Genetic algorithm aided gray-APSK constellation optimization

The amplitude phase shift keying (APSK) constellation with Gray labeling (Gray-APSK) is able to achieve considerable shaping gains in comparison with the quadrature amplitude modulation (QAM) counterpart under both iterative and non-iterative scenarios. Nevertheless, the Gray-APSK constellation with the fixed radius cannot provide the optimal performance for every code rate of practical usage. In this paper, we propose a method based on the genetic algorithm to optimize the parameters of Gray-APSK constellations by maximizing the mutual information for the targeted code rate. Certain performance gains can be achieved via optimizing the radius of the Gray-APSK, as verified by both the average mutual information (AMI) analysis and the error performance evaluation.