Performance enhancement of 3D-GS-CAP based on improved swarm intelligence algorithm and spherical Fibonacci grid in VLC system

This paper proposes an optimization scheme for three-dimensional geometrically shaped carrier-less amplitude phase (3D-GS-CAP) modulation based on the multi-strategy improved honey badger algorithm (MIHBA) and the spherical Fibonacci grid (SFG), with its performance experimentally verified in a visible light communication (VLC) system. To reduce the geometric complexity of artificial constellation design, the scheme uses the uniform distribution of the spherical surface and the swarm intelligence algorithm to optimize the three-dimensional (3D) spherical surface. Aiming at the problem that the HBA algorithm is prone to local optimality, four improvement strategies are proposed in this paper to optimize the 3D constellation distribution and improve the anti-noise ability of the signal. In addition, in order to solve the high computational complexity of existing optimization algorithms, SFG is introduced innovatively in this paper. The universality of the method to constellation points makes it efficient to optimize under different orders without complicated parameter adjustment. The experimental results show that when the bit error rate threshold is 5E^-4, the maximum Vpp operating range of 3D-MIHBA is extended by 59.98 % and the bias current is increased by 23.4 mA compared with the traditional 3D-CAP-16 signal. This shows that the scheme enhances the anti-noise capability of the system and greatly improves the flexibility of constellation optimization under low complexity.