Geometric Constellation Shaping

Technical Explanation

Standard QAM constellations place points on a rigid square grid (e.g., a 16-point square). Information theory proves that the optimal input distribution for a Gaussian channel is a circularly symmetric Gaussian cloud. Geometric Constellation Shaping moves points away from square grids into concentric rings and multi-dimensional spheres, delivering up to 1.53 dB of shaping gain (the ultimate 'ultimate shaping gain' in communication theory) without changing transmitter power.

Key Functions

• Placing signal constellation points in non-uniform concentric rings and multi-dimensional spheres
• Capturing up to 1.53 dB of theoretical shaping gain over conventional square QAM
• Significantly reducing required transmit power for battery-constrained mobile devices
• Optimized multi-dimensional (4D/8D) constellations for coherent optical and Sub-THz links
• Standardized modulation formats for extreme spectral efficiency in 6G physical layers