Abstract

Rapid urbanization has intensified air pollution levels, creating an urgent need for intelligent, real-time monitoring systems that can effectively track and analyze pollutants across dynamic city environments. Conventional Internet of Things (IoT)-based sensing frameworks often face challenges such as latency, limited processing power, and inefficient data management when deployed at large scales. Recent advances in quantum communication and edge artificial intelligence (Edge AI) have opened new avenues for developing highly adaptive and secure environmental monitoring architectures. Despite the proliferation of drone-assisted monitoring systems, most existing models rely on centralized cloud computing, resulting in network bottlenecks and delayed responses. Furthermore, data collected from heterogeneous sensors often lack accuracy due to noise interference and spatial inconsistencies, limiting the reliability of real-time pollution detection and source localization. This study proposes an Edge- Enabled Quantum Drone Sensor Network (Q-DSN) integrated with Convolutional Neural Networks (CNNs) to perform decentralized pollution detection and classification. Quantum key distribution (QKD) enhances communication security among drones, while CNN-based feature extraction processes multispectral data from gas sensors and high-resolution cameras. The edge layer employs a lightweight AI model for on-site prediction, reducing latency and dependence on cloud computation. In addition, an adaptive routing protocol optimizes data transmission between drones and ground stations. Simulation and field-level evaluations demonstrated that the proposed Q-DSN achieved a detection accuracy of 81–84%, a latency reduction to 2.1–2.7 seconds, and an energy efficiency improvement of 20–35% compared to conventional IoT-based and drone-assisted monitoring systems. Communication reliability reached 92% through quantum-secured channels, while edge-level inference enabled real-time classification of pollutants such as CO2, NO2, PM2.5, and VOCs across a 2 km² urban testbed.

Authors

N. Devakirubai
R P Sarathy Institute of Technology, India

Keywords

Quantum Drone Networks, Edge AI, Convolutional Neural Networks, Urban Pollution Detection, Secure Environmental Monitoring