Joint Position and Resource Optimization for Multi-UAV-Aided Relaying Systems

Qinbo Chen

doi:10.1109/access.2020.2965162

Abstract

Unmanned aerial vehicles (UAVs) can enhance the overall performance of terrestrial communication systems due to their high possibility of the line of sight (LoS) links and on-demand deployments. The advantages of the UAV position optimization now attract attentions from researchers to study the cooperation among static UAV relays. In this paper, we study a relaying system, where multiple UAVs establish a cooperative UAV relay network to help some transmitters (Txs) communicate with their corresponding receivers (Rxs) by employing orthogonal frequency division multiple access (OFDMA). We maximize the paired Tx-Rxs' minimum rate through solving a non-convex, information causality constraints involved problem by optimizing the UAV relays' locations, nodes' powers, and bandwidth allocations together. An iterative algorithm is proposed by using block coordinate descent (BCD) and successive convex approximation (SCA) methods. Moreover, we also propose a new position initialization method, discuss some special cases of our problem to show some insights for future works, and prove the convergence of our algorithm. Simulations show the effectiveness of our proposed algorithm, some interesting trade-offs about the optimized UAVs' locations, and the influence of available resources.

Highlights

Unmanned aerial vehicles (UAVs) evolve from bomb carriers and reconnaissance equipments, into civilian applications such as children toys, package delivery vehicles, or communication assistants
In the literature, UAVs have been widely applied as base stations [2], [15], [16], [19]–[22], [28], [29], relays [3], [17], [18], [24]–[27], wireless chargers [4], data collectors/disseminators [5], [23], and mobile edge computing (MEC) servers [6]
In [2], a UAV works as a mobile base station to help a ground base station offload data to maximize the minimum user rate by optimizing its flying radius, bandwidth allocation, and user partitioning

Summary

INTRODUCTION

Unmanned aerial vehicles (UAVs) evolve from bomb carriers and reconnaissance equipments, into civilian applications such as children toys, package delivery vehicles, or communication assistants. References in sensor networks usually assume that each node has a fixed communication range, and they do the system design based on this assumption to meet their specified objectives, which means that the system performance based on SNR needs further investigation in these works Their algorithms cannot be directly applied to UAV position optimization problems. When these UAVs are involved in relay networks, we need to establish suitable connections among them to maximize the system throughput These interconnected links make troubles in many respects, such as handling multi-hop links’ information causality, finding general ways to handle these interconnected links involved problems rather than using heuristic methods whose performance or complexity can be improved in general, and balancing loads among UAVs. the better performance can be achieved by jointly optimizing the UAVs’ positions and all available resources, such as nodes scheduling and power allocations, and this has not been well studied yet.

SYSTEM MODEL

POSITION OPTIMIZATION

CONVERGENCE ANALYSIS

VIII. CONCLUSION