Performance Modeling and Optimization of Connected and Autonomous Vehicles with Reliable Wireless Connectivity

Vehicle-to-infrastructure (V2I) communication contributes to safe and efficient mobility of connected autonomous vehicles (CAVs). In fully automated traffic streams, speed optimization of CAVs is a fundamental challenge. On one hand, increasing the CAVs' speed improves traffic flow, whereas, on the other hand, it increases communication handovers as the CAVs switch from one base station (BS) to another, thus reducing communication data rates. Therefore, a trade-off exists between the communication data rates and CAV traffic flow. In this thesis, I answer the question of determining the optimal active BS density which maximizes the traffic flow subject to CAVs' data rate constraints. Specifically, the proposed optimization framework is designed to maximize the average traffic flow through an aggregate macroscopic traffic flow model while optimizing the active BS density and average CAV speed with network connectivity constraints and optimize individual CAV speeds to maximize average traffic flow through a microscopic traffic flow model.