Performance Analysis And Optimization For Sidelink Communication In Unlicensed Spectrum Based On Stochastic Geometry

The broad proliferation of mobile devices and the diversity of wireless Internet applications have driven an ever-increasing demand for network capacity and spectrum resources in the fifth generation(5G)networks.However,due to the scarcity of licensed spectrum and the abundance of unlicensed spectrum resources,the utilization of unlicensed spectrum has been considered one of the promising technologies.Recently,New Radio-based access to unlicensed spectrum(NR-U),which aims to extend the applicability of NR to unlicensed spectrum as a general-purpose technology,is in a standardization process started by the 3rd Generation Partnership Project(3GPP).Sidelink(SL)transmission supports direct communications between proximal devices over the PC5 interface.Compared with NR-U,SL over Unlicensed Spectrum(SL-U)can further increase network capacity.This paper utilizes stochastic geometry to model and theoretically analyze the SL-U network,and the performance of the coexistence network is optimized based on the theoretical analysis,so as to propose reasonable coexistence schemes in unlicensed spectrum and provide theoretical references for system parameter setting.In this paper,three research contents are included,namely the modeling and dynamic resource allocation of the 3-D SL-U network,the joint power and access control for the SL-U network,and the modeling and performance analysis of the relay-assisted SL-U network.In the first part of this paper,considering the smart homes or offices in urban areas,a 3-D indoor coexistence network model is established using a homogeneous Poisson point process.The medium access probability and the spectral efficiency of a typical device are respectively derived.Based on the duty-cycle access mode,a dynamic resource allocation scheme is proposed,which maximizes the system throughput while guaranteeing fair coexistence.Secondly,a spectrum reuse mode to incorporate SL communication into the unlicensed spectrum is investigated to further improve resource utilization.The Wi-Fi network is modeled by Matem Cluster Process(MCP),and the medium access probability of SL transmitter and Wi-Fi access point(AP)are derived,as well as the spectral efficiency of a typical device when the SL transmitter can access the channel and cannot access the channel.Moreover,the SL power control and access control are combined to maximize the spectral efficiency of the coexistence system while limiting the interference between networks.A suboptimal algorithm is studied integrating neural network and K-means clustering algorithm to provide theoretical references for the power and channel threshold setting of practical SL devices.In the third part of the paper,the relay technology is introduced into the SL-U network,and the relay-assisted SL-U network transmission scheme is proposed.The medium access probability and the spectral efficiency of SL and Wi-Fi devices are derived.Finally,the dynamic adjustment strategy of the relay distance threshold is proposed to improve the total spectral efficiency of the coexistence system.