Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

URLLC-Aware Proactive UAV Placement in Internet of Vehicles

Chen-Feng Liu, Nirmal D. Wickramasinghe, Himal A. Suraweera, Mehdi Bennis, Merouane Debbah

TL;DR

This work targets UAV placement in Internet of Vehicles to maximize video resolution under URLLC constraints on maximal delay. It combines generalized extreme value theory to model the worst-case delay across a time block with Gaussian process regression to map UAV position and video resolution to GEV parameters, enabling offline learning and online decision-making. The authors propose a proactive framework that jointly selects image resolutions and places the UAV by solving a relaxed optimization via generalized projected gradient descent, achieving near-optimal delay performance in even vehicle distributions and substantial reductions in worst-case delay for biased distributions. Numerical results validate the EVT-GPR characterization and demonstrate practical gains in reliability and video quality, highlighting the approach's adaptability to changing vehicle patterns and its potential for real-time IoV deployments.

Abstract

Unmanned aerial vehicles (UAVs) are envisioned to provide diverse services from the air. The service quality may rely on the wireless performance which is affected by the UAV's position. In this paper, we focus on the UAV placement problem in the Internet of Vehicles, where the UAV is deployed to monitor the road traffic and sends the monitored videos to vehicles. The studied problem is formulated as video resolution maximization by optimizing over the UAV's position. Moreover, we take into account the maximal transmission delay and impose a probabilistic constraint. To solve the formulated problem, we first leverage the techniques in extreme value theory (EVT) and Gaussian process regression (GPR) to characterize the influence of the UAV's position on the delay performance. Based on this characterization, we subsequently propose a proactive resolution selection and UAV placement approach, which adaptively places the UAV according to the geographic distribution of vehicles. Numerical results justify the joint usage of EVT and GPR for maximal delay characterization. Through investigating the maximal transmission delay, the proposed approach nearly achieves the optimal performance when vehicles are evenly distributed, and reduces 10% and 19% of the 999-th 1000-quantile over two baselines when vehicles are biased distributed.

URLLC-Aware Proactive UAV Placement in Internet of Vehicles

TL;DR

This work targets UAV placement in Internet of Vehicles to maximize video resolution under URLLC constraints on maximal delay. It combines generalized extreme value theory to model the worst-case delay across a time block with Gaussian process regression to map UAV position and video resolution to GEV parameters, enabling offline learning and online decision-making. The authors propose a proactive framework that jointly selects image resolutions and places the UAV by solving a relaxed optimization via generalized projected gradient descent, achieving near-optimal delay performance in even vehicle distributions and substantial reductions in worst-case delay for biased distributions. Numerical results validate the EVT-GPR characterization and demonstrate practical gains in reliability and video quality, highlighting the approach's adaptability to changing vehicle patterns and its potential for real-time IoV deployments.

Abstract

Unmanned aerial vehicles (UAVs) are envisioned to provide diverse services from the air. The service quality may rely on the wireless performance which is affected by the UAV's position. In this paper, we focus on the UAV placement problem in the Internet of Vehicles, where the UAV is deployed to monitor the road traffic and sends the monitored videos to vehicles. The studied problem is formulated as video resolution maximization by optimizing over the UAV's position. Moreover, we take into account the maximal transmission delay and impose a probabilistic constraint. To solve the formulated problem, we first leverage the techniques in extreme value theory (EVT) and Gaussian process regression (GPR) to characterize the influence of the UAV's position on the delay performance. Based on this characterization, we subsequently propose a proactive resolution selection and UAV placement approach, which adaptively places the UAV according to the geographic distribution of vehicles. Numerical results justify the joint usage of EVT and GPR for maximal delay characterization. Through investigating the maximal transmission delay, the proposed approach nearly achieves the optimal performance when vehicles are evenly distributed, and reduces 10% and 19% of the 999-th 1000-quantile over two baselines when vehicles are biased distributed.
Paper Structure (13 sections, 1 theorem, 5 equations, 6 figures, 1 table, 1 algorithm)

This paper contains 13 sections, 1 theorem, 5 equations, 6 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Our Contribution
  4. System Model and Problem Formulation
  5. System Architecture
  6. Problem Formulation
  7. Maximal Delay Characterization
  8. Generalized Extreme Value Distribution
  9. GEV Distribution Parameters Characterization
  10. GPR for Position-Parameter Mapping
  11. URLLC-Aware Resolution Selection and UAV Placement
  12. Numerical Results
  13. Conclusion

Key Result

Theorem 1

Let $T(1),T(2),\cdots$ be a stationary process with the identical marginal distribution and define $T_{\max}=\max\{T(1),\cdots,T(k)\}$. As $k\to\infty$, $T_{\max}$ can be approximately characterized by a GEV distribution $G(t;\mu,\sigma,\xi)$, i.e., $\Pr\{T_{\max}> t\}\approx G(t;\mu,\sigma,\xi)\equ

Figures (6)

  • Figure 1: Studied IoV network with a UAV in which the longitudinal and latitudinal roads are divided into six zones.
  • Figure 2: Learned GEV distribution parameters $\hat{\mu}$, $\hat{\sigma}$, and $\hat{\xi}$ (in Zone 3) versus $\mathbf{e}_{\rm u}$ given that the estimation errors are not significant.
  • Figure 3: Empirical and approximate CCDFs of the maximal delay $\tilde{T}^{(\max)}$ in Zone 1 at $\bar{\mathbf{e}}_{\rm u}=(90\,\hbox{m},60\,\hbox{m})$.
  • Figure 4: Numerical and predicted CCDFs of the maximal delay $\tilde{T}^{(\max)}$ in Zone 6 at $\mathbf{e}_{\rm u}=(45\,\hbox{m}, 45\,\hbox{m})$.
  • Figure 5: CCDFs of the maximal transmission delay in different UAV placement schemes.
  • ...and 1 more figures

Theorems & Definitions (1)

  • Theorem 1