Abstract
In modern Internet of Things (IoT) applications, the freshness of data—quantified by the Age of Information (AoI)—is often more critical than traditional throughput metrics. This paper presents a rigorous mathematical framework based on stochastic geometry to analyze the Peak Age of Information (PAoI) in UAV-assisted IoT networks. By modeling device distributions via a Matérn Cluster Process (MCP) and comparing resource allocation strategies, the study reveals that time-splitting generally outperforms bandwidth-splitting for maintaining data freshness. Furthermore, the research explores how device correlation can be leveraged to minimize PAoI in clustered environments.
Introduction
The proliferation of IoT devices in critical monitoring applications (e.g., industrial safety, environmental sensing) has shifted the focus from high data rates to data "freshness." A delayed status update can render information obsolete. The Age of Information (AoI) measures the time elapsed since the generation of the last successfully received update. This paper specifically targets the Peak AoI (PAoI), representing the maximum age reached just before a new update arrives, providing a "worst-case" metric for system performance.
Deploying Unmanned Aerial Vehicles (UAVs) as mobile base stations offers a flexible solution for collecting data from clustered IoT devices. However, this introduces complex challenges in interference management and resource allocation.
System Model and Methodology
The authors employ a multi-disciplinary approach to model the network:
- Spatial Modeling: IoT devices are distributed according to a Matérn Cluster Process (MCP). Drones are positioned to hover directly above the centers of these clusters to maximize Line-of-Sight (LoS) probability.
- Device Types: The framework considers two scenarios: Uncorrelated devices, where each sensor monitors a unique process, and Correlated devices, where multiple sensors monitor the same physical process.
The analysis incorporates a comprehensive channel model (LoS/NLoS, Nakagami-m fading) and uses the SINR meta-distribution to handle spatial randomness and interference.
Resource Allocation Strategies
A core contribution of the paper is the comparison of two primary load models for serving devices within a cluster:
- Bandwidth Splitting: The available frequency spectrum is divided among active devices.
- Time Splitting: Devices take turns using the full bandwidth for shorter durations.
Key Insights
- Superiority of Time-Splitting: Despite higher instantaneous interference, time-splitting results in a lower mean PAoI. The reduction in transmission time per update outweighs the negative impact of increased interference, especially since UAVs benefit from strong LoS links that dominate over inter-cluster interference.
- Leveraging Correlation: For correlated devices, increasing the number of sensors per cluster can actually improve PAoI at low arrival rates. The UAV can effectively utilize the freshest update available from any device in the group.
- Environmental Impact: In "Highrise Urban" settings, the drop in LoS probability causes PAoI to rise sharply due to frequent transmission failures and retransmissions.
Conclusion
This research provides a robust analytical framework for optimizing data freshness in UAV-IoT networks. The findings suggest that time-splitting is the preferred resource allocation strategy for minimizing PAoI. Future work will extend this model to include UAV mobility, battery constraints, and alternative queuing disciplines like Last-Come-First-Served (LCFS) to further enhance real-time monitoring capabilities.
Summary updated