Lotka-Volterra distributed power control model for OCDMA systems

Dynamic continuous Lotka-Volterra model is used to build two new distributed power control algorithms (DPCA) for optical code division multiple access systems (OCDMA) communication systems, namely DPCA-LV and DPCA-LV* algorithms. The DPCAs features are described, while comparative performance analyses are carried out involving both proposed LV-based DPCAs, the classical Sigmoid DPCA (by Uykan and Uykan), and the Matrix inversion method (MIM) by Tarhuni, including the convergence speed, the normalized mean square error (NMSE), the average consumption of energy per user and complexity measured by the floating-point operations (flops). Under conditions of estimated errors, the proposed DPCA-LV* exhibits minor discrepancy regarding the ideal power vector solution and better convergence, considering both fixed and adaptive factor convergence regarding the DPCA-LV, the conventional Sigmoid DPCA, and MIM approach. The superiority of the DPCA-LV* in terms of performance-complexity tradeoff is more remarkable under high dimension optical networks.