Intelligent Intersection Management (IIM) aims to optimize urban traffic flow but face inherent limitations based on its operational mode. Sequential IIM suffers from inefficiencies under high demand, rigid cyclic operations, and limited scalability. Parallel IIM struggles with turning conflicts, unbalanced traffic flows, and strict lane discipline requirements. Synchronous IIM relies heavily on precise timing, reducing flexibility. To address these challenges, we proposed the Software-Defined Intelligent Intersections (SDI 2 ) framework, an SDN-based IIM solution that intelligently orchestrates intersection management through centralized control. The SDI 2 framework dynamically switches between parallel and synchronous modes based on real-time traffic patterns, effectively accommodating individual vehicles and groups. In this work we evaluate SDI 2 against conventional, adaptive, and reactive approaches across various traffic densities. Results show that SDI 2 improves traffic flow by reducing green signal reaction times and minimizing unnecessary waiting delay by up to 3.5s/veh, reducing energy waste by 7g/veh, and associated PM x emissions by 0.5mg/veh on average, when compared to the synchronous approach, which was the second-best performing IIM.
Reddy et al. (Thu,) studied this question.