Fedqos cuts communication and improves safety in car ai training
QoS-Aware Federated Learning for Multimodal In-Cabin Interaction in Smart Vehicles
Machine Learning
Summary
Smart cars use many sensors to understand what’s happening inside, but sending and combining all this data safely can be tricky because the car’s network changes and resources like energy are limited. The authors created FedQoS, a way to train AI models asynchronously so that learning only happens when the car’s sensors and energy levels are safe. This method also decides when to send updates based on how useful and costly they are to transmit. Their tests show FedQoS keeps model accuracy close to traditional methods while reducing network problems, communication, and delay, making it better suited for real car environments.
What this means in practice
- •For automotive engineers: Manage in-car AI training efficiently by adapting computation and communication to vehicle resource constraints to improve safety and reduce communication load.
- •For network schedulers: Optimize data transmission from vehicles by selectively sending updates that balance importance and network conditions to reduce latency and overhead.
Authors
Baran Can Gül, Mert Nakıp, Nasser Jazdi, Michael Weyrich
Abstract
Modern smart vehicles leverage multimodal sensors, ranging from high-bandwidth vision systems to low-rate physiological monitors, to provide personalized in-cabin services. However, integrating high-fidelity multimodal fusion with collaborative training is often hindered by the heterogeneous and time-varying Quality of Service (QoS) constraints of vehicular networks. Standard Federated Learning (FL) approaches enforce rigid synchronous rounds that fail to account for these resource asymmetries, leading to safety-critical timing violations and energy exhaustion. In this paper, we propose FedQoS, a novel asynchronous, event-triggered FL framework that decouples local computation from global communication via a two-phase gating mechanism. First, we introduce a resource-aware training gate that initializes local learning only when sensing buffers and energy reserves meet safety thresholds, preventing ML tasks from compromising core vehicle mobility. Second, a QoS-aware transmission policy gates uplink updates based on an efficiency score that balances model novelty against instantaneous latency and energy costs. Locally, clients optimize an objective featuring a staleness-aware proximal term that dynamically adjusts the global anchor strength based on update age. Extensive experiments on multimodal vehicular datasets demonstrate that FedQoS achieves competitive personalized accuracy with only marginal performance loss compared to FedAvg, while substantially reducing QoS violations, cutting communication overhead by 76.7\%, and lowering latency cost by 26.0\%, demonstrating a highly favorable accuracy and efficiency balance for real-world vehicular deployments.