Rhombus-Shaped Constellations for Lower-Error Symbol Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data transmission methods suffer from high bit error ratios and low accuracy due to the mismatch between the designed decision regions in constellation diagrams and the true distribution of received symbols, especially in scenarios with AWGN and multiplicative channels with phase noise.
Innovation Solution
Redesign the constellation diagram to match the true distribution of received symbols by using rhombus-shaped decision regions for two bits and nested rhombus-shaped regions for (2+M) bits, ensuring unequal diagonal lengths and non-intersecting rhombuses, and demodulate based on log-likelihood ratios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a uniform ring or sector decision region is used in conventional constellation diagrams, then the demodulation process is simple, but the bit error ratio is high and transmission accuracy is low due to mismatch with the true distribution of received symbols
Solution Approach 1:
The patent applies asymmetry by transforming the conventional symmetric uniform ring or sector decision regions into asymmetric rhombus-shaped decision regions. The rhombus shape with unequal diagonal lengths better matches the asymmetric true distribution of received symbols caused by phase noise and AWGN, thereby improving transmission accuracy while maintaining reasonable demodulation complexity
Solution Approach 2:
The patent implements local quality by creating different decision region shapes for different symbol groups. Specifically, it divides the constellation into multiple rhombuses with different orientations and shapes, where each rhombus corresponds to a specific bit combination. This localized optimization allows each decision region to match the local symbol distribution characteristics, improving overall demodulation accuracy
2Ease of manufacture
If the constellation diagram is designed with symmetric uniform regions, then the modulation scheme is simple to implement, but it cannot adapt to the skewed distribution of received symbols under phase noise and AWGN
Solution Approach 1:
The patent applies dynamics by making the decision region boundaries adaptive to the channel conditions. The rhombus-shaped decision regions are designed with specific geometric properties that dynamically adjust to the skewed symbol distribution caused by phase noise. The decision boundaries are no longer fixed uniform sectors but are shaped to follow the probability density contours of the received symbols under realistic channel conditions
Solution Approach 2:
The patent implements parameter changes by modifying the geometric parameters of the decision regions. Instead of using uniform ring or sector shapes with equal angular widths, the patent uses rhombus shapes with unequal diagonal lengths and specific vertex positions. These parameter changes allow the decision regions to better capture the skewed distribution of received symbols, improving adaptability to phase noise and AWGN while maintaining a systematic modulation framework
Data Source
AI summary
Provided are a symbol sending method, a symbol receiving method, a sending device, a receiving device, and a storage medium. The symbol sending method includes the following: modulating two bits according to a model of a first constellation diagram to obtain a transmission symbol, where the model of the first constellation diagram satisfies the following: four vertices of the first constellation diagram form a rhombus, the lengths of the two diagonal lines of the rhombus are unequal, and each vertex corresponds to one value of the two bits; and sending the transmission symbol.