when compared to a Feedforward Equalizer (FFE) since DFFE uses decision outputs of slicers as inputs to a finite impulse response (FIR) unlike FFE which uses actual analog signal inputs. The digital outputs of the tentative decision slicers are multiplied with tap coefficients to reduce noise. Further, since digital outputs are used as the multiplier inputs, the multipliers effectively work as adders which are less complex to implement. The decisions at the outputs of the tentative decision slicers are tentative and are used in a FIR filter to equalize the signal; the equalized signal may be provided as input to the next stage slicers. The bit-error-rate (BER) of the final stage decisions are lower or better than the BER of the previous stage tentative decisions.


