Distributed Spectrum Sensing Structure Design Based on Distributed Deep Neural Network

Computational resources and environmental interference are practical issues that restrict the improvement of spectrum efficiency for IoT devices. A distributed spectrum sensing structure based on distributed deep neural network is proposed in this paper, with the basic idea of using information entropy to determine the spectrum sensing effectiveness of individual devices, and achieving joint spectrum sensing by deploy parts of the neural network on multiple platforms, thereby providing accurate input for efficient spectrum application strategy generation.