Abstract
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 xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">staleness</i> , 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.
Highlights
I NITIALLY meant for military uses, followed by a boom in commercial entertainment usage, the flying drones or unmanned aerial vehicles (UAVs) have a growing importance in the world of communications
Due to their flexibility, low altitude platform (LAP) drones are useful to act as on-demand drone small cells (DSCs) for wireless communication support
Staleness is a cardinal metric for our setup since all nodes are assumed to possess useful data and, stragglers cannot be dropped. This creates asynchrony between the amount of learning each robot does. We model this asynchrony by the largest difference of epochs computed among the learners, which has been shown to be key for the performance of the generation of asynchronous federated learning (FL) [30]–[32] The maximum staleness comes as a consequence of the asynchrony of such an FL implementation which is an issue that we want to tackle by implementing path planning in the duration of a single FL round
Summary
I NITIALLY meant for military uses, followed by a boom in commercial entertainment usage, the flying drones or unmanned aerial vehicles (UAVs) have a growing importance in the world of communications. The use of drones for wireless communication purposes has received a surge of attention [1], [2] due to their excellent coverage to outdoor users. Adhering to the surge of interest in enabling connected intelligence [3], [4], one can envision the use of such aerial platforms as an effective means to implement collaborative federated learning (FL), where few geographically scattered nodes collaborate to improve a common machine learning (ML) model. Such an FL use case requires periodic and high bandwidth communications
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
