Energy-Efficient Techniques for UAVs in Communication-based Applications

TL;DR

This paper reviews energy-efficient techniques for UAVs in communication applications, focusing on challenges like battery limitations and exploring recent solutions including GSP and ML for future improvements.

Contribution

It provides a comprehensive overview of current energy-saving methods for UAV communication systems and discusses future research directions with advanced signal processing and machine learning.

Findings

Analysis of energy constraints in UAV communication applications

Comparison of recent energy-efficient solutions

Identification of promising future research directions

Abstract

Unmanned Aerial Vehicles (UAVs), which are at the forefront of cutting-edge technology, have unmatched potential for pioneering applications in a wide range of disciplines. In particular, in the field of cognitive radio (CR), which is a key aspect in the implementation of the new 5G telecommunication technology. The integration between the drone and CR consolidates the drone's capabilities at the heart of the remarkably promising Internet-of-Things (IoT) technology supported by CR. The highly dynamic network topologies, weakly networked communication links, reliable line-of-sight (LOS) communication links, and orbital or flight paths are characteristic features of UAV communication compared to terrestrial wireless networks. Nevertheless, the implementation of this system is constrained by several severe challenges, such as energy efficiency, battery power limitation, spectrum handover, propagation channel modeling, routing protocols, security policy, and delay setbacks. In this paper, we consider the impact of energy scarcity faced by the UAV in various CR applications. We also analyze the impact of energy scarcity on communication-based applications and present the general problem of battery limitation. Finally, we give an overview and comparison between recent solutions proposed by researchers both in the field of communication and based on batteries and consider possible future directions according to the state of the art, such as novel Graph Signal Processing (GSP) and machine learning (ML).