Research On Coexistence Mechanism Between LTE And Wi-Fi

In order to meet the high-growth demand for mobile data traffic,mobile operators are seeking to use the unlicensed spectrum as a supplement to the licensed spectrum.Therefore,the current LTE/LTE-A technology is proposed to operate on unlicensed spectrum,which can greatly increase the spectrum resources and further meet current and future demand for mobile data traffic.On the one hand,it can improve the spectrum utilization of unlicensed spectrum.On the other hand,it can expand the capacity of the mobile communication system and further provide better quality of service for users.This thesis mainly focuses on the coexistence problem between LTE and Wi-Fi on unlicensed spectrum.Firstly,based on the characteristics of the licensed assisted access(LAA)mechanism recommended by 3GPP,a new two-dimensional markov chain model is proposed,and the performance of this mechanism is deeply analyzed based on the proposed mathematical model.Secondly,based on the fact that the contention window size(CWS)in LAA has a significant influence on the coexistence system,a chaotic Q-learning algorithm is proposed to intelligently select CWS.The main research work of this thesis is summarized as follows:First,the key technologies in LTE and Wi-Fi networks are introduced.In view of the fairness of resource allocation between LTE and Wi-Fi networks,the current fair coexistence scheme of LTE and Wi-Fi networks and its challenges are detailedly introduced.Second,considering the characteristics that the 3GPP LAA scheme is different from other LAA schemes and IEEE 802.11 distributed coordination function(DCF)protocol in Wi-Fi systems,the 3GPP LAA scheme is modeled as a new two-dimensional discrete Markov model in this thesis,the collision probability and the channel occupancy utilization are further derived based on the proposed model.To further analyze the performance of the 3GPP LAA scheme,we also derive the expressions of the same performance metric for the linear backoff based LAA scheme and the fixed contention window based LAA scheme.The simulation and numerical results prove the validity of the proposed Markov chain model in this thesis,and also show that the 3GPP LAA scheme is the friendliest one among the three LAA schemes to Wi-Fi networks.Third,considering the significant influence of CWS on the performance of coexistence mechanism,this thesis proposed a chaotic Q-learning based LAA scheme.Based on the proposed scheme,LAA SBSs can learn from the environment and select CWS intelligently according to the current environment.The simulation results show that the proposed chaotic Q-learning LAA scheme in this thesis outperforms the other LAA schemes such as the 3GPP,linear fixed LAA schemes and listen before talk(LBT)-adaptive scheme in terms of throughput,collision probability,fairness and delay.