Decision Feedback Equalizer Precursor ISI Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing decision feedback equalization (DFE) techniques, such as feed-forward equalization (FFE), face limitations in reducing precursor inter-symbol interference (ISI) as they can amplify noise and crosstalk, necessitating an improved method for effective precursor ISI reduction in integrated circuit (IC) devices.

Innovation Solution

The proposed solution involves an enhanced decision feedback equalizer (EDFE) with a precursor cancellation block that uses a comparison and selection circuit to subtract weighted postcursor decisions from the analog input signal, comparing the signal against various threshold inputs to select the appropriate digital output, thereby reducing precursor ISI while minimizing noise and crosstalk amplification.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If feed-forward equalization (FFE) is used to reduce precursor ISI, then precursor ISI reduction is achieved, but noise and crosstalk are amplified

Engineering Contradiction:
Improveprecursor ISIVSAvoidnoise and crosstalk amplification
Core Design Contradiction:
Object-affected harmful factorsVSObject-generated harmful factors

Solution Approach 1:

The patent employs decision feedback equalization (DFE) where decisions made on previous symbols are fed back to cancel their interference on current symbols. The postcursor decisions are subtracted from the analog input signal to reduce precursor ISI without the noise amplification problem of FFE, as the feedback mechanism uses already-decoded information to predict and cancel interference.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent extracts and separates the postcursor interference component from the received signal by identifying and subtracting the weighted postcursor decisions from the analog input signal. This isolation allows precise cancellation of the harmful precursor ISI while leaving the desired signal and noise components unaffected.

Inventive Principle:
Principle #2Taking out (Extraction)

2Object-affected harmful factors

If multiple comparators and selection circuits are added to implement precursor cancellation, then precursor ISI reduction improves, but device complexity increases

Engineering Contradiction:
Improveprecursor ISI reductionVSAvoidcomparison and selection circuit complexity
Core Design Contradiction:
Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The patent divides the equalization function into separate blocks: a subtraction block for postcursor cancellation, a postcursor decision block for generating decisions, and a precursor cancellation block with comparators for threshold-based selection. This segmentation allows each block to perform a specific function efficiently, reducing overall system complexity while achieving effective precursor ISI reduction.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent uses dynamic threshold selection where the comparison thresholds are adaptively determined based on the postcursor coefficient h1 and precursor coefficient hm1. The selection circuit dynamically chooses among multiple threshold inputs (-h1-hm1, +h1-hm1, -h1+hm1, +h1+hm1) based on the current signal conditions, allowing the system to optimize performance without requiring a fixed complex circuit structure.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3295631B1Decision feedback equalization with precursor inter-symbol interference reduction
Publication Date: 2020.08.19 XILINX INC
  • EP3295631B1 patent drawingFigure 1
  • EP3295631B1 patent drawingFigure 2
  • EP3295631B1 patent drawingFigure 3

AI summary

In a receiver (100), a decision feedback equalizer (120) provides weighted postcursor decisions (121) to a subtraction block (122) for subtraction from an analog input signal (101) to provide an analog output signal (123). A postcursor decision block (130) compares the analog output signal (123) against positive and negative values (104, 105) of a postcursor coefficient for providing first and second possible decisions (136, 137) for selecting a current postcursor-based decision (116) therebetween responsive to a previous postcursor-based decision (117). A precursor cancellation block (108) receives the analog output signal (123), the previous postcursor-based decision (117) and the current postcursor- based decision (116) for providing a digital output signal (124) for a previous sample of the analog input signal (101). The precursor cancellation block (108) includes comparators (211 -214) for receiving the analog output signal (123) and for respectively receiving threshold inputs (201 -204) different from one another for providing possible digital outputs (215-218) for the analog output signal (123). The selection stage (230) is coupled for receiving the possible digital outputs (215-218) for selection of the digital output signal (124).