Simulation-Based Traffic Management Model To Minimize The Vehicle Congestion in
Simulation-Based Traffic Management Model To Minimize The Vehicle Congestion in
Abstract—The proliferation of technological advancements lenging. The development of the transportation infrastructure
within urban areas attracts individuals to urban centers, thereby of the country increases with the number of vehicles. However,
resulting in a surge of vehicular congestion during periods of the composition of the infrastructure of the road, including
high demand. Nevertheless, the problem of congestion at points
of intersection plays a substantial role in the initiation and spread the traffic signs and traffic signal lights or stop lights, plays a
of traffic congestion. The majority of conventional methods for prominent role in congestion management and traffic control,
traffic control require a greater degree of computational intricacy. as per the conventional traffic-management strategy [2]. Traffic
However, the duration of traffic light cycles remains consistent signal systems are implemented in metropolitan areas as a
and equal for all directions with lane control, which causes means to address, regulate the traffic flow and alleviate issues
unnecessary vehicle congestion. In this paper, we propose a model
that additionally utilizes an approach to dynamically adjust related to traffic congestion [1], [3], [7]. Most traditional forms
traffic light schedules in four-way intersections in real-time, of traffic control necessitated a higher level of computing
taking into consideration variations in traffic density. We have complexity. Nevertheless, the amount of time splitting of traffic
also considered mainly two traffic scenarios, such as light traffic lights remains consistent and uniform across all directions with
conditions and heavy traffic conditions. In our simulation experi- lane controlling [2]. The lack of uniformity in patterns of
ment, it is clearly seen that our proposed approach of dynamically
changing the traffic light schedule outperforms conventional traffic from various directions leads to an imbalance in time
traffic control with a fixed time duration. Therefore, our proposed allocation, thereby causing inefficiency in resource utilization.
model has the potential to serve as the preferred choice for These systems adhere to inefficient traffic policies and are
transportation planning in smart-city-oriented transport systems. unable to effectively manage traffic in all cities with the current
Keywords—Traffic Signal, Vehicle Congestion, Traffic Control, system.
Green-Signaling
This conventional system lacks the capability to accurately
I. I NTRODUCTION detect the presence of vehicles on each route. Consequently,
whenever, a route is empty, the traffic signal operation for
In connection with the rise in human population, the rapid that particular route inefficiently consumes time. Hence, urban
expansion of automobiles has caused congestion, pollution, areas are seeking a more viable traffic management system,
and logistical transportation delays in major metropolitan areas alternative to the existing. This traffic management system
[1], [2]. Therefore, real-time, accurate, and efficient trans- would enable the efficient and effective way of resource
portation, a comprehensive transportation management system utilisation by improving safety and reducing delays. The
is essential [3]. The goal of smart transportation engineering external factors such as traffic volume, driver’s intention and
is to comprehensively manage resources and vehicles in the speed of the vehicle would decide in the traffic congestion.
transportation field through the integration of vehicle network- These systems have the potential to mitigate traffic congestion
ing technology, artificial intelligence (AI) technology, Fog and improve transportation efficiency. The intelligence and
Computing, Cloud Computing and Embedded sensor-based long-term viability of those systems are contingent upon the
technology [1], [2], [4], [5]. This integration of technology efficacy of the implemented solutions. However, the issue of
paves the way toward a more mechanized and intelligently congestion at intersections is a significant factor contributing
managed and automated system. This could be achieved with to the initiation and propagation of traffic jams.
the proper management of two dependable elements, such To the best of our knowledge, this is the first paper present-
as congestion management and traffic control. Congestion, ing a model to address the traffic congestion particularly, based
delays, and an increased death toll are some of the problems on the density of traffic. This study details the planning and
that arise from improper management of them [2], [6]. execution through simulation of an adaptive traffic light system
With the increasing expansion of road infrastructures and in the context of congestion management. Initially, in this
the proliferation of automobiles, the task of effectively manag- paper, we propose this model to make use of the simulation-
ing traffic and transport networks becomes increasingly chal- based method to automatically revise traffic light schedules
979-8-3503-5813-1/23/$31.00 © 2023 IEEE in real-time to account for changes in traffic density. The
227
Authorized licensed use limited to: Zhejiang University. Downloaded on August 01,2024 at 15:30:13 UTC from IEEE Xplore. Restrictions apply.
