Vehicle-to-Vehicle Wireless Communications for Virtual Traffic Lights

Road safety and traffic congestions are among major concerns today. The Global status report on road safety 2013 indicates a worldwide total number of road traffic deaths of 1.24 million per year. A large part of them occurs at intersections. Intersection management is also critical to reduce the traffic jams in our cities. However, the percentage of intersections that are governed by traffic lights is inevitably limited and significantly increasing the number of traffic lights is not realistic due to the heavy costs of deployment and maintenance. Besides this, traffic lights generally work under fixed or slowly varying time settings, thus inefficiencies are likely to occur.

Today's wireless communication technologies allow to implement a radically new approach to this issue: through accurate positioning systems and short range wireless communications, vehicles could communicate their own position to each other and they could autonomously coordinate the intersection management in a safe and efficient way. A Virtual Traffic Light (VTL) could then be shown to drivers to control who should stop and who goes next.

Supported by the European EIC ICT Labs, in the "Next generation car-to-X" activity, the Istituto di Elettronica ed Ingegneria dell'Informazione e delle Telecomunicazioni (IEIIT) designed, implemented, and tested a Virtual Traffic Light algorithm for intersection management.
Dedicated hardware and software have been installed inside each vehicle, running the VTL algorithm. Communication has been provided through a new standard belonging to the Wi-Fi family, denoted IEEE 802.11p and operating at 5.9 GHz. Thanks to this technology and a GPS, each vehicle communicates its own position in real time and receives the position of the other vehicles approaching the same intersection; inside each vehicle, the algorithm runs and defines the priorities. Then, a direct communication among vehicles is performed to agree upon the algorithm result, and a green or red virtual traffic light is shown to the driver.
In addition to tests conducted in our laboratories adopting virtual coordinates, we validated the application in a real field trial.

Besides improving intersection management where traffic lights are not deployed, this technology can be applied to control fleets of autonomous vehicles in closed areas, such as airports. Furthermore, it is also an important step towards intersection management for the next generation transportation systems, where vehicles will be driverless.

The implemented algorithm

The algorithm flow chart is shown on the left and the algorithm details can be found in [A].
The main principles are as follows.

1) Each vehicle is either a leader (the vehicle of each road which is the nearest one to the intersection) or a follower (otherwise); the next leader to cross the intersection is the intersection leader;

2) The implemented algorithm acts in two phases: i) phase one: when vehicles approach an empty intersection, the first vehicle approaching is the first to cross (parameters are added to modify the priority of each road); this phase only involves the leaders; ii) phase two: the vehicle with priority (i.e., the intersection leader) defines the next vehicle to cross: either its first follower or the leader of another road;

3) To guarantee safe crossing of the intersection, the priority is either agreed (phase one) or received (phase two) using unicast messages with acknowledgment;

4) A number of parameters allow to optimize the intersection management and timers are used to cope with possible message losses.

Download area

- Few slides here [2 MB]

- Test with virtual coordinates, 3 vehicles, here [7.9 MB]

- Test with virtual coordinates, 6 vehicles, here [11.3 MB]

- Test on field, free parking area (no obstacle impairing the communication), 3 vehicles, here [4.9 MB]

- Test on field, rural area with line of sigth (no obstacle impairing the communication), 2 vehicles, here [4.1 MB]

- Test on field, private area with trees and other small obstacles impairing the communication, 2 vehicles, here [81.8 MB] 

- Test on field, private area with a high building close to the intersection impairing the communication, 2 vehicles, here [130.1 MB] 

- Test on field, residential area with buildings impairing the communication, 4 vehicles, here [80 MB] 


The java code with the VTL algorithm is available here (includes our modifications to the GCDC code).

The html, php, and other files used for laboratory tests are available here.


[A] A. Bazzi, A. Zanella, B. M. Masini, "A Distributed Virtual Traffic Light Algorithm Exploiting Short Range V2V Communications", accepted for publication in Elsevier Ad Hoc Networks, DOI:10.1016/j.adhoc.2016.06.006. Available here.

[B] A. Bazzi, A. Zanella, B. M. Masini, G. Pasolini, "A Distributed Algorithm for Virtual Traffic Lights with IEEE 802.11p", EuCNC 2014, Bologna, June 2014. DOI:10.1109/EuCNC.2014.6882621

Short Bio         Publications        Other Links