Modelagem de zonas de ancoragem para disseminação de conteúdo flutuante em redes veiculares utilizando métricas de centralidade temporais

Abstract

Vehicle networks are large networks composed of vehicles with technologies that enable the exchange of information. These networks have great capacity for the production and dissemination of content, however, the constant topological changes make it difficult to exchange information between nodes and can invalidate the information, generating information losses, packet delays, improper disconnections, among other problems. In vehicular networks there are places where the content generated by the sensor nodes reaches a larger audience, these places which we call Anchorage Zones are used for exchanging relevant information with greater connectivity. The information disseminated in an anchor zone is known as Floating Content because it has local, temporal and spatial validity. We developed a study using complex network metrics such as, Betweenness and Degree together with flow measures, and, we incorporated the network’s temporal aspect to these metrics to determine the best zones for exchanging information on a vehicular network. In our evaluations we use real databases to statically evaluate the performance of our approach, and simulate the behavior of a vehicular network when considering our approach. The results show that the methodology used is effective for characterizing anchor zones in vehicular networks, because the information extracted with the discretized time is better approximated to reality, eliminating edges that no longer exist between the nodes of the network, differently from the aggregate time that stores information for edges that no longer exist.