Elsevier

Ad Hoc Networks

Volume 107, 1 October 2020, 102265
Ad Hoc Networks

Enhancing intelligence in traffic management systems to aid in vehicle traffic congestion problems in smart cities

https://doi.org/10.1016/j.adhoc.2020.102265Get rights and content

Abstract

One of the main challenges in urban development faced by large cities is related to traffic jam. Despite increasing efforts to maximize the vehicle flow in large cities, to provide greater accuracy to estimate the traffic jam and to maximize the flow of vehicles in the transport infrastructure, without increasing the overhead of information on the control-related network, still consist in issues to be investigated. Therefore, using artificial intelligence method, we propose a solution of inter-vehicle communication for estimating the congestion level to maximize the vehicle traffic flow in the transport system, called TRAFFIC. For this, we modeled an ensemble of classifiers to estimate the congestion level using TRAFFIC. Hence, the ensemble classification is used as an input to the proposed dissemination mechanism, through which information is propagated between the vehicles. By comparing TRAFFIC with other studies in the literature, our solution has advanced the state of the art with new contributions as follows: (i) increase in the success rate for estimating the traffic congestion level; (ii) reduction in travel time, fuel consumption and CO2 emission of the vehicle; and (iii) high coverage rate with higher propagation of the message, maintaining a low packet transmission rate.

Introduction

Currently, there has been a rapid increase in the number of vehicles transiting in large cities. This increase attributed by population growth and the use of vehicles as mode of transport [1], [2], [3], causes several problems in the transport system such as congestions, CO2 emission, noise and others. Moreover, there are other problems that affect the traffic [4] such as traffic diversion and the drivers’ lack of safety due to an event on the urban environment. Such problems seriously affect the economy when considered as a whole. In Brazil, the estimated cost caused by congestion in the cities surpasses R$ 80 billion (BRL) per year [5]. On the other hand, in the European Union, such cost accounts for approximately 2% of its Gross Domestic Product (GDP) [6], and in the USA it accounts for more than U$ 160 billion [7].

It should be noted that the issue of traffic jam in the cities is not a question easily solved. Although the expansion of the transport network assisting in the congestion issue, this expansion may be unfeasible due to high financial costs, the imposed environmental and geographic restrictions as well as the long time for improving the transport infrastructure. Such restraints can be minimized by Traffic Management Systems (TMS) [4], [8], [9], which emerge as a promising alternative to assist in the congestion issue in the cities. TMS can be defined as a special type of service of Vehicular Ad-hoc Networks (VANETs), equipped with communication, detection, and processing technologies to collect data on traffic aiming at improving it, not only concerning certain vehicles, but rather the entire transport system of the city.

Different solutions have been proposed to deal with the problem of traffic jam in cities. Some solutions recommend the best individual routes to avoid congestion [10], [11], which can generate congestion in other areas. Others have been limited to detecting congestion [12], [13], or redirecting vehicles [14], [15], or proposing solutions for specific scenarios [16], [17], such as urban or highway environments, solving only part of the problem. In addition, there are studies whose authors model machine learning algorithms to estimate road congestion [18], [19], [20]. However, such model should comprise a large amount of data, obtained in an implicit manner and provided by TMS, to monitor traffic.

Although there are considerable efforts concerning the matter, increasing the accuracy of correct classification of congested roads to maximize the vehicle flow in the transport infrastructure arises a new research question, namely: how to provide greater accuracy for estimating traffic jam, maximizing the vehicle flow in the transport infrastructure, without increasing the overhead of control information in the network? This question consists in a new issue that has been addressed in literature, and which we investigate in this article.

