| Network throughput has always been the most important performance metric of traditional wireless communications.However,a large number of new applications are emerging,such as industrial automation,high-precision remote control technology,etc.,which have more stringent requirements of communication reliability and delay.Due to the short length of the transmitted block of these applications,the traditional wireless transmission technology is no longer applicable.To deal with this challenge,the Ultra-reliable and lowlatency communication(URLLC)technology provides an effective solution.This technology is based on the theory of finite block-length information,which have a more accurate description of short-term information transmission.On the other hand,unmanned aerial vehicle UAV-based wireless communication has recently been considered as a communication assistance technology for next-generation mobile communication networks with great application potential.Compared with traditional solutions,UAV-based wireless communication has many new advantages,such as flexible on-demand deployment,highly controllable 3D airspace,line-of-sight characteristics of air-ground channels and high network flexibility.When URLLC is combined with UAV-assisted technology,it will produce huge system advantages.First,URLLC will make the UAV-assisted communication network more accurate and efficient,so it will be suitable for more novel scenarios.Then,UAV-assisted technology will enhance the coverage of URLLC networks.Therefore,studying how to make full use of the advantages of the two technologies to achieve URLLC based on UAV-assisted technology has important practical and research value.Based on the above reasons,this paper studies a UAV-assisted communication network with URLLC,in which one UAV acts as a base station to collect information from a set of ground sensors,and then send the processed data(e.g.control signals)to their respective ground actuators.The main contents of this thesis are as follows:1)First,we investigate the offline algorithm to jointly design the UAV’s location and resources optimization allocation for UAV-assisted URLLC in Internet of Things(Io T)network.Our goal is to minimize the maximum packet error rate of the sensor-actuator communication pair by jointly optimizing the UAV’s position,communication latency and power allocation,subject to the latency and the UAV’s maximum total power constraints to improve the performance of the communication system.This problem is a non-convex problem,which is generally difficult to be solved.We divide it into three sub-problems and solve them iteratively,and finally obtain the local optimal solution of the original problem.The simulation results show that the optimal allocation of wireless resources can compensate the disadvantages caused by the difference of UAV’s deployment,and reveal that in a two-hop communication system,the combined uplink and downlink optimized system performance will be significantly improved compared to the performance of the system with single link optimization.2)Then,we extend the research in Part I to the mobile ground equipment scenario,and propose an online algorithm of UAV trajectory and wireless resource optimization allocation based on reinforcement learning with imperfect network state information.When the ground equipment can move freely,the traditional optimization method is no longer applicable since its movement state and channel information are unpredictable.Therefore,we apply novel machine learning methods to solve this problem.Specifically,we discretize the UAV’s flight duration into specific time slots,use traditional optimization methods to allocate the UAV transmit power in each time slot,and use machine learning methods to select the UAV’s flight direction between adjacent time slots.Then we can obtain a local optimal solution of the original problem in an iterative manner.Finally,the numerical simulation results verify the effectiveness of our proposed algorithm,which shows that the reinforcement learning algorithm can be successfully applied to the optimization of the UAV trajectory and UAV can effectively avoid obstacles and choose a trajectory with greater gain.In the context of the Io T applications,this thesis conduct online and offline research on UAV-assisted URLLC communications.The research results show that the combination of these two technologies can effectively improve the system’s performance.Furthermore,we look forward the development of this technology,e.g.,with UAV-based computing,large-scale UAV group control technology. |
