Multipath intelligent transmission strategy for UAM dynamic networks

As an important vehicle for the low-altitude economy, the flying cars play a crucial role in vigorously developing urban air mobility (UAM) and building an efficient low-altitude intelligent network. However, the high mobility of flying cars introduces instability in transmission links due to high-dynamic network changes. To address these challenges, a multipath intelligent transmission strategy was proposed to achieve efficient data transmission. Firstly, a multipath intelligent routing algorithm based on reinforcement learning was introduced, which was not only proven to effectively enhance routing efficiency and reduce routing latency but also capable of dynamically segmenting data packets according to different transmission requirements to adapt to changes in the physical environment. To further improve the adaptability of the routing algorithm to dynamic environments, an environmental validation mechanism was designed to evaluate the compatibility between current routing strategies and dynamic network, enabling self-adaptive adjustment of routing strategies during the transmission process. Simulation experiments demonstrate that, under various scenarios, the multipath intelligent routing strategy based on reinforcement learning (MIRSRL) effectively reduces end-to-end transmission delay while also increasing the probability of successful data recovery.