Optimizing satellite broadcast scheduling problem using the competitive Hopfield neural network

In this paper the competitive Hopfield neural network method for finding a broadcasting schedule in the satellite system will be described. The satellite broadcast scheduling (SBS) problem is known as an NP-complete problem. Communication links between satellites and ground terminals are provided in a repetition of time slots. The goal of the proposed algorithm is to find the broadcasting schedule of satellites with the maximum number of broadcasting time slots under the constraints. A competitive learning rule provides a highly effective means for obtaining a resonance solution and is capable of reducing the time-consuming effort to obtain coefficients. The proposed method can always satisfy the problem constraints and guarantee the viability of the solutions for the SBS problem. The competitive mechanism simplifies the network complexity. The proposed method is greatly suitable for implementation on a digital machine. Furthermore, the competitive Hopfield neural network method permits temporary energy increases to escape from local minima. Simulation results show that the competitive Hopfield neural network method can improve system performance and with fast convergence and high reliability.