Energy-efficient and solar powered mission planning of UAV swarms to reduce the coverage gap in rural areas: The 3D case

Abstract

Although the percentage of people living outside a broadband network has more than halved in recent years, around 10% of the global population does not have access to the Internet. This lack of coverage is particularly concentrated in rural and low-income areas, in which the lack of a cost-effective electricity supply is the main barrier to expanding network coverage. To tackle this problem, this work proposes a theoretical model based on a self-sustainable 5G network architecture in which the mission planning of a swarm of Unmanned Aerial Vehicles (UAVs) is efficiently scheduled to provide cellular coverage over the territory and to reduce the required energy consumption. A Mixed Integer Linear Programming is provided to formalize the problem of minimizing the energy consumption required by the swarm of UAVs that are able to provide coverage by operating at different altitudes. In order to practically solve the problem, a Genetic Algorithm is defined and evaluated over a realistic scenario. Results indicate that a higher granularity in the number of altitudes at which UAVs can provide coverage increases the percentage of territory that is covered, while a small penalty on the energy consumption must be paid compared to the case in which an unique altitude over the ground is considered.

Introduction

Mobile networks are the most common method to access the Internet for the majority of the world’s population. In fact, mobile technology is the easiest (and in many cases the only) way to allow low-income populations and rural residents to be connected [1]. Although the mobile industry connects over 3.5 billion people around the world (47% of the global population), around 10% of the population is still not covered by mobile broadband, i.e., a 3G connection or higher. This percentage, also known as coverage gap, is mostly concentrated in rural and remote areas (e.g., Sub-Saharan Africa or South Asia), where up to 40% of the population cannot have access to 3G connectivity [2]. If no actions are performed to reduce the coverage gap in the near future, the risk of reinforcing existing inequalities among people living in those areas will still be present.

The main reason associated to the coverage gap is the prohibitive cost of mobile broadband deployment in rural and low-income areas. The cost of construction and management of the required network infrastructure in a rural area is, on average, double compared to the case of an urban deployment. Moreover, the revenue obtained in these locations is normally of the order of $10$ times lower than the one retrieved in a city. In particular, the cost of a mobile network infrastructure, including both operator’s capital (CapEx) and operational expenditure (OpEx), can be divided into three areas: (i) the mobile Base Station (BS) deployment; (ii) the back-haul technology connecting the user to the core network; and (iii) the energy that is required to make these elements functional, both supply and storage.

In the recent years, the networking research community has exploited the emergence of Unmanned Aerial Vehicles (UAVs) to propose a new concept of cellular network, in which UAVs carry a BS to provide cellular coverage over a portion of territory [3], [4]. Solutions such as the Altaeros SuperTower [5] and Loon [6] aim at reducing the operational costs of providing cellular coverage to disperse populations. UAV-based cellular networks provide a better spatial and temporal flexibility compared to traditional-fixed cellular networks, since specific areas and time spans can be selected for coverage provisioning. Nevertheless, a major drawback arises in the main component of the architecture: the UAVs battery depletion derived from their coverage operation [3]. UAVs are generally equipped with limited batteries that must be frequently recharged [7], [8]. In most cases, recharging operations normally require the connection of ground sites to the electricity grid, which would increase the associated cost.

To overcome this challenge, renewable energy solutions, particularly the solar-powered ones, are an interesting common solution for many rural deployments. They can provide energy to ground sites designed to function off-grid or to serve specific remote rural communities [8]. Therefore, exploiting the optimal installation of a set of Solar Panels (SPs) in specific locations can serve as the basis for the proposal of sustainable UAV-based cellular networks able to provide coverage in rural and low-income areas. In [8], the authors propose an energy-efficient UAV mission planning scheme to solve the coverage problem in rural areas throughout time. In particular, the objective of such work is to maximize the energy stored by the set of UAVs and the one stored in the batteries of the ground sites, while ensuring the coverage and energy constraints. However, the impact of the UAVs altitude on the energy consumption is not considered, i.e., the multi-period mission planning is formulated and solved considering a 2D case scenario. As reported in the literature, ascending/descending actions must be carefully considered when proposing an energy-efficient mission planning solution for UAV swarms [9], [10]. These actions clearly impact UAVs battery levels and must be taken into account to prevent from potential battery depletion before the UAV with low battery moves to a ground site in order to recharge it.

In this work we define and solve the 3D problem, in which UAVs are able to fly and provide coverage at different altitudes. The motivation is derived by the fact that if an UAV is hovering at a higher altitude, the coverage range is bigger compared to the case where the UAV altitude is lower. Nevertheless, the higher the altitude of the UAV, the higher energy consumption required to provide coverage. This trade-off must be carefully evaluated and it is a fundamental aspect in energy-efficient UAV-based solutions. In this context, several questions emerge, such as: Is it possible to define a model to cover a set of areas by means of a swarm of UAVs that can be placed at different altitudes and manage their energy consumption in an efficient manner? How to leverage the trade-off between the altitudes at which UAVs should be placed and the energy consumption required for a successful and energy-efficient mission planning? Is it possible to define a strategy to solve this problem in tractable time? These issues are not targeted by [8], while in this paper they are faced and analyzed to a large extent.

In particular, a theoretical model is provided to schedule energy-efficient missions by exploiting UAV-based 5G networks. The proposed model is composed of an energy measurement tool which relies on different specifications for UAV actions, coverage, and path loss. In addition, the formalization of the 3D Energy-Efficient UAVs Mission Planning (3DEE-UMP) problem is provided. Eventually, an heuristic based on Genetic Algorithms (GAs) is defined to solve the problem in a reasonable amount of time. The obtained results reveal that the possibility of considering a higher number of altitudes improves the percentage of the target scenario that is covered by the set of UAVs, while a small penalty on the energy consumption must be paid w.r.t. the case in which only one altitude over the ground is considered.

The remainder of the paper is organized as follows. Section 2 reviews the related work. In Section 3, the proposed UAV-based 5G network architecture is defined. Section 4 presents the problem formulation that aims at solving the problem at hand. The GA-based heuristic is described in Section 5. Section 6 reports the performance evaluation of the proposed solution. Finally, Section 7 concludes our work and introduces future research lines.