proposed model is also built around three primary components: layer, employs a variety of data sources and assessment of
vehicles (the number of vehicles), infrastructure (the four- the traffic conditions. Secondly, fog layer, utilizes this data
way junction), and events (the interchange of the vehicles to make effective decisions. Thirdly, cloud layer, serves as
and traffic lights). This model can be the go-to option for a permanent storage solution for the data analytics.The final
transportation planning in smart-city-based transport systems. layer offers auxiliary services for the purpose of traffic man-
Further, this would lessen traffic delays and congestion, lessen agement. In this study, the authors discuss a novel algorithm
the likelihood of accidents, and enhance the quality of the designed for the purpose of effectively managing congestion
travel experience. at intersections. Additionally, this implementation enhance the
The simulation used in the study is more explainable and decision-making process, ensuring reliability and accuracy.
easily understandable according to the nature of the traffic. The efficiency of the proposed smart-traffic light algorithm
Our findings are as follows. The traditional signaling method was also assessed, along with the accuracy of the tweet
to open the road with Green signal for the small duration of classification and image-processing algorithms [4].
fixed time is more efficient than the large duration of fixed The authors in [2], introduce the conceptualization and ex-
time. Further, our proposed signaling method to open the road ecution of an Adaptive Traffic Management system to address
with Green signal considerably outperforms when compared all potential challenges within the transportation system to
to traditional signaling method. identify inadvertent anomalies. The proposed model contin-
The paper opens by conducting a comprehensive literature uously updates traffic signal schedules based on the volume
review in Section II. In Section III, we outline the methodology of traffic and estimated movements from nearby intersections.
used in our research work. Next, we perform the simulation The results of the experiment indicate that the system exhibited
based experiment to evaluate our results in Section IV. Finally, superior performance compared to the conventional traffic-
this paper is concluded with the conclusion presented in management strategy in smart-city-based transport systems.
Section V. The effective use of big data has been suggested as a
potential solution for addressing the problem of congestion. A
II. L ITERATURE R EVIEW novel approach for predicting Short-Term Traffic Congestion
The explicit contributions mentioned in this section are (STTC) for every 5 minutes over 1 hour, by introducing a
using cutting-edge technologies with recent studies conducted hybrid model that combines the Xception architecture (for
to examine and to provide an appropriate solution of reducing feature detection within the dataset) with a Support Vector
traffic congestion in cities. Machine (SVM) classifier (for enhancing the accuracy of
In [6], authors focused on mitigating congestion in urban output prediction) is discussed in [7]. This is achieved through
areas by utilizing a multi-variable model to estimate the appro- the implementation of weight regulation techniques and a
priate charge. The data utilized in this study was obtained and finely-tuned binary hyperplane mechanism.
examined primary congested zones from surveys. The model
evaluation comprised of Fisher multiple regression analysis III. M ETHODOLOGY
and the coefficient of determination. The aforementioned vali- In this section, we illustrate a highway traffic signaling
dations have demonstrated that the model possesses statistical model that we developed to reduce the vehicle congestion.
significance. A four-way junction of Road-1, Road-2, Road-3, and Road-4
The authors in [5], present a proposed framework for a with double lane and various traffic scenarios are shown in
smart traffic management system that leverages contemporary Fig. 1. In this study, we consider the left-handed entry and
technologies including the Internet of Things (IoT), cloud right-handed exit transportation method. Solid arrows in each
computing, 5G, and data science to enhance traditional traffic road denote the direction of flow of vehicles. Dotted arrows
management systems and effectively address the identified is- from one road to all other roads denote the direction of flow
sues. The proposed solution aims to promote mobility through while a particular road is opened with the Green signal.