With this in mind, we propose TRAFFIC, a solution of inter-vehicle communication for estimating the congestion level to maximize the vehicle traffic flow in the transport system. TRAFFIC is based on an ensemble of classifiers that aims at increasing reliability at the time of vehicle traffic classification. Thus, this classification is used as an input to the proposed dissemination mechanism. The dissemination mechanism propagates information based on the Publish/Subscribe communication module that, in addition to sending messages to groups of interest, avoids the problem of broadcast storms. In order to ensure the efficiency of TRAFFIC regarding the traffic jam issue, the proposed solution was compared with other solutions well-established in literature. Results obtained from simulations show increase in the correct classifications of congestion levels, with reduction in the drivers’ travel time, in fuel consumption of vehicles, and in CO2 emission. Moreover, the results show a satisfactory performance of the computational resources of the network, presenting a shorter response time with greater coverage and low overhead in the information dissemination.

The remainder of this article is organized as follows. In Section 2, we present the main studies on detection of vehicle congestion according to TMS. In Section 3, we present the TRAFFIC model, and its performance assessment and obtained results are presented in Section 4. Finally, in Section 5, we present the conclusions and further studies to be conducted.

Section snippets

Related works

Several solutions have been proposed to face the congestion problem in the last few years [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28]. Besides bringing greater comfort and safety to drivers [4], [29], these solutions aim to optimize the traffic of vehicles in the traffic system, to reduce travel time, fuel consumption, and the amount of CO2 emissions in the atmosphere. However, modeling a mechanism to estimate the congestion level with increase in correct classifications

TRAFFIC: Inter-vehicular solution to aid in the vehicle traffic congestion problem in cities

In this section we present TRAFFIC, an inter-vehicular solution to aid in the vehicle traffic congestion problem in cities. To do so, we will introduce the problem formulation as well as an overview of the solution. Additionally, we will describe the data model to estimate congestion on urban roads and the proposed mechanism to detect traffic jam. Finally, we will introduce the mechanism to disseminate data and how the recommendation of a new route is carried out.

Performance evaluation and methodology

In this section, we present the results of the performance evaluation as well as the methodology used to generate results. To this end, TRAFFIC was validated in three steps. In the first step, Section 4.1, we assessed the TRAFFIC capacity in estimating the road congestion level. In the second and third steps, Section 4.2, we performed an assessment of traffic management in TMS using TRAFFIC and an assessment of computer resources of the network in TMS using TRAFFIC, Sections 4.2.1 and 4.2.2

Conclusion and future research

Research shows that one of the main challenges in urban development faced by large cities is related to vehicle congestion. Despite increasing efforts to maximize the flow of vehicles in large cities, to provide greater accuracy to estimate the traffic jam and maximize the vehicle flow in the transport infrastructure without increasing the overhead of information on the control network still remain as issues that are worth researching, and which we addressed in our study. Therefore, we proposed

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

Rodolfo I. Meneguette would like to thanks CNPq (process 407248/2018-8 and 309822/2018-1) for funding the research project. Jó Ueyama would like to thank FAPESP for funding his research project, FAPESP Grant ID 2018/17335-9.

Geraldo P. Rocha Filho ([email protected]) is an Assistant Professor at the Department of Computer Science (CiC) at University of Brasılia (UnB). He received his Ph.D. in Computer Science from the University of São Paulo (USP) in 2018. He received his M.Sc. from the USP in 2014. He was also a post-doctoral research fellow at the Institute of Computing at UNICAMP before joining the UnB. His research interests are wireless sensor networks, vehicular networks, smart grids, smart home and machine

References (43)

  • R. Bauza et al.

    Traffic congestion detection in large-scale scenarios using vehicle-to-vehicle communications

    J. Netw. Comput. Appl.

    (2013)
  • R.S. Schwartz et al.

    A directional data dissemination protocol for vehicular environments

    Comput. Commun.

    (2011)
  • E. Ahmed et al.

    The role of big data analytics in internet of things

    Comput. Netw.

    (2017)
  • G.D. Junior et al.

    Simulação de controle adaptativo de tráfego urbano por meio de sistema multiagentes e com base em dados reais

    Revista Brasileira de Computação Aplicada

    (2015)
  • R. Du et al.

    The sensable city: a survey on the deployment and management for smart city monitoring

    IEEE Commun. Surv. Tutor.

    (2018)
  • Y. Liu et al.

