With the evolution of vehicle networking technology towards advanced autonomous driving, communication latency has become one of the critical factors affecting road safety and collaborative control stability. In order to solve the problem of time delay runaway in the high real-time scenario, a dual-layer control strategy system is constructed: Multi-access edge computing nodes are deployed at the network layer to enable local traffic offloading and establish security control slices;at the transport layer, a novel dynamic retransmission mechanism is developed by combining delay prediction and compensation algorithm;Finally, empirical analysis of a 100 km/h high-speed platooning scenario demonstrated that compared with the traditional LTE-V2N control group, the experimental group had a 62.82% reduction in average latency from 23.4 ms to 8.7 ms;it compressed delay jitter from ± 9.8 ms to ± 1.2 ms, representing an 87.76% reduction.
在网络层优化策略体系中,引入移动边缘计算策略,并将其作为时延控制的基础架构。具体实践过程中,通过重构5G基站的分布式单元与中央单元部署模式,使具备用户面功能(user plane function,UPF)处理能力的智能路侧单元(road side unit,RSU)能够直接连接车载终端的数据。在该技术的加持下,用户面功能从传统核心网下沉至路侧单元,进而规避核心网网关的迂回路径。基于UPF下沉的移动边缘计算架构如图1所示。