Weighted Adaptive Equalizer for Transition Error Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing adaptive equalizers do not effectively address the higher error rates of certain received data, particularly those with level transitions, leading to increased system error rates in communication systems.
Innovation Solution
A weighted adaptive equalizer that generates a weighted signal based on the expected error rate and utilizes this signal to adjust equalization coefficients, along with a projecting unit to filter noise, thereby reducing the averaged error rate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a fixed coefficient adjustment factor τ is used in the LMS operation, then the equalization coefficients can be updated systematically, but the system cannot adapt to different channel environments effectively, leading to higher error rates for certain received data
Solution Approach 1:
The patent makes the coefficient adjustment factor dynamic by introducing a weighted signal that varies according to the characteristics of received data. The weighted signal generator produces different weights for different received signals based on their expected error rates, allowing the LMS operation to adapt its step size dynamically. This resolves the contradiction by transforming the fixed τ into a variable parameter that adapts to different channel conditions and data types.
Solution Approach 2:
The patent changes the parameter τ from a fixed value to a variable weighted signal that reflects the expected error rate of received data. By modifying this key parameter based on data characteristics (such as level transitions), the system achieves better adaptability while maintaining systematic coefficient updates. The weighted signal effectively modulates the adjustment factor to optimize performance for different received signals.
2Loss of time
If the coefficient adjustment factor τ is set to a greater number for quick convergence, then the equalization coefficients enter stable state faster, but the probability of non-convergence increases and system error rate rises
Solution Approach 1:
The patent dynamically adjusts the coefficient adjustment factor based on the weighted signal that reflects the expected error rate. During initial convergence phases, the weighted signal may allow larger adjustments for faster convergence, while during stable operation or when errors are detected, it reduces the adjustment factor to ensure convergence reliability. This dynamic behavior resolves the contradiction between convergence speed and reliability.
Solution Approach 2:
The patent transforms the fixed adjustment factor τ into a variable parameter that changes based on system state and error conditions. The weighted signal modulates this parameter to achieve optimal convergence behavior - allowing aggressive updates when appropriate and conservative updates when reliability is concerned. This parameter transformation enables the system to balance convergence speed and reliability adaptively.
3Device complexity
If uniform equalization coefficients are used for all received signals, then the equalization unit structure remains simple, but certain received data with level transitions experience higher error rates
Solution Approach 1:
The patent applies local quality by introducing differentiated weighted signals for different received data characteristics. Instead of treating all received signals uniformly, the weighted signal generator produces specific weights for specific data types (e.g., higher weights for signals with level transitions that are more prone to errors). This allows the system to maintain simple overall structure while achieving localized optimization for problematic signal types.
Solution Approach 2:
The patent changes the uniform coefficient application into a differentiated parameter approach where the weighted signal modulates the effective adjustment for different received signals. The underlying equalization coefficient structure remains simple, but the weighted parameter variation enables targeted improvement for specific data types without complicating the overall system architecture.
Data Source
AI summary
An adaptive equalizer and the related method are disclosed. The adaptive equalizer is capable of adjusting its own equalization coefficients, and includes a reference signal generator for generating a reference signal according to a first reference source, an equalization unit for generating an equalized signal by processing a received signal through a plurality of equalization coefficients, a weighted signal generator for generating a weighted signal according to a second reference source, and a coefficient adapting circuit for adjusting the equalization coefficients according to the reference signal, the equalized signal, the weighted signal, and the received signal.