A Decentralized Signal Control for Non-Lane-Based Heterogeneous Traffic Under V2I Communication

Queue- and delay-based back-pressure control algorithms are appropriate for homogeneous traffic with strict lane discipline. Every traffic movement does not form a separate queue and vehicular dimensions are highly varied in non-lane-based heterogeneous traffic conditions. In this paper, we propose a novel decentralized, area occupancy-based back-pressure control for a non-lane-based heterogeneous traffic network under V2I communication. The concept of spatial area occupancy and temporal area occupancy combined with back-pressure (SAO-BP and TAO-BP) control are appropriate for heterogeneous traffic in the absence of lane discipline. We conducted an extensive simulation study to illustrate the efficacy of the proposed algorithms using the Vissim-MATLAB simulation environment. Our results suggest that the proposed SAO-BP and TAO-BP control algorithms significantly improve the performance of traffic network compared with vehicle actuated control and aggregated queue-based back-pressure control deployed on the same network.