UAV-Assisted Heterogeneous Networks for Capacity Enhancement

TL;DR

This paper explores using UAVs as intermediate nodes in heterogeneous networks to enhance capacity, coverage, and load balancing, demonstrating significant spectral efficiency and delay improvements through a neural-based UAV assignment approach.

Contribution

It introduces a neural-based method for UAV assignment in heterogeneous networks, improving coverage and traffic offload compared to traditional ground-based systems.

Findings

Spectral efficiency increased by up to 38%.

Delay reduced by up to 37.5%.

Multiple UAVs improve load balancing and connectivity.

Abstract

Modern day wireless networks have tremendously evolved driven by a sharp increase in user demands, continuously requesting more data and services. This puts significant strain on infrastructure based macro cellular networks due to the inefficiency in handling these traffic demands, cost effectively. A viable solution is the use of unmanned aerial vehicles (UAVs) as intermediate aerial nodes between the macro and small cell tiers for improving coverage and boosting capacity. This letter investigates the problem of user demand based UAV assignment over geographical areas subject to high traffic demands. A neural based cost function approach is formulated in which UAVs are matched to a particular geographical area. It is shown that leveraging multiple UAVs not only provides long range connectivity but also better load balancing and traffic offload. Simulation study demonstrate that the proposed approach yields significant improvements in terms of 5th percentile spectral efficiency up to 38\% and reduced delays up to 37.5\% compared to a ground-based network baseline without UAVs.