Predictive Intelligent Transportation: Alleviating Traffic Congestion in the Internet of Vehicles

Abstract

Due to the limitations of data transfer technologies, existing studies on urban traffic control mainly focused on isolated dimension control such as traffic signal control or vehicle route guidance to alleviate traffic congestion. However, in real traffic, the distribution of traffic flow is the result of multiple dimensions whose future state is influenced by each dimension’s decisions. Presently, the development of the Internet of Vehicles enables an integrated intelligent transportation system. This paper proposes an integrated intelligent transportation model that can optimize predictive traffic signal control and predictive vehicle route guidance simultaneously to alleviate traffic congestion based on their feedback regulation relationship. The challenges of this model lie in that the formulation of the nonlinear feedback relationship between various dimensions is hard to describe and the design of a corresponding solving algorithm that can obtain Pareto optimality for multi-dimension control is complex. In the integrated model, we introduce two medium variables—predictive traffic flow and the predictive waiting time—to two-way link the traffic signal control and vehicle route guidance. Inspired by game theory, an asymmetric information exchange framework-based updating distributed algorithm is designed to solve the integrated model. Finally, an experimental study in two typical traffic scenarios shows that more than 73.33% of the considered cases adopting the integrated model achieve Pareto optimality.

A systematic review of congestion control in internet of vehicles and vehicular ad hoc networks: Techniques, challenges, and open issues

“…It is crucial to modify each vehicle's data rate to manage network load and avoid channel overload 19,45,69 . Congestion control improves responsiveness, bandwidth utilization, and fair use of network resources 20,30,62 . Prioritization strategy for secure and nonsecure IVC communications can be managed at the control and service channels 73 .…”

“…Zhang et al 62 proposed a transit that is an intelligently connected model based on their relationship in feedback regulation, which can simultaneously increase the effectiveness of prediction road control methods and prediction vehicle route assistance to reduce congestion. The challenges with this model come from how difficult it was to construct a matching solution algorithm that can achieve Pareto optimality for multidimension management.…”

“…According to the investigated articles of this section, we can say that the common innovation between them is the creation of a machine learning‐based framework, which has helped to a good extent in reducing congestion in the VANETs and IoV networks. For example, we can mention the road monitoring system 61 and the road congestion prediction system 62 , which are helpful for congestion reduction in the VANETs and IoV networks. But there are still critical limitations in this category that invoke more research 61 .…”

Transmission line UAV inspection multi-link data congestion control method

“…According to the above steps, we can calculate the priority of data flow of multiple links for UAV inspection [11][12][13], and make reasonable scheduling and arrangement.…”

A unified model for the fairness mechanism‐based coordinated vehicle route guidance

“…Traffic congestion is a phenomenon that vehicles cannot travel at a free speed due to excessive road loads [1]. It leads to air pollution, fuel consumption, and other direct or indirect economic losses in almost all modern cities [2]. Improving traffic management systems becomes a considerable challenge for city managers.…”