the utilization of roadside messaging agents, which provide In Fig. 1a, Road-1 is opened with the Green traffic signal
up-to-date traffic information regarding congestion and unfore- and given priority for the transportation. Therefore, the vehi-
seen traffic incidents. Therefore, time will be saved particularly cles in Road-1 flow from the Road-1 to all other roads Road-2,
during peak hours. As an integral component of proposed Road-3, and Road-4 while Road-2, Road-3, and Road-4 are
methodology, each signal dashboard is equipped with traffic closed. This is the maximum possible chances of the flow of
data with the use of a case study. The findings of the case vehicles in a four-way roads. There are also other possibilities
study indicate that the proposed strategy demonstrates superior that we addressed in our study, such as the vehicles flow to
performance compared to the existing alternatives. any one road (say, to Road-2) or any two roads (say, to Road-3
In [8], the authors presented several technologies such as and Road-4).
artificial intelligence (AI), machine learning, and IoT with Similarly, in Fig. 1b, the direction of flow from Road-2 to
their associated components embedded connected gadgets, IoT all other roads is shown while Road-2 is opened and Road-3,
detectors, and smart traffic-control systems for the purpose of Road-4, and Road-1 are closed. These similar pattern can also
collecting and analyzing data. The paper presents a framework be seen in Fig. 1c and Fig. 1d when Road-3 and Road-4 are
that consists of four distinct layers. The first layer, sensing opened with the Green signal respectively.
228
Authorized licensed use limited to: Zhejiang University. Downloaded on August 01,2024 at 15:30:13 UTC from IEEE Xplore. Restrictions apply.
In all the scenarios of the existing traffic signaling system addition to that Fig. 2 depicts the number of vehicle currently
shown in Fig. 1, equal time duration will be given to open available in each road. In such scenario, Road-1, Road-2,
all four roads with the Green signal in a Round-robin manner. Road-3, and Road-4 consist of 9, 3, 15, and 7 vehicles
Suppose, Road-1 is opened for 15 Seconds, then all the other respectively. Instead of assigning a fixed duration to give
roads are also opened 15 Seconds individually in a minute. priority to vehicles on each road, we calculate the duration
Though the time duration given to open all roads is varied, of activating Green signal based on the density of vehicles
the time duration is not systematically calculated without in each road. For example, as shown in Fig. 2, there are 34
considering the traffic density where vehicles are waiting for vehicles in all roads. If a minute is divided in the ratio of
Green signal in the road. This causes unnecessary vehicle vehicles in all four roads, the duration of Green signal for
congestion due to an inefficiency of traffic signaling system. Road-1 is approximately 16 Seconds. This is approximately
Therefore, we develop a model to allocate time for the 5 Seconds, 26 Seconds, and 13 Seconds for the duration of
vehicles waiting in the road for the Green signal which is more Green signal of Road-2, Road-3, and Road-4 respectively.
efficient to reduce the vehicle congestion during the traffic In contrast, providing 15 Seconds equally to open Road-
signal transportation. 2 with 3 vehicles and to open Road-3 with 15 vehicles is
Typical four-way road infrastructure with similar entry and obviously inefficient. Instead of that, providing 5 Seconds for
exit methods is shown in Fig. 2 as explained in Fig. 1. In Road-2 with 3 vehicles and 26 Seconds for Road-3 with 15
Road-4
Road-2
Road-2
Road-3 Road-3
(a) (b)
Road-1 Road-1
Road-4
Road-4
Road-2
Road-2
Road-3 Road-3
(c) (d)
Fig. 1: Road infrastructure with several traffic scenarios when (a) Road-1 is opened, (b) Road-2 is opened, (c) Road-3 is
opened, and (d) Road-4 is opened with Green signal.
229
Authorized licensed use limited to: Zhejiang University. Downloaded on August 01,2024 at 15:30:13 UTC from IEEE Xplore. Restrictions apply.
Road-1 Algorithm 1 Allocate time duration for a traffic signal in a
highway based on the density of traffic.
