Research on spectrum prediction technology of BP neural network based on genetic algorithm optimization

In order to use the electromagnetic spectrum more reasonably and effectively, and make users use the spectrum access more effectively, this paper first introduces the basic process and the basic elements of genetic algorithm. Because genetic algorithm is a multiple iterative search algorithm, it optimizes BP neural network, and uses those individuals with better adaptability through selection, crossover, mutation and other operations, that is, to optimize and select the optimal weight and threshold of BP neural network, improve the accuracy of spectrum prediction, reduce prediction error, and verify the effectiveness of the algorithm through simulation.