Skip to Main content Skip to Navigation

Mobility and traffic models for VANETs

Abstract : The ever-growing traffic congestion is becoming a serious issue that needs to be urgently addressed. The growing number of traffic accidents, the environmental footprint of transport, commercial speed of public transportation and traffic jams are few examples of the adverse consequences of congestion. This calls for an upgrade of the transport system in order to ensure road users safety and effectively enhance the transportation infrastructure. ITS have grown in importance as a promising solution to the congestion issue. These systems rely on the most advanced technologies to provide intelligent and robust safety services that seek to prevent road incident that may threaten the life of passengers. One of the most stringent requirements of these applications is an error-free and reliable transmission of data packets. VANET were designed specifically to provide a wireless communication infrastructure to allow vehicles and road equipment to exchange traffic data. The particular feature of this network is the highly dynamic mobility which results in frequent changes in the topology and density of the network. This has negative effects on the network performance which does not allow to cater safety applications requirements. In this thesis, we address specifically channel access methods for VANET that are based on TDMA method. TDMA has been proven the most suitable access technique for VANET as it allows a single node to access the channel at any time slot. However, conventional TDMA-based protocols might encounter difficulties in a dynamic networks such as access collision and unfair use of resources. Hence, a good understanding of mobility will allow the design and evaluation of channel access methods that are efficient and robust even in a mobile environment. Although faithful mobility models are found in the literature, they fail to accurately capture some aspects of vehicular mobility. The traffic behaviour is influenced by several factors such as road layout, speed limits, traffic rules and individual vehicle’s behaviour. Consequently, it is compulsory to include all these features in a mobility model for accurate results. In this context, we develop in this thesis stochastic Markov chain models based on real vehicle traces collected by RSUs using V2X communication to emulate vehicular behaviour in both urban and highway roads. The proposed models have the twofold benefit of modelling and predicting traffic. Using a direct numerical resolution technique, traffic density, waiting queue lengths, travel times and delays are predicted. The predicted traffic density is then exploited to design a Traffic-aware TDMA channel access method that aims to reduce access collisions and enhance resource utilization through mobility prediction and clustering. To evaluate the performance of the proposed method, a queue-based mobility simulation framework was developed using the SimEvents toolbox. The simulation framework allows the generation of synthetic measures relevant to the assessment of road network performance. The TA-TDMA MAC protocol was then implemented and compared with an existing MAC protocol called VeMAC, under different scenarios and environments. The proposed solution has shown better results than the VeMAC protocol in terms of efficiency and robustness against topological changes.
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Friday, April 1, 2022 - 3:51:24 PM
Last modification on : Monday, April 4, 2022 - 8:33:25 AM


Version validated by the jury (STAR)


  • HAL Id : tel-03627931, version 1


Oumaima El Joubari. Mobility and traffic models for VANETs. Networking and Internet Architecture [cs.NI]. Université Paris-Saclay, 2022. English. ⟨NNT : 2022UPASG018⟩. ⟨tel-03627931⟩



Record views


Files downloads