Things are interconnected using information and communication technologies in smart cities, forming Internet of Things (IoT). The Internet of Vehicles (IoV) refers to an IoT application, where the urban vehicle fleet forms a worldwide network, using V2X (Vehicle-to-Everything) communications. The 5G is the new generation of cellular networks that will eliminate the bounds of bandwidth, performance, and latency limitations. IoV is one of the high-priority application domains for 5G. Among the under development IEEE Standard regarding 5G, the IEEE P1931.1 standard (named also Real-time Onsite Operations Facilitation (ROOF) Standard) seems to be very promising for IoV requirements. This paper proposes ROOF-based Named Data Vehicular Networking (RND − V_n), a named data networking (NDN) architecture for IoV. In addition to the proposal, we provide SeCrND_n (Searchable Encryption for Content Retrieval in NDN), a searchable encryption technique for NDN content retrieval. Furthermore, we propose the intelligent Named Data Caching (iNDC), a machine-learning–based data caching technique for ROOF-based named data networking. The iNDC predicts the number of content requests, such that popular contents are kept as long as possible on roadside units. The proposed iNDC is also used to predict the storage capacity required by each roadside unit. A performance study was conducted to evaluate the performance of machine learning algorithms applied to iNDC. The results show that linear and ridge regressions are the most efficient in terms of content popularity prediction. To predict the capacity of new roadside units, iNDC provides better accuracy using k-Nearest Neighbors.

在智慧城市中使用信息和通信技术将事物互连，形成物联网（IoT）。车联网 (IoV) 是指一种物联网应用，其中城市车队形成一个全球网络，使用 V2X（车辆到一切）通信。5G 是新一代蜂窝网络，它将消除带宽、性能和延迟限制的界限。车联网是 5G 的高优先级应用领域之一。在正在制定的有关 5G 的 IEEE 标准中，IEEE P1931.1 标准（也称为实时现场操作便利 (ROOF) 标准）似乎对 IoV 要求非常有希望。本文提出了基于 ROOF 的命名数据车辆网络 ( R N D − V _n)，一种用于 IoV 的命名数据网络 (NDN) 架构。除了提案之外，我们还提供了S e C r N D _n（NDN 内容检索的可搜索加密），这是一种用于 NDN 内容检索的可搜索加密技术。此外，我们提出了智能命名数据缓存 (iNDC)，这是一种基于机器学习的数据缓存技术，用于基于 ROOF 的命名数据网络。该INDC预测内容请求，使得流行的内容尽可能长时间地保持在路边单元的数量。提议的iNDC还用于预测每个路边单元所需的存储容量。进行了一项性能研究，以评估应用于iNDC的机器学习算法的性能。结果表明，线性回归和岭回归在内容流行度预测方面是最有效的。为了预测新路边单元的容量，iNDC使用 k-最近邻提供了更好的准确性。