Distributed and Autonomous Flow Routing Based on Deep Reinforcement Learning
| Autor: | Sima Barzegar, Marc Ruiz, Luis Velasco |
|---|---|
| Přispěvatelé: | Universitat Politècnica de Catalunya. Departament de Teoria del Senyal i Comunicacions, Universitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors, Universitat Politècnica de Catalunya. GCO - Grup de Comunicacions Òptiques |
| Rok vydání: | 2022 |
| Předmět: |
Deep reinforcement learning
Distributed Control Deep learning Deep learning Enginyeria de la telecomunicació [Àrees temàtiques de la UPC] Distributed-control Autonomous systems Reinforcement learnings Distributed parameter control systems Reward function Reinforcement learning Flow routing Autonomous system Optimal flows Service Quality Aprenentatge profund |
| Zdroj: | 2022 27th OptoElectronics and Communications Conference (OECC) and 2022 International Conference on Photonics in Switching and Computing (PSC). |
| Popis: | A DRL approach with a specific reward function is proposed for autonomous flow routing operating on multilayer scenarios. Illustrative results reveal that the DRL achieves optimal flow routing in terms of cost and service quality. © 2022 IEICE. The research leading to these results has received funding from the H2020 B5G-OPEN (G.A. 101016663), the MINECO-NextGenerationEU TIMING (TSI-063000-2021-145), the MICINN IBON (PID2020-114135RB-I00), and the ICREA Institution. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|