Multi-objective Optimization for Data Collection in UAV-assisted Agricultural IoT

Lingling Liu,Aimin Wang,Geng Sun,Jiahui Li,Hongyang Pan,Tony Q. S. Quek

2024-03-04

Abstract:The ground fixed base stations (BSs) are often deployed inflexibly, and have high overheads, as well as are susceptible to the damage from natural disasters, making it impractical for them to continuously collect data from sensor devices. To improve the network coverage and performance of wireless communication, unmanned aerial vehicles (UAVs) have been introduced in diverse wireless networks, therefore in this work we consider employing a UAV as an aerial BS to acquire data of agricultural Internet of Things (IoT) devices. To this end, we first formulate a UAV-assisted data collection multi-objective optimization problem (UDCMOP) to efficiently collect the data from agricultural sensing devices. Specifically, we aim to collaboratively optimize the hovering positions of UAV, visit sequence of UAV, speed of UAV, in addition to the transmit power of devices, to simultaneously achieve the maximization of minimum transmit rate of devices, the minimization of total energy consumption of devices, and the minimization of total energy consumption of UAV. Second, the proposed UDCMOP is a non-convex mixed integer nonlinear optimization problem, which indicates that it includes continuous and discrete solutions, making it intractable to be solved. Therefore, we solve it by proposing an improved multi-objective artificial hummingbird algorithm (IMOAHA) with several specific improvement factors, that are the hybrid initialization operator, Cauchy mutation foraging operator, in addition to the discrete mutation operator. Finally, simulations are carried out to testify that the proposed IMOAHA can effectively improve the system performance comparing to other benchmarks.

Information Theory,Signal Processing

What problem does this paper attempt to address?

This paper attempts to address multiple challenges faced when using Unmanned Aerial Vehicles (UAVs) to assist data collection in agricultural Internet of Things (IoT). Specifically, the paper focuses on the following issues:

1. **Uneven device transmission rates**: When a UAV hovers at a location, devices closer to the UAV can upload data at higher transmission rates, while devices farther away have lower transmission rates. This leads to an imbalance in transmission rates among different devices.

2. **Uneven device energy consumption**: The varying distances between different devices and the UAV result in inconsistent energy consumption for data transmission. Devices farther away need to consume more energy to transmit data, affecting the overall energy efficiency of the system.

3. **UAV energy consumption**: UAVs consume energy both during movement and hovering. The energy consumption during movement depends on the UAV's speed, while the energy consumption during hovering is related to the hovering time. These energy consumptions impact the communication performance and system lifespan of the UAV.

To address the above challenges, the paper proposes a multi-objective optimization problem (UDCMOP), aiming to collaboratively optimize the UAV's hovering positions, visiting sequence, speed, and the transmission power of devices, while achieving the following goals:
- Maximizing the minimum transmission rate of all devices.
- Minimizing the total energy consumption of all devices.
- Minimizing the total energy consumption of the UAV.

By solving this multi-objective optimization problem, the paper hopes to improve the efficiency and energy efficiency of data collection in agricultural IoT.