QAM Blind Equalizer Using Multiple Modulus Regions for Faster Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional blind equalizers for Quadrature Amplitude Modulation (QAM) receivers face challenges in convergence speed, especially in severe channel conditions with high constellation sizes, as they rely on a single constant modulus criterion that slows down adaptation to optimal tap weights.
Innovation Solution
Implementing a blind equalizer with multiple constant modules that divide the QAM constellation into subgroups with different radii, using a coarse decision engine to select the closest pre-calculated modulus for error calculation, thereby improving convergence by reducing the adaptation error signal magnitude.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a single constant modulus criterion is used in conventional blind equalizers, then the system overhead is reduced and blind equalization is achieved, but the convergence speed becomes slow especially in severe channel conditions with high constellation sizes
Solution Approach 1:
The QAM constellation is segmented into multiple sub-constellations based on symbol radius groups. Instead of using a single constant modulus criterion for all symbols, the equalizer divides the constellation space and applies separate constant modulus values (K1, K2, ..., Km) to different radius groups. This segmentation allows the equalizer to converge faster by providing more targeted error correction for each subgroup, resolving the contradiction between convergence speed and structural complexity.
2Productivity
If higher QAM sizes are adopted to meet increasing bandwidth demand, then the data transmission capacity increases, but the difficulty of convergence becomes greater and more difficult under severe channel conditions
Solution Approach 1:
For high-order QAM constellations (64-QAM, 256-QAM, etc.), the invention segments the large constellation into multiple smaller sub-constellations based on symbol radius. Each sub-constellation is assigned a specific constant modulus value, making the convergence process more manageable. This segmentation reduces the difficulty of detecting and measuring the correct symbol values, allowing the equalizer to handle higher QAM sizes effectively while maintaining convergence capability under severe channel conditions.
Solution Approach 2:
The invention applies different constant modulus values (local quality parameters) to different regions of the constellation diagram based on symbol radius groups. Instead of using a uniform approach for all symbols, each radius group receives a tailored constant modulus value that optimizes the error correction for that specific region. This local quality approach enables the equalizer to maintain high productivity with higher QAM sizes while reducing convergence difficulty.
Data Source
AI summary
Methods and apparatuses for blind equalizers with multiple constant modules. In one embodiment, a circuit, includes: a filter to produce an output based on an input that represents a symbol being received, the symbol being one of a Quadrature Amplitude Modulation (QAM) constellation; a decision engine coupled to the filter to generate a result indicating one region of a plurality of regions in a QAM constellation diagram, the output of the filter being in the indicated region which includes a plurality of symbols of different radii in the constellation diagram; and an error reduction engine coupled to the decision engine and the filter to reduce a difference between a selected one of a plurality of constants and a modulus of the output; where each of the plurality of constants correspond to one of the plurality of regions; and the selected one of the plurality of constants is selected according to the result of the decision engine.