Um framework baseado em Fog Computing para redução de dados em um sistema de detecção de tráfego para VANETs

Abstract

With the growth in the number of vehicles worldwide in recent years, the adoption of technologies to deal with the consequences that this volume of vehicles can generate for large cities, such as increased congestion on highways, has become necessary. Vehicular Ad-Hoc Networks (VANETs) present themselves as a promising technology in this scenario, assisting in the formation of vehicular networks capable of interconnecting vehicles and infrastructure to understand and deal with traffic congestion. Considering this, the amount of data generated by this environment increases proportionally as the number of vehicles on the roads grows. Consequently, sending data from the vehicular environment to the structure that identifies congestion can become increasingly burdensome from a network utilization perspective, leading to overloads and increased latency, thus hindering quick decision-making. Thus, we propose in this work the construction of a Framework aimed at identifying vehicle congestion, with an approach to reducing, at the fog layer, the data generated by a VANET and then sending only the most relevant data to the cloud for decision-making. In addition to congestion detection, with historical data in time series format, we perform congestion prediction using ARIMA. To work with data reduction, the Framework utilizes simple sampling algorithms and clustering techniques (DBSCAN and XMEANS). The results showed that the use of clustering algorithms in this Framework achieved a significant level of accuracy in detecting traffic congestion along with a pronounced reduction in network usage compared to the Baseline algorithm.