Bit-Edge Zero Forcing Equalizer for Low-Complexity ISI Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing communication systems face significant challenges in minimizing Inter-Symbol Interference (ISI) at bit centers, leading to performance issues and increased complexity due to the need for numerous filter taps and large coefficient values, which are difficult to implement effectively, especially in high-speed data links.

Innovation Solution

The Bit-Edge Zero Forcing Equalizer (BE-ZFE) approach, which adjusts filter tap coefficients based on communication channel characteristics to minimize ISI at bit edges rather than bit centers, allowing for fewer filter taps and smaller coefficients, thereby reducing device size and parasitic loading.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional equalization methods are used to minimize ISI at bit centers, then decision accuracy is improved, but device complexity and parasitic loading increase due to numerous filter taps and large coefficient values

Engineering Contradiction:
Improvedecision accuracyVSAvoidfilter complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the target parameter for equalization from bit center ISI minimization to bit edge ISI minimization. This parameter change allows the use of fewer filter taps (e.g., 3 taps instead of traditional larger numbers) with smaller coefficient values, thereby reducing device complexity and parasitic loading while still achieving effective ISI cancellation through the modified equalization target

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent extracts and addresses only the critical ISI components at bit edges rather than attempting to minimize ISI across the entire bit period. By focusing equalization efforts specifically on bit edge zero-crossings, the system achieves effective ISI cancellation with reduced filter complexity, eliminating the need for numerous filter taps required by traditional approaches

Inventive Principle:
Principle #2Taking out (Extraction)

2Reliability

If more filter taps are used to reduce ISI, then signal quality is improved, but hardware implementation becomes more difficult and parasitic loading increases

Engineering Contradiction:
Improvesignal qualityVSAvoidimplementation difficulty
Core Design Contradiction:
ReliabilityVSEase of manufacture

Solution Approach 1:

By changing the equalization target from bit center to bit edge, the patent achieves effective ISI cancellation with fewer filter taps and smaller coefficient values. This parameter change simplifies hardware implementation, reduces parasitic loading, and makes the system more manufacturable while maintaining signal quality through the focused bit edge zero-crossing approach

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

Instead of trying to minimize ISI at the traditional bit center sampling point, the patent inverts the approach by targeting bit edges for ISI minimization. This inversion allows the use of simpler filter structures with fewer taps, making hardware implementation easier while still achieving the essential goal of ISI cancellation through the alternative sampling strategy

Inventive Principle:
Principle #13The other way round (Inversion)

Data Source

PatentUS7653127B2Bit-edge zero forcing equalizer
Publication Date: 2010.01.26 XILINX INC
  • US7653127B2 patent drawing
  • US7653127B2 patent drawing
  • US7653127B2 patent drawing

AI summary

Bit-Edge Zero Forcing Equalizer. A novel solution is presented by which a BE-ZFE (Bit-Edge Zero Forcing Equalizer) is employed to drive an error term within a data signal to an essentially zero value. This new BE-ZFE looks at values of data that occur at the bit edges of a data signal and drives the associated error term to zero. The new BE-ZFE is appropriately implemented within communication systems that are phase (or jitter) noise limited. Some examples of such communication systems include high-speed serial links one type of which serviced using a SERDES (Serializer/De-serializer) where data that is originally in a parallel format is serialized into a serial data stream and then subsequently de-serialized back into a parallel data stream.