    Exploring data validity in transportation systems for smart cities

    IEEE Commun. Mag.

    (2017)
  • R. Meneguette, R. E. De Grande, A. A. F. Loureiro, Intelligent Transportation Systems, Springer International...
  • M. Cintra

    A crise do trânsito em são paulo e seus custos

    GVExecutivo

    (2013)
  • K.M. H. Allen, M. Stonehill, A summary of the proceedings from the United Nations climate change conference in Doha,...
  • W.T.J. Texas Transportation Institute, D.L. Schrank 2015 urban mobility scorecard, mobility report,...
  • A.M. Souza et al.

    Traffic management systems: a classification, review, challenges, and future perspectives

    Int. J. Distrib. Sens. Netw.

    (2017)
  • F. Cunha, G. Maia, H.S. Ramos, B. Perreira, C. Celes, A. Campolina, P. Rettore, D. Guidoni, F. Sumika, L. Villas, R....
  • R. Doolan et al.

    VANET-enabled eco-friendly road characteristics-aware routing for vehicular traffic

    Vehicular Technology Conference (VTC Spring), 2013 IEEE 77th

    (2013)
  • R. Doolan et al.

    EcoTrec–a novel VANET-based approach to reducing vehicle emissions

    IEEE Trans. Intell. Transp. Syst.

    (2017)
  • S. Wongcharoen et al.

    Twitter analysis of road traffic congestion severity estimation

    Computer Science and Software Engineering (JCSSE), 2016 13th International Joint Conference on

    (2016)
  • A. Nottle et al.

    Distributed opportunistic sensing and fusion for traffic congestion detection

    First International Workshop on Distributed Analytics InfraStructure and Algorithms for Multi-Organization Federations

    (2017)
  • J.S. Pan et al.

    Proactive vehicle re-routing strategies for congestion avoidance

    2012 8th IEEE International Conference on Distributed Computing in Sensor Systems

    (2012)
  • J.S. Pan et al.

    DIVERT: a distributed vehicular traffic re-routing system for congestion avoidance

    IEEE Trans. Mob. Comput.

    (2017)
  • R.I. Meneguette et al.

    Enhancing intelligence in inter-vehicle communications to detect and reduce congestion in urban centers

    (2015)
  • J.N. Severino et al.

    UrbTrans: a system for mobility management of school vans for cities in the interior

    Proceedings of the 10th Latin America Networking Conference

    (2018)
  • G. Araujo et al.

    Um protocolo de identificação e minimização de congestionamentos de trafego para redes veiculares

    Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos-(SBRC)

    (2014)
  • Cited by (43)

    • Fusion of engineering insights and emerging trends: Intelligent urban traffic management system

      2022, Information Fusion
      Citation Excerpt :

      It allows to improve the capacity of urban traffic management and to clear congestion in the intersection of densely populated cities. To estimate congestion level, Rocha Filho et al. modeled a set of classifiers in [102] since relying on an individual classifier not only depends on adjustments in the parameters of the classification but also implies characteristic errors that can lead to incorrect classifications depending on the problem or on the used dataset. Therefore, the authors aim to maximize the accuracy of the correct classification and information dissemination.

    • Predictive Congestion Control based on Collaborative Information Sharing for Vehicular Ad hoc Networks

      2022, Computer Networks
      Citation Excerpt :

      TMS can collect, process, and communicate data related to drivers, roads, traffic jams, and others to improve TMS applications in this connected vehicular scenario. Addressing vehicular traffic management using VANETs has driven many recent works [12]. In common, these techniques require systems with sophisticated communication protocols, managing data synchronization to support decision-making mechanisms.

    View all citing articles on Scopus

    Geraldo P. Rocha Filho ([email protected]) is an Assistant Professor at the Department of Computer Science (CiC) at University of Brasılia (UnB). He received his Ph.D. in Computer Science from the University of São Paulo (USP) in 2018. He received his M.Sc. from the USP in 2014. He was also a post-doctoral research fellow at the Institute of Computing at UNICAMP before joining the UnB. His research interests are wireless sensor networks, vehicular networks, smart grids, smart home and machine learning.

