Centralised Versus Decentralised Traffic Optimisation of Urban Road Networks: A Simulation Study

Abstract

One of the pivotal challenges presented to urban road traffic controllers is the effective utilisation of transport infrastructure, as a result of growing urbanisation, the finite network capacity, and of the increasing number of road vehicles. In this context, the arrival of connected autonomous vehicles (CAVs) represents a unique opportunity for a fundamental change in urban traffic optimisation, and urban traffic control should take an active role in integrating CAVs into the mobility ecosystem in order to maximise benefits. Traditional approaches, commonly exploited by SATNAVs, are based on a decentralised logic, where each vehicle decides the route to follow in isolation, possibly by considering the current network conditions. The arrival of connected vehicles would allow the exploitation of centralised traffic optimisation, where a central urban traffic controller can suggest routes to vehicles by taking into account the current network conditions, and predicted future evolution. This paper introduces a centralised approach for traffic optimisation of urban road networks, and presents an extensive evaluation of the capabilities of centralised and decentralised approaches. Evaluation is based on a validated and calibrated SUMO simulation model of the town centre of Milton Keynes, United Kingdom.