A Neurodynamic Approach for SWIPT Power Splitting Optimization

Abstract Simultaneous wireless information and power transfer (SWIPT) systems using energy from RF signals can effectively solve the shortage of devices. However, existing SWIPT optimization methods numerical algorithms are difficult to non-convex problem adapt dynamic communication circumstances. In this paper, a duplex neurodynamic method is used address system’s partitioning issue. The rate maximization system framed as biconvex problem. A recurrent neural network concurrently execute local search update initial state by particle swarm get global optimum. experimental results demonstrate that neurodynamic-based maximizes while satisfying minimal harvesting requirement in variety channel states.