    Rodolfo I. Meneguete is an Assistant Professor of the Institute of Mathematics and Computer Science (ICMC) at the University of São Paulo (USP). He received his Bachelor’s degree in Computer Science from the University of São Paulo, Brazil, in 2006. He received his master’s degree in 2009. He received his doctorate from the University of Campinas (Unicamp), Brazil, in 2013. He did his post-doctorate in the PARADISE Research Laboratory, University of Ottawa, Canada, in 2017. His research interest are in the areas of vehicular networks, resources management, flow of mobility and vehicular clouds.

    José R. Torres Neto Received the Ph.D degree in computer science and computational mathematics at Institute of Mathematical and Computer Sciences, University of São Paulo, São Carlos, SP. He has M.S. degree in computer science from Institute of Computing, University of Campinas (IC-UNICAMP), Campinas, Brazil, in 2015 and B.S. degree in computer science from State University of Piauí (UESPI), Teresina, Brazil, in 2014. His research interests include computational intelligence on IoT environments, middleware for IoT, affective computing, fog computing, mobile computing, distributed systems, ubiquitous computing and monitoring systems.

    Alan Valejo is a postdoctoral fellow at the University of São Paulo. Received the BSc, M.S. and Ph.D. degrees in computer science from University of São Paulo, Brazil, in 2011, 2014 and 2019, respectively. His current research is focused on multilevel methods applied to problems in data science, large-scale complex networks analysis, machine learning, and data mining.

    Li Weigang is a professor and chair of the Department of Computer Science at the University of Brasília (UnB), Brazil. He received his Ph.D. from the Aeronautics Institute of Technology (ITA), Brazil, in 1994. He is a researcher with grant from Brazilian National Council for Scientific and Technological Development (CNPq). He coordinated various research projects from CAPES, CNPq, FINEP, FAPESP and FAPDF and the industry projects with Atech and Boeing Company/Brazil. His research interests include artificial intelligence with emphasis on computation model in air traffic management and data analytics.

    Jó Ueyama is a  Professor of the Institute of Mathematics and Computer Science (ICMC) at the University of São Paulo (USP). Prof. Ueyama is also a Brazilian Research Council (CNPq) fellow. He completed his Ph.D. in computer science at the University of Lancaster (England) in 2006. Before joining USP, he was a research fellow at the University of Kent at Canterbury (England). Jó has published 53 journal articles and more than 100 conference papers. His main research interest includes Computer Networks, Security, and Blockchain.

    Gustavo Pessin got his D.Sc. in Computer Science at the University of Sao Paulo, as a member of the Mobile Robotics Lab. During his D.Sc. Pessin carried out research with the Robotics Lab, at the Heriot-Watt University, Edinburgh, UK and the Communication and Distributed Systems Group, at the Universität Bern, Switzerland. In 2015, Pessin had a Visiting Scholar position within the Media Lab at the Massachusetts Institute of Technology. Currently, Pessin is an Associated Researcher within the Robotics Lab, at the Vale Institute of Technology. The bulk of his research is related to intelligent systems and mobile robots.

    Leandro A. Villas is an Associate Professor and director in the Institute of Computing at the University of Campinas (Unicamp), Brazil. In 2012 he received the title of Doctor of Computer Science at the Federal University of Minas Gerais, having worked in the area of data collection in wireless sensor networks. He spent a year in the PARADISE laboratory at SITE (School of Information Technology and Engineering) at the University of Ottawa, Canada, as part of his sandwich doctorate. In 2007 he received the title of Master in Computer Science from the Federal University of São Carlos, having worked in the area of routing solutions for wireless sensor networks. Leandro has published 50+ papers in international journals and 120+ in conferences. Six of those papers received the best paperaward. Moreover, he received the Latin America Young Professional Award, IEEE Communications Society, and the Excellence Award from the Institute of Computingat Unicamp.

    Fully documented templates are available in the elsarticle package on CTAN.

    View full text