Prediction of Consecutive Road Node Congestion Based on Queueing Model

With the urbanization of modern cities, traffic congestion has become an inevitable challenge to authorities. A large volume of road traffic demands more space than the available road capacity. To ease the situation, researches have involved in addressing this issue. As a result, Intelligent Transportation System (ITS) as a promising paradigm has emerged. To alleviate traffic congestion, theories and models can be found in the open literature. Compared to the statistical approach of congestion evaluation approaches, this paper develops a congestion prediction model of consecutive road nodes. We present the Hurst parameter estimation of road traffic at rush hours which suggests strong self-similar characteristic. Hence the analytical model examines the queueing performance of consecutive road node by taking traffic self-similarity into account. The validation of the developed model is demonstrated via comparisons between analytical results and simulations on real road traffic data. The developed analytical model is further applied on scheduling the road traffic system dynamically and reasonably.