Network Resources Optimization through Regional Computing forVehicular Big Data

Abstract

The transportation system is shifting toward an Intelligent Transportation System (ITS). Research shows that a singleAutonomous Vehicle (AV) can generate up to 1 Terabyte (TB) datadaily. Data processing occurs in real-time at the vehicle edge to facilitate navigation. However, some of this data is sent to the cloudfor real-time navigation support for other AVs. Even slight delayscan have significant consequences for these vehicles. All AV datais also necessary in the cloud for navigation and training purposes.Unfortunately, the massive volume of Vehicular Big Data (VBD)congests public networks, making real-time communication unfeasible. This study proposes Regional Computing (RC) as a solution,enabling real-time updates and navigation for AVs. Compared toedge servers, RC offers sufficient resources and considerably reduceddelays and costs compared to Cloud Computing (CC). RC substantially reduces communication delay, cost, and network congestion.