Privacy-Preserving Drone Navigation Through Homomorphic Encryption for Collision Avoidance

2025-07-19

AI Summary

To address commercial privacy leakage—specifically exposure of flight routes and delivery destinations—in shared-path drone delivery within residential areas, this paper proposes a privacy-preserving cooperative collision-avoidance scheme based on homomorphic encryption (HE). The scheme is the first to apply HE to trajectory intersection detection for drone obstacle avoidance, enabling direct ciphertext-domain path comparison without decryption, thereby ensuring both privacy preservation and real-time responsiveness. Algorithmic optimizations and lightweight virtual-machine-based simulation significantly reduce cryptographic computation overhead and communication load. Experimental evaluation demonstrates that, compared to a garbled-circuit-based baseline, the proposed scheme achieves approximately 3.2× higher computational throughput and reduces communication volume by 67%. Formal security analysis confirms its resilience against semi-honest adversaries.

Technology Category

• Machine Learning: Privacy
• Search and Optimization: Heuristic Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Security and Privacy: Data transparency and provenance
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

As drones increasingly deliver packages in neighborhoods, concerns about collisions arise. One solution is to share flight paths within a specific zip code, but this compromises business privacy by revealing delivery routes. For example, it could disclose which stores send packages to certain addresses. To avoid exposing path information, we propose using homomorphic encryption-based comparison to compute path intersections. This allows drones to identify potential collisions without revealing path and destination details, allowing them to adjust altitude to avoid crashes. We implemented and tested our approach on resource-limited virtual machines to mimic the computational power of drones. Our results demonstrate that our method is significantly faster and requires less network communication compared to a garbled circuit-based approach. We also provide a security analysis of the approach against potential attacks.

Problem

• Prevent drone collisions without revealing flight paths
• Protect business privacy in shared delivery routes
• Enable secure path intersection computation using encryption

Innovation

• Homomorphic encryption for path intersection computation
• Resource-limited virtual machine implementation
• Faster and less network communication required