Graph-Enabled Intelligent Vehicular Network Data Processing

Intelligent vehicular network (IVN) is the underlying support for connected vehicles and smart city, but there are several challenges IVN data processing due to dynamic structure of network. Graph processing, as one essential machine learning big paradigm, which provide a set scheme, well-designed data. In this paper, we discussed research motivated us address these by using graph technologies. We explored characteristics widely used algorithms frameworks on GPU. Furthermore, proposed graph-based optimization technologies processing. The experimental results show GPU can archive excellent performance