LTE-U Interference Coexistence Technology For Large-Capacity And Massive-Connection

With the rapid development of wireless communication technologies,the fifth generation of mobile communication technology(5G)comes into being.The main usage scenarios of 5G are grouped into three categories:enhanced mobile broadband(eMBB),massive machine-type communications(mMTC)and ultra-reliable and low-latency communications(URLLC).LTE-U(LTE in Unlicensed Spectrum)extends Long Term Evolution(LTE),which operates in licensed bands,to unlicensed bands with more abundant frequency resources.LTE-U technology can use the unlicensed band as the auxiliary carrier for traffic offloading,and make up for the shortage of licensed band resources.It is also possible to meet the high rate and low latency requirements in the large-capacity and massive-connection scenarios.Since the current unlicensed band is mainly used by Wi-Fi systems,the primary problem is to solve the coexistence problem of LTE-U and Wi-Fi systems.This thesis studies basic theories and key technologies of coexistence between LTE-U and Wi-Fi systems under the premise of prioritizing Wi-Fi system QoS in large-capacity massive-connection scenarios.The main work is as follows:(1)Based on the deep learning theory,a traffic perception and prediction scheme for Wi-Fi system is proposed.According to the self-similarity of Wi-Fi traffic,the Markov modulation Bernoulli process(MMBP)is used to model the arrival process of Wi-Fi packets.Furthermore,a traffic sensing and prediction method based on LSTM neural network is proposed.The data set is constructed by the historical data perceived by LTE-U and used to train the traffic prediction network.(2)In order to reduce the the complexity of user access selection and resource allocation methods when LTE-U and Wi-Fi systems coexist,a dynamic blank subframe resource allocation model based on reinforcement learning is proposed.The model aims to maximize the coexistence system capacity,and realizes the judgment of the state transition of the LTE-U/Wi-Fi coexistence system through the Q learning algorithm,so as to select the optimal subframe allocation mode selection according to the transition state.(3)The coexistence problem of multiple i-Fi nodes and LTE-U in dense networks is studied.In order to reduce the adjacent interference of LTE-U to Wi-Fi,a dynamic frequency selection algorithm based on interference constraint and joint utility function is proposed.Specifically,the LTE-U base station comprehensively considers the interference constraint,the channel idle coefficient and the service priority.Then the working channel is selected when the utility function is optimal.Finally,the coexistence mechanism of LTE-U and Wi-Fi based on deep learning is evaluated.The simulation results show that the proposed LTE-U adaptive blank subframe mechanism and its coexistence control mechanism with Wi-Fi system can be effective in achieving fair coexistence between the LTE-U system and the Wi-Fi system.