Resource Allocation For High-throughput Solar-powered UAV Based On Power And Spectrum Joint Cognition

With the reduction of costs and the development of related technologies,UAVs have aroused widespread interest in the wireless communication industry due to their high mobility and line-of-sight communication characteristics.In order to achieve more efficient spectrum utilization and high-throughput communication,this thesis took the Solarpowered Unmanned Aerial Vehicle(SUAV)communication system as the research object,taking into account the random arrival characteristic of solar energy and the random communication channel fading during the flight.This thesis proposed a resource allocation algorithm that maximizes the system throughput.First of all,this thesis introduced the power cognition technology.Specifically,in each time slot,SUAVs will perform three power cognition tasks including power perception,power decision-making,and power movement in sequence,so that SUAVs can perceive changes of solar energy density through interaction with the surrounding environment,and make decisions related to flight and data transmission action based on the learned knowledge.Subsequently,this thesis used the reinforcement learning mechanism to describe this decision-making process,and proposed a power cognition SUAV resource allocation scheme in dynamic environment.This thesis jointly optimized the SUAV’s flight trajectory,energy harvesting and data transmission.The simulation results show that the proposed power cognition technology can improve the throughput and solar energy utilization of the SUAV system.Furthermore,this thesis established a system model of SUAV-assisted cellular communication based on power and spectrum joint cognition,taking into account the random arrival of solar energy,SUAV flight status,quality of service(Qo S)requirements,and energy causality constraints.This thesis jointly optimized the SUAV’s threedimensional flight trajectory,flight speed,energy capture,transmit power,and time allocation factors,the object function is to maximize the communication throughput of the SUAV system.In view of the non-convexity of the optimization problem,after equivalent simplification,this thesis performed the Constrained Stochastic Successive Convex Approximation(CSSCA)algorithm to convert it into a standard form of convex optimization problem,and propose an iterative resource allocation algorithm to achieve high-throughput SUAV system accordingly.Finally,the simulation of the proposed algorithm and two baseline strategies are performed,thus confirmed the effectiveness and superiority of the algorithm proposed in this thesis.The final flight simulation results of SUAV system revealed the trade-off between solar energy harvesting and high-throughput wireless communication.