Shaghaghi et al. / Front Inform Technol Electron Eng in press
3
intersection send their real-time movement infor-
mation repeatedly to the Road Side Unit (RSU) at the
intersection and it estimates the traffic density of the
intersection. In this approach, the basic traffic signal
timing framework is used including two offshoots:
density measurement for signal timing, and traffic
signal sign propagation. Though the idea is re-
markable, it is confronted with excessive load of the
network due to the propagation of a large amount of
broadcast packets. Moreover, frequent variation of
vehicle speeds cause inaccuracy in the estimation of
vehicles’ position near the intersection and impreci-
sion of density assessment for traffic signal timing.
Similarly, V2I communications are used by some
authors (Gradinescu et al. , 2007). In this approach,
while vehicles travel and use one-hop car-to-car
communication, the RSUs detect the movement
information exchanged between them to estimate the
traffic density near an intersection. Though this ap-
proach can decrease the vehicles’ delay at an inter-
section, two noticeable difficulties exist: excessive
propagation of broadcast packets, and inaccurate
perceiving of vehicle’s platoon density close to the
intersection. In a similar way, (Priemer and Friedrich,
2009) suggested the deployment of RSUs beside the
loop detectors along the road as a multi-detection
method. In this approach, vehicles in the communi-
cation range of RSUs broadcast their movement
information; thus traffic information will be assessed
from both the RSU and loop detector devices using
dynamic programing and complete enumeration.
This method, in addition to having similar network-
ing deficiencies as the former approach, suffers from
redundancy in assessment of vehicles’ information,
which causes the time-consuming procedure of de-
tecting iterative information and slowing down the
traffic measurement. In another study by (Nafi and
Khan, 2012), a different approach called intelligent
road traffic signaling system (IRTSS) is proposed. In
this approach, while vehicles approach the RSU at
the intersection, they receive periodic broadcast
messages including traffic information and the cur-
rent traffic signal sign. Then each vehicle replies its
movement information. Accordingly, the traffic
controller calculates the traffic saturation at the in-
tersection and provides the appropriate traffic signal
timing. Though this approach is efficient in de-
creasing the vehicle waiting time and optimizing fuel
consumption, the absence of a comprehensive algo-
rithm for the traffic signal timing and the rise of
network contention besides packet loss in congested
traffic conditions cannot be neglected. In another
study by (Kwatirayo et al. , 2013), a different ap-
proach by taking advantage of V2I communication
for broadcasting the vehicles’ information to the
RSU at the intersection was proposed. This approach
concentrates more on obtaining realistic traffic in-
formation by using a specific algorithm that consid-
ers the relative positions of vehicles from the inter-
section and groups them into specific platoons with
the same destination. Though this approach is effi-
cient in reducing vehicles’ travel time and delay at
the intersection, the effects of message loss and
network contention were not applied in the experi-
ments. In a different study by (Pandit et al. , 2013),
the deployment of RSUs at the intersection was
suggested. Thereby while vehicles approach the
intersection periodic broadcasting of their movement
information is done. This approach formulates the
traffic signal timing as a job scheduling service. This
method groups the vehicles to equally-sized platoons
according to their location and speed, names them as
jobs, and then employs an algorithm called the
“oldest job first” in order to control the green light
interval for each movement. This approach has been
effective in decreasing the vehicles’ delay in light
and moderate congestion scenarios; however, it has
not been very effective under heavy traffic loads. In
the study by (Li et al. , 2014), different types of
traffic control policies by taking advantage of V2I
communication were investigated. In this study the
use of both feedback and feedforward— measure-
ment for the prediction of traffic demands prior to
entering the monitoring area, characteristics of traf-
fic controlling, is suggested. Also the sending of
information regarding vehicles’ desired route from
their origin to their destination as well as the com-
mon movement information that is periodically sent
to the RSUs is hypothesized. Accordingly, the traffic
controller perceives the entire trajectory that vehicles
are aiming to travel and sketches upcoming traffic in
a series of time slices to define the effective traffic
signal timings. However, the proposed approach
involves high demands of information propagation
and large sizes of transmission data, which are con-