V-Edge is an innovative framework for vehicular networks and communications, focused on “Vehicular Edge Intelligence” (VEI): artificial intelligence and advanced use cases on board the vehicle and at the network edge. The framework uses exclusively unlicensed frequency bands, and is able to enable several relevant use cases in the vehicular sector, such as computer vision, video streaming, low latency information exchange and other use cases based on Machine Learning. The solution has already been developed into a low-cost off-the-shelf hardware-based PoC on which the framework has been integrated.

Recent developments in the vehicular sector will require systems using advanced AI and Deep Learning techniques as well as transferring large amounts of data between vehicles with low latency.

V-Edge is an innovative framework capable of enabling the practical use of on-vehicle intelligence and the streaming of large quantities of data at low latency that uses only unlicensed bands, solving problems present in "infrastructure-based" networks. V-Edge combines various wireless technologies while maintaining stable connectivity and reducing final costs.

V-Edge includes an intelligent module to optimize data exchange and establish the best nodes to send information to, an advanced wireless stack for exchanging information such as vehicle dynamics and kinematics. This allows for optimized task offloading in the absence of infrastructure support and fixed nodes that manage a centralized approach.

• Vehicular communication
• Artificial intelligence on the vehicle
• Sharing of low latency information, also relating to the resources available on board
• Task offloading for advanced use cases (e.g. Object Detection)

• Use of unlicensed frequencies
• Latencies reduced by up to 65% compared to MEC/cloud solutions
• Lower costs
• Support for VEI use cases even in the absence of support from the infrastructure