Federated Learning With a Drone Orchestrator: Path Planning for Minimized Staleness
In this paper, we investigate the problem of scheduling transmissions for spatially scattered nodes that contribute to a collaborative federated learning (FL) algorithm via wireless links provided by a drone. In the considered system, the drone acts as an orchestrator, coordinating the transmissions...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Open Journal of the Communications Society |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9400376/ |
| _version_ | 1818573543074430976 |
|---|---|
| author | Igor Donevski Nithin Babu Jimmy Jessen Nielsen Petar Popovski Walid Saad |
| author_facet | Igor Donevski Nithin Babu Jimmy Jessen Nielsen Petar Popovski Walid Saad |
| author_sort | Igor Donevski |
| collection | DOAJ |
| description | In this paper, we investigate the problem of scheduling transmissions for spatially scattered nodes that contribute to a collaborative federated learning (FL) algorithm via wireless links provided by a drone. In the considered system, the drone acts as an orchestrator, coordinating the transmissions and the learning schedule within a predefined deadline. The actual schedule is reflected in a planned path: as the drone traverses it, it controls the distance and thereby the data rate to each node. Hence, the model is structured such that the drone orchestrator uses the path (trajectory) as its only tool to achieve fairness in terms of learning <italic>staleness</italic>, which reflects the learning time discrepancy among the nodes. Using the number of learning epochs performed at each learner as a performance indicator, we combine the average number of epochs computed and staleness into a balanced optimization criterion that is agnostic to the underlying FL implementation. We consider two methods for solving the complex trajectory planning optimization problem for static nodes: (1) successive convex programming (SCP) and (2) deep reinforcement learning (RL). Considering the proposed criterion, both methods are compared in three specific scenarios with few nodes. The results show that drone-orchestrated FL outperforms an immobile deployment by providing improvements in the range of 57% to 87.7%. Additionally, RL-guided trajectories are generally superior to SCP provided ones for complex node arrangements. |
| first_indexed | 2024-12-15T00:12:44Z |
| format | Article |
| id | doaj.art-bb6371a5a84b4e9fba679862510bb1c0 |
| institution | Directory Open Access Journal |
| issn | 2644-125X |
| language | English |
| last_indexed | 2024-12-15T00:12:44Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Open Journal of the Communications Society |
| spelling | doaj.art-bb6371a5a84b4e9fba679862510bb1c02022-12-21T22:42:31ZengIEEEIEEE Open Journal of the Communications Society2644-125X2021-01-0121000101410.1109/OJCOMS.2021.30720039400376Federated Learning With a Drone Orchestrator: Path Planning for Minimized StalenessIgor Donevski0https://orcid.org/0000-0002-3935-6376Nithin Babu1https://orcid.org/0000-0002-9757-4726Jimmy Jessen Nielsen2https://orcid.org/0000-0001-6664-7198Petar Popovski3https://orcid.org/0000-0001-6195-4797Walid Saad4https://orcid.org/0000-0003-2247-2458Department of Electronic Systems, Aalborg University, Aalborg, DenmarkDepartment of Electronic Systems, Aalborg University, Aalborg, DenmarkDepartment of Electronic Systems, Aalborg University, Aalborg, DenmarkDepartment of Electronic Systems, Aalborg University, Aalborg, DenmarkDepartment of Electrical and Computer Engineering, Wireless@VT, Virginia Tech, Blacksburg, VA, USAIn this paper, we investigate the problem of scheduling transmissions for spatially scattered nodes that contribute to a collaborative federated learning (FL) algorithm via wireless links provided by a drone. In the considered system, the drone acts as an orchestrator, coordinating the transmissions and the learning schedule within a predefined deadline. The actual schedule is reflected in a planned path: as the drone traverses it, it controls the distance and thereby the data rate to each node. Hence, the model is structured such that the drone orchestrator uses the path (trajectory) as its only tool to achieve fairness in terms of learning <italic>staleness</italic>, which reflects the learning time discrepancy among the nodes. Using the number of learning epochs performed at each learner as a performance indicator, we combine the average number of epochs computed and staleness into a balanced optimization criterion that is agnostic to the underlying FL implementation. We consider two methods for solving the complex trajectory planning optimization problem for static nodes: (1) successive convex programming (SCP) and (2) deep reinforcement learning (RL). Considering the proposed criterion, both methods are compared in three specific scenarios with few nodes. The results show that drone-orchestrated FL outperforms an immobile deployment by providing improvements in the range of 57% to 87.7%. Additionally, RL-guided trajectories are generally superior to SCP provided ones for complex node arrangements.https://ieeexplore.ieee.org/document/9400376/Drone trajectory optimizationwireless communicationsfederated learningdrone small cellsstaleness minimizationreinforcement learning |
| spellingShingle | Igor Donevski Nithin Babu Jimmy Jessen Nielsen Petar Popovski Walid Saad Federated Learning With a Drone Orchestrator: Path Planning for Minimized Staleness IEEE Open Journal of the Communications Society Drone trajectory optimization wireless communications federated learning drone small cells staleness minimization reinforcement learning |
| title | Federated Learning With a Drone Orchestrator: Path Planning for Minimized Staleness |
| title_full | Federated Learning With a Drone Orchestrator: Path Planning for Minimized Staleness |
| title_fullStr | Federated Learning With a Drone Orchestrator: Path Planning for Minimized Staleness |
| title_full_unstemmed | Federated Learning With a Drone Orchestrator: Path Planning for Minimized Staleness |
| title_short | Federated Learning With a Drone Orchestrator: Path Planning for Minimized Staleness |
| title_sort | federated learning with a drone orchestrator path planning for minimized staleness |
| topic | Drone trajectory optimization wireless communications federated learning drone small cells staleness minimization reinforcement learning |
| url | https://ieeexplore.ieee.org/document/9400376/ |
| work_keys_str_mv | AT igordonevski federatedlearningwithadroneorchestratorpathplanningforminimizedstaleness AT nithinbabu federatedlearningwithadroneorchestratorpathplanningforminimizedstaleness AT jimmyjessennielsen federatedlearningwithadroneorchestratorpathplanningforminimizedstaleness AT petarpopovski federatedlearningwithadroneorchestratorpathplanningforminimizedstaleness AT walidsaad federatedlearningwithadroneorchestratorpathplanningforminimizedstaleness |