Constellation Subset Selection: Theories and Algorithms

Constellation subset selection (CSS) is a new topic emerging in the adaptive modulation communications. How to choose the subset of the original constellation phasers appears to be challenging and interesting to the researchers. There hardly exists any literature which studies the feasibility and the solution of constellation subset selection in details. Here we dedicate to this problem in both theoretical exploration and algorithm design. In this paper, we introduce the detailed theoretical analysis regarding the mathematical properties of the commonly-used constellations and then facilitate the constellation subset selection problem. The CSS problem can be formulated as the maximization of the minimum inter-phaser distance within a constellation subset subject to the symmetry and rate constraints. Based on our problem formulation, we design two algorithms to solve this problem thereby. Our complexity analysis evinces the effectiveness of the proposed algorithms.