Remote ID based UAV collision avoidance optimization for low-altitude airspace safety

Ziye Jia, Yian Zhu, Qihui Wu, Yao Wu, Lei Zhang, Sen Yang, Zhu HanChinese 2025 · DOI

TL;DR

This paper introduces a novel framework for Unmanned Aerial Vehicle (UAV) collision avoidance that leverages Remote ID (Remote Identification) technology to enable inter-UAV communication and optimize transmission delays. By employing a multi-agent deep reinforcement learning approach, UAVs can dynamically adapt their communication protocols, leading to a significant reduction in delays and improved safety in low-altitude airspace.

What Problem Does This Paper Address?

The increasing prevalence of Unmanned Aerial Vehicles (UAVs) in various applications, from package delivery to urban monitoring, brings with it a critical challenge: ensuring safe operations, especially in crowded low-altitude airspaces. Traditional sensor-based collision avoidance methods are often costly and impractical for large-scale deployments. While communication-based approaches like Remote ID offer a scalable solution for monitoring and information exchange, they are hampered by communication delays. These delays, influenced by factors like protocol choice, interference, and packet loss, can significantly degrade the effectiveness of collision avoidance systems, which rely on timely information for accurate decision-making. This paper aims to bridge this gap by developing a system that not only uses Remote ID for situational awareness but also actively optimizes communication delays to enhance collision avoidance performance.

Methods

The core of this work is the Distributed Multi-UAV Collision Avoidance (DMUCA) framework. This framework operates in three phases:

  1. Observe: UAVs broadcast their identity, position, and velocity using Remote ID, utilizing protocols like Bluetooth Low Energy (BLE) 4, BLE 5, or Wi-Fi.
  2. Orient: UAVs predict the trajectories of nearby UAVs to mitigate the impact of information propagation delays.
  3. Decide: Collision-free velocities are calculated using the Optimal Reciprocal Collision Avoidance (ORCA) method, followed by a path recovery mechanism.

To address the challenge of communication delays, the authors propose a Multi-Agent Deep Q-Network (MADQN)-based Adaptive Transmission Mode Configuration (MADQN-ATMC) algorithm. This algorithm allows each UAV to act as an independent agent that learns to autonomously select the optimal communication protocol (BLE 4, BLE 5, or Wi-Fi) and adjust its message transmission rate. This is achieved by transforming the problem into a Markov Decision Process (MDP) and training the agents using a value-based reinforcement learning approach. The reward function is designed to balance local communication efficiency with global transmission delay minimization.

Datasets

Not discussed in the paper. The simulations utilize a 1km x 1km airspace with five UAVs following pre-planned trajectories generated using MATLAB's waypoint trajectory toolbox.

Key Insights and Ideas

  • Remote ID as a Foundation for Collision Avoidance: The paper demonstrates that Remote ID, beyond its regulatory purpose, can be effectively utilized as a communication backbone for distributed multi-UAV collision avoidance.
  • The Criticality of Communication Delay: The research strongly emphasizes that communication delay is not a static parameter but a dynamic factor that directly impacts collision avoidance accuracy and safety. Longer delays lead to reduced predictive accuracy and a higher risk of near-misses.
  • Adaptive Protocol Selection is Key: Instead of relying on a single, fixed communication protocol, dynamically selecting the most suitable protocol (BLE 4, BLE 5, or Wi-Fi) based on environmental conditions (e.g., airspace density) is crucial for minimizing delays.
  • MADQN for Decentralized Optimization: The MADQN-ATMC algorithm proves effective in enabling UAVs to learn optimal communication strategies in a decentralized manner, adapting to dynamic environments without requiring centralized control.
  • Balancing Local and Global Objectives: The reward function's design, which balances local communication efficiency with global transmission delay, is essential for achieving robust and globally optimal performance.

Key Results

  • The proposed DMUCA framework, combined with the MADQN-ATMC algorithm, successfully avoids collisions while maintaining safe separation distances, even under significant communication delays.
  • The adaptive protocol selection mechanism achieved an average delay reduction of 32% compared to the optimal fixed protocol configuration.
  • In high-density scenarios, MADQN-ATMC achieved an average delay of 1,855.28ms, closely matching the optimal fixed Wi-Fi configuration.
  • In low-density scenarios, MADQN-ATMC achieved an average delay of 246.82ms, near the optimal delay for fixed BLE 4.
  • In dynamic-density environments, MADQN-ATMC achieved an average delay of 1050.46ms, outperforming fixed protocols and other baseline methods.
  • The MADQN-ATMC algorithm demonstrated superior convergence speed and consistency compared to baseline methods like MADDPG and IAC.

Limitations and Future Work

  • Limitations: The paper acknowledges that the current work focuses on a simplified scenario with a limited number of UAVs and assumes constant velocity for trajectory prediction. The impact of more complex environmental factors like dynamic obstacles and system uncertainties is not fully explored.
  • Future Work: The authors plan to extend the framework to more complex scenarios, including non-cooperative UAVs, dynamic obstacles, and system uncertainties. They also aim to investigate the influence of hardware differences on Remote ID platform performance through real-world deployments.

Summary updated

Abstract

With the rapid development of Unmanned Aerial Vehicles (UAVs), it is paramount to ensure safe and efficient operations in open airspaces. The Remote Identification (Remote ID) is deemed an effective real-time UAV monitoring system by the federal aviation administration, which holds potentials for enabling inter-UAV communications. This paper deeply investigates the application of Remote ID for UAV collision avoidance while minimizing communication delays. First, we propose a Remote ID based Distributed Multi-UAV Collision Avoidance (DMUCA) framework to support the collision detection, avoidance decision-making, and trajectory recovery. Then, the average transmission delays for Remote ID messages are analyzed, incorporating the packet reception mechanisms and packet loss due to interference. The optimization problem is formulated to minimize the long-term average communication delay, where UAVs can flexibly select the Remote ID protocol to enhance the collision avoidance performance. To tackle the problem, we design a multi-agent deep Q-network based adaptive communication configuration algorithm, allowing UAVs to autonomously learn the optimal protocol configurations in dynamic environments. Finally, numerical results verify the feasibility of the proposed DMUCA framework, and the proposed mechanism can reduce the average delay by 32% compared to the fixed protocol configuration.
BibTeX
@inproceedings{jia2025remote,
  author = {Ziye Jia and Yian ZHU and Qihui Wu and Yao Wu and Lei Zhang and Sen Yang and Zhu Han},
  title = {{Remote ID based UAV collision avoidance optimization for low-altitude airspace safety}},
  booktitle = {Chinese},
  year = {2025},
  doi = {10.1016/j.cja.2025.103841},
}
Figure