Title: Performance evaluation of smart cooperative traffic lights in VANETs

Authors: Laécio Rodrigues; Firmino Neto; Glauber Gonçalves; André Soares; Francisco Airton Silva

Addresses: Universidade Federal do Piauí, R. Cícero Duarte, No. 905, Junco, Picos, PI, 64607-670, Piauí, Brazil ' Universidade Federal do Piauí, R. Cícero Duarte, No. 905, Junco, Picos, PI, 64607-670, Piauí, Brazil ' Universidade Federal do Piauí, R. Cícero Duarte, No. 905, Junco, Picos, PI, 64607-670, Piauí, Brazil ' Universidade Federal do Piauí, R. Cícero Duarte, No. 905, Junco, Picos, PI, 64607-670, Piauí, Brazil ' Universidade Federal do Piauí, R. Cícero Duarte, No. 905, Junco, Picos, PI, 64607-670, Piauí, Brazil

Abstract: Vehicular ad hoc network (VANET) is an emerging new type of network, consisting of vehicles as mobile nodes and temporary communication links among these nodes. One of the crucial topics in VANETs is related to how to use traffic lights to optimise vehicle mobility. The traffic lights can work cooperatively to reduce traffic jams by communicating with the vehicles. However, the architecture of smart traffic lights offers challenges related to network latency restrictions and resource constraints. This paper presents a performance evaluation of a cooperative smart traffic light using a stochastic Petri net (SPN) model. The proposed model can calculate the mean response time, resource utilisation, and the number of requests discards. Three case studies are presented to illustrate how useful the model can be. Besides, we conduct real experiments to validate the proposed model by using micro-controllers (Raspberry Pi) that emulate traffic lights. The model is highly flexible, allowing developers and system administrators to calibrate eighteen parameters.

Keywords: analytical modelling; mean response time; MRT; performance; vehicular ad hoc networks; stochastic Petri net; SPN.

DOI: 10.1504/IJCSE.2021.115650

International Journal of Computational Science and Engineering, 2021 Vol.24 No.3, pp.276 - 289

Received: 06 Jun 2020
Accepted: 16 Oct 2020

Published online: 04 Jun 2021 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article