By N Viswesh & B Gopi | Published on January 15, 2025
Unmanned Aerial Vehicles (UAVs) or drones are increasingly utilized in mission-critical areas like surveillance and disaster relief, necessitating advanced distributed coordination algorithms to ensure consensus even amidst challenges such as hardware malfunctions and communication delays. Traditional consensus mechanisms struggle to adapt to the dynamic environments of drone networks, characterized by intermittent connectivity, high mobility, and energy constraints. This calls for the development of specialized algorithms that maintain resilience, efficiency, and energy efficiency amidst rapid topology changes and unreliable communication. While protocols like Practical Byzantine Fault Tolerance (PBFT) offer high reliability under attacks, they suffer from substantial communication overhead, and Delegated Byzantine Fault Tolerance (DBFT), while reducing communication costs, may compromise fault tolerance in high-risk scenarios.
The proposed D2BFT model is a simulation framework for studying fault-tolerant consensus in distributed drone networks under the Multi-Agent Reinforcement Learning Proximal Policy Optimization (MARLPPO) mechanism that presents a scalable and efficient solution to real-world applications of drone-based surveil- lance, revealing better performance measurements in experimental evidence. D2BFT unites the essential elements of Practical Byzantine Fault Tolerance (PBFT) and Delegated Byzantine Fault Tolerance (DBFT) to decisively achieve consen- sus among drones, effectively managing varying fault percentages and accom- modating an increasing number of drones. The hybrid system provides a strong agreement mechanism that guarantees reliability and efficiency, ensuring that drones can communicate and function cohesively despite failures or malicious behavior within the network. Generally, D2BFT offers useful information on the efficiency and resilience of consensus mechanisms in autonomous systems with fault-prone environments.
12
5