RIS-Empowered Topology Control for Decentralized Federated Learning in Urban Air Mobility

Urban air mobility (UAM) expands vehicles from the ground to the near-ground space, envisioned as a revolution for transportation systems. Comprehensive scene perception is the foundation for autonomous aerial driving. However, UAM encounters the intelligent perception challenge: high-perception learning requirements conflict with the limited sensors and computing chips of flying cars. To overcome the challenge, federated learning (FL) and other collaborative learning have been proposed. It enables resource-limited devices to conduct onboard deep learning (DL) collaboratively. But traditional FL relies on a central integrator for DL model aggregation, which is difficult to deploy in dynamic UAM environments. The fully decentralized learning schemes may be the intuitive solution while the convergence of decentralized learning cannot be guaranteed. Accordingly, this article explores reconfigurable intelligent surfaces (RISs)-empowered decentralized FL (DFL), taking account of topological attributes to facilitate the DFL performance with convergence guarantee. Several DFL topological criteria are proposed for optimizing the transmission delay and convergence rate. Subsequently, we innovatively leverage the RIS link construction and deconstruction ability to remold the current network based on the proposed topological criteria. This article rethinks the functions of RIS from the perspective of the network layer. Furthermore, a deep deterministic policy gradient-based RIS phase shift control algorithm is developed to reshape the communication network. Simulation experiments are conducted over MobileNet-based multiview learning to verify the efficiency of the DFL framework.