PRESTO

PRESTO: Predictive Real-Time Network Management for Enhanced Automotive and Transport Services

The overall objective of the PRESTO project is to address the most relevant aspects of predicting QoS in cellular-vehicular-to-everything C-V2X for network optimization and management. Particularly, we are interested in the reliability and real-time nature of these predictions for advanced vehicular control services (hence the name PRESTO). To approach the exciting goal above, we will investigate fundamentally new ML methods, including a distributed/federated learning model of a system of vehicles that maintain knowledge about their own “neighborhood” and use the C-V2X cloud to arrive at federated orchestration of vehicle actions. We will fundamentally extend ML methods to predict if packets can be delivered within a desired latency window and if wireless resources must be provided or reconfigured in order for C-V2X services to remain available.