This work presents a multidimensional signal constellation framework for reliable low-SNR wireless communication based on hypercube decision regions and increased minimum Euclidean distance between signal vectors. The proposed approach investigates geometric methods for improving energy efficiency in communication systems without relying exclusively on long error-correcting codes. Analytical estimates of error probability and energy efficiency are derived for multidimensional signal constellations with large normalized signal duration. The paper discusses the relationship between signal dimensionality, decision-region geometry, and communication reliability, and compares the proposed framework with conventional spherical-shell signaling approaches. Potential applications include energy-efficient wireless systems, low-SNR communications, satellite communications, and high-reliability data transmission systems.
Bykhovskiy et al. (Wed,) studied this question.