A Novel Wireless Interference Identification and Scheduling Method based on Convolutional Neural Network

Abstract

Wireless interference identification plays a key role in improving the performance of mobile communication systems in terms of empowering smarter scheduling. This paper proposes to apply the convolutional neural network (CNN) to identification of wireless interference, by constructing a novel multi-level identifier which works on three different time granularities and combines the results. Exploiting the powerful feature extraction ability of CNN, the proposed approach can identify and locate 7 types of interference with high accuracy, and an adaptive threshold is calculated based on the identification result for smart scheduling. Simulation results verify that the proposed multi-level method can improve the accuracy of interference identification significantly, and achieve smart scheduling as well as increase the throughput of the network.