In our research, we emphasize how vehicle characteristics critically shape our dynamic scheduling scheme for V2V communications-enhanced federated learning (FL). The influence of these characteristics is twofold:

High-Speed Channel Variability Due to Vehicle Movement: Unlike many FL studies that assume a static or slowly changing channel state in traditional wireless networks, the movement of vehicles introduces rapid changes in channel conditions. This dynamic environment necessitates a scheduling algorithm capable of adapting to these changes. Our proposed dynamic scheduling algorithm is designed to dynamically schedule the upload of models by vehicles at each time slot, ensuring that model transmissions remain efficient and resilient to fluctuating channel states throughout a training round.

青衫晚上好~

happy middle-autumn festival！~

Morning ~~~

晚上好~！

早安！~