Deployment Of Uavs In Emergency Communication Networks

Generally,the ground communication infrastructures are destroyed to varying degrees after the disasters.Existing infrastructures can not satisfy the urgent and huge communication demand,which results in network congestion and increase of communication interruption probability.The rescue work after the disaster highly relies on the communication network.Therefore,rapid establishment of a reliable emergency communication network for the disaster-affected area is vital to rescue work.Due to the characteristics of high mobility,easy deployment,and economical price of unmanned aerial vehicle(UAV),UAV-assisted communication network is considered to be an effective solution for emergency communications after a disaster.However,limited battery capacity of UAVs and the nondeterminacy of user distribution in the post-disaster scenario bring huge challenges to the design of UAV base station(BS)deployment strategy.This thesis studies the trajectory planning problem of UAV BSs in emergency communication networks,and proposes effective algorithms for different performance goals.The main contributions of this thesis are as follows:1.We construct the system model to describe the UAV-assisted emergency communication network,which considers the special user distribution after the disaster.We also propose appropriate performance metrics to measure the performance of the proposed algorithm.2.We study a UAV trajectory planning problem when the user distribution is unknown.The goal is to choose the position where the UAV hover each time to achieve the maximum system throughput while satisfying the power constraint.Since the user distribution is unknown,we adopt the online learning algorithm to carry out real-time trajectory planning,which means the user distribution information obtained after selecting the hovering position each time is used to guide the subsequent selection.We transform the problem into a multi-armed bandit(MAB)problem,and propose two distant-aware trajectory planning algorithms.Simulation results show that the proposed algorithms can effectively improve the system throughput.3.We study a trajectory planning problem for multiple UAVs.The goal is to maximize the coverage of ground users while satisfying the power constraints.We propose an iterative algorithm based on the idea of power balancing,which makes full use of all the UAVs by allocating the workloads in a balanced way.The algorithm assigns ground users to each UAV,thus transforming the multi-UAV trajectory planning problem into multiple independent travelling salesman problems(TSP),and reassigns the workloads to further balance the power consumption.Simulation results show that the proposed algorithm can effectively improve the coverage of ground users.In summary,this thesis studies the deployment of UAV BSs in emergency communication networks,and proposes effective algorithms according to different performance goals.Numerical results show that the proposed algorithms can effectively improve the performance of the emergency communication network.