The Fog Computing was proposed to extend the computing task to the network edge in lots of Internet of Things (IoT) scenario, such as Internet of Vehicle (IoV). However, the unbalanced data processing requirement caused by the uneven distribution of vehicles in time and space limits the service capability of IoV. To enhance the flexibility and data processing capability, we propose a hybrid fog architecture which composed by fog computing radio access network (F-RAN) and Vehicular Fog Computing (VFC), which is called VF-based F-RAN. In addition, we propose a heuristic algorithm enhanced by deep learning to optimize the computation offloading in this hybrid architecture. The simulation result reveals that the proposed hybrid fog architecture with the heuristic algorithm can effectively improve the data processing efficiency and balance the Quality of Experience (QoE).

中文翻译：

---

基于F-RAN的车雾处理能力和QoE驱动的优化计算分流方案

提出雾计算是为了将计算任务扩展到许多物联网（IoT）场景中的网络边缘，例如车联网（IoV）。然而，由于车辆在时间和空间上的不均匀分布所导致的不平衡的数据处理要求限制了IoV的服务能力。为了提高灵活性和数据处理能力，我们提出了一种由雾计算无线接入网（F-RAN）和车载雾计算（VFC）组成的混合雾架构，称为基于VF的F-RAN。此外，我们提出了一种通过深度学习增强的启发式算法，以优化此混合体系结构中的计算分流。