STUDY OF DIGITAL MODULATION RECOGNITION IN UNDERGROUND MINE CHANNEL BASED ON HIGHER ORDER CUMULANT

摘要

Coal mine is still one of the main energy sources in many countries in the world,and its safety production has always been the most concerned problem of coal enterprises.In order to determine the signal modulation in the underground mine environment,an algorithm based on higher order cumulant is proposed in this study.In which,the Deep Neural Network (DNN) model is adopted and the method can be used to effectively identify the variety of digital modulation signals in mine fading channel.Considering the recognition problem of many kinds of digital modulation,such as MASK,MPSK,MFSK,MQAM,OFDM in underground mine environment,we analyze the influence of Nakagami fading channel and shadow fading channel on higher order cumulant.The expression of higher-order cumulant is derived after the downhole fading channel,which is used to construct the training DNN model of characteristic parameters.Simulation results show that the classification performance based on higher order cumulant and DNN model is excellent at low SNR.The correct recognition rate of the proposed method is more than 90% when SNR is - 2dB and 100% when SNR is more than 5dB in two kinds of mine fading environments.