Guidelines for the Design of Vehicular Cloud Infrastructures for Connected Autonomous Vehicles

Initiatives all around the world, funded by industries and governments, have devoted tremendous efforts to develop connected and autonomous vehicles (CAVs). Nowadays, different CAV prototypes are being tested on the roads. The natural next step in the evolution of this area will be the provisioning of infotainment applications for enjoyable commutes and trips. In this article, we discuss the motivation for vehicular clouds for content delivery aimed to support applications for CAVs. First, we review the main concepts of vehicular clouds. Consequently, we discuss the needs of vehicular clouds to support content distribution for applications in CAV scenarios. In addition, we highlight fundamental challenges regarding vehicular cloud that need to be tackled: self-organization, service discovery and management, and resource discovery and allocation. By doing so, we analyze the current works in the literature and portray their limitations. Moreover, we provide some guidelines to deal with the challenges of vehicular clouds for CAV applications. Finally, we present future research directions that might be considered for the design of large-scale vehicular cloud infrastructures for CAV applications.