Reinforcement learning optimizes UAV flight and routing for delay-tolerant networks

Researchers have developed a novel approach called JUROR to optimize communication in delay-tolerant networks (DTNs) by jointly controlling unmanned aerial vehicle (UAV) flight and opportunistic routing. This method uses a reinforcement learning framework, specifically Proximal Policy Optimization (PPO), to enlarge future communication opportunities through discrete UAV movements. The system is framed as a partially observable Markov decision process, allowing decentralized execution by individual nodes while a centralized critic uses global statistics during training. Simulations show JUROR outperforms existing methods like PRoPHET and MaxProp in various traffic scenarios. AI

IMPACT This research could lead to more efficient data transmission in challenging network environments, improving the reliability of communication for applications relying on UAVs.

RANK_REASON The item is a research paper published on arXiv detailing a new method for optimizing network communication. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Delay-tolerant networks with network coding: How well can we simulate real devices?
• juror
• Markov decision process
• MaxProp
• Prophet
• Proximal Policy Optimization
• store-carry-forward
• unmanned aerial vehicle

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →