Research On Spectrum Utilization Methods For LTE-LAA Networks

The contradiction between the growing mobile users as well as the explosive data amount and the exhausted licensed spectrum resources has raised a high demand for improving the spectrum efficiency and extending the usable frequency range,which gives birth to the Long Term Evolution on Unlicensed spectrum(LTE-U)and Licensed-Assisted Access(LAA).However,the existing academic researches on LTE-LAA system mainly focus on the performance analysis and optimization based on the perfect spectrum sensing while the sensing errors in actual scenario are inevitable and will cause the transmission collision,then deteriorate the system performance.In this thesis,the throughput analysis and optimization of LTE-LAA networks with imperfect spectrum sensing are firstly studied.The unlicensed spectrum is modeled as a M/M/1 quequing system,and then the probabilities of two channel states and four state transition probabilities are described,based on which the detection probability and false-alarm probability are deduced.Next,in the unsaturated Wi-Fi system and LTE-LAA system,the channel access process with binary exponential backoff are modeled by Discrete-Time two-dimensional Markov Chain(DTMC),according to which the transmission probability and channel occupied probability of LAA users and Wi-Fi users are analyzed.Finally,the throughput of LTE-LAA system is maximized by optimizing the sensing duration and sensing threshold under the interference constraint.Simulation results confirm the significant impact of the imperfect spectrum sensing on the LTE-LAA networks,and show that the optimization of sensing duration and sensing threshold can effectively improve the throughput of LTE-LAA system.Furthermore,considering the different energy supply and computing ability of base stations(BS)and users(UE)in LTE-LAA system,this thesis designs a joint BS-UE spectrum detection framework,and then analyzes and optimizes the LTE-LAA network throughput with BS spectrum prediction assisting.In the joint BS-UE spectrum detection framework,LAA BS performs spectrum sensing and predicting based on Hidden Markov Model(HMM)at the beginning.Next with the predicting sequences and predicting accuracy from LAA BS,LAA UE detects the unlicensed spectrum based on the priori knowledge.The throughput of LTE-LAA system is analyzed and optimized with a model similar to the previous subsection and simulation results show that the introduction of BS spectrum prediction can help LAA UE improve the accuracy of spectrum sensing and reduce the collision probability.Finally,the adaptive frame structure in LAA-D2D network and the user synchronous channel access method based on pseudo random sequence are designed.The LAA frame is divided into dedicated slot,spare time slot and shared slot,and different resource allocation strategy and user access control mechanism are adopted for three slots to improve the resource utilization rate of unlicensed spectrum.For the synchronized channel access of D2D users,with the same random seeds assigned by the BS,the sender and receiver generate the same pseudo random sequence according to the same algorithm,and then synchronizedly access to the same mini resource block(MRB)on the unlicensed spectrum.The simulation results show that the time slot division of LAA-D2D frame structure can improve the resource utilization of unlicensed spectrum and the synchronous access mechanism of D2D users based on pseudo random sequence can increase the access probability of users.