CNN-based idle spectrum detection for 5G broadcasting services

Abstract

With the rapid growth of consumer demand for high-quality audio-visual media content such as 4K/8K and VR/AR., broadcasting services are becoming diversified and refined and the existing broadcast television network is difficult to meet the demand. 5G link will become an important communication channel for broadcasting services. Among them, the detection of idle spectrum is particularly important. In this paper, we use a convolutional neural network (CNN) module to solve the task of detection of idle spectrum. Experiment demonstrates that the CNN module is better than the traditional energy detection (ED) module. We compare CNN with ED when the signal-to-noise ratios (SNR) varies constantly to validate the scheme proposed in this paper.