Input: T = (R, V ), R: roads, V : vehicles, e: number of en-
tering vehicles, x: number of exiting vehicles, Cycle i =
(e, x, R) with traffic flow in all roads.
Output: Finding the duration of Green signal of each road
based on the density of traffic.
Road-4
Road-2
1: Step-1
2: For all roads in the traffic signal
3: if Rk is closed with the Red signal with fixed duration
where k=1 to 4 then
4: Find, number of entering vehicles Ve on road Rk
5: end if
6: if Rk is opened with the Green signal with fixed duration
where k=1 to 4 then
Road-3 7: Find, number of entering vehicles Ve on road Rk
8: Find, number of exiting vehicles Vx on road Rk
Fig. 2: Typical four-way road infrastructure with different 9: end if
density of traffic 10: Step-2
11: Find, the density of traffic for each road Rk
12: Step-3
vehicles is much efficient.
13: Find, the duration of Green signal in each road Rk based
Further, to implement time duration allocation method based
on the density of traffic.
on the density of traffic in each road, we proposed an algorithm
14: Step-4
shown in Algorithm 1. This algorithm consists of number of
15: if Rk is opened with the Green signal with calculated
roads, number of vehicles, number of entering vehicles, and
duration then
number of exiting vehicles as inputs. This algorithm is used
16: Allow, entering vehicles Ve on road Rk
to calculate the time duration for a traffic signal in a highway
17: Allow, exiting vehicles Vx from road Rk
based on the density of traffic and allow the vehicles according
18: else
to such time duration. Therefore, this approach definitely gives
19: Allow, entering vehicles Ve on road Rk
significant impact on the traffic signaling system to reduce
20: Reject, exiting vehicles Vx from road Rk
the vehicle congestion. The numerical results to justify this
21: end if
approach are given in the next section.
230
Authorized licensed use limited to: Zhejiang University. Downloaded on August 01,2024 at 15:30:13 UTC from IEEE Xplore. Restrictions apply.
20 20
Traditional Green signalling method - 20 Seconds per road Traditional Green signalling method - 20 Seconds per road
18 Proposed Green signalling method - Total duration of 80 Seconds per cycle 18 Proposed Green signalling method - Total duration of 80 Seconds per cycle
16 Traditional Green signalling method - 15 Seconds 16 Traditional Green signalling method - 15 Seconds
Number of Vehicles
Number of Vehicles
Proposed Green signalling method - Total duration of 60 Seconds per cycle Proposed Green signalling method - Total duration of 60 Seconds per cycle
14 Traditional Green signalling method - 10 Seconds 14 Traditional Green signalling method - 10 Seconds
12 Proposed Green signalling method - Total duration of 40 Seconds per cycle 12 Proposed Green signalling method - Total duration of 40 Seconds per cycle
10 10
8 8
6 6
4 4
2 2
0 0
10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100
Number of Cycles Number of Cycles
(a) (b)
20 20
Traditional Green signalling method - 20 Seconds per road Traditional Green signalling method - 20 Seconds per road
18 Proposed Green signalling method - Total duration of 80 Seconds per cycle 18 Proposed Green signalling method - Total duration of 80 Seconds per cycle
16 Traditional Green signalling method - 15 Seconds 16 Traditional Green signalling method - 15 Seconds
Number of Vehicles
Number of Vehicles
Proposed Green signalling method - Total duration of 60 Seconds per cycle Proposed Green signalling method - Total duration of 60 Seconds per cycle
14 Traditional Green signalling method - 10 Seconds 14 Traditional Green signalling method - 10 Seconds
12 Proposed Green signalling method - Total duration of 40 Seconds per cycle 12 Proposed Green signalling method - Total duration of 40 Seconds per cycle
10 10
8 8
6 6
4 4
2 2
0 0
10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100
Number of Cycles Number of Cycles
(c) (d)
Fig. 3: Comparison between Traditional Green signaling methods with 20, 15, and 10 Seconds per road against Proposed
Green signaling methods with total time duration 80, 60, and 40 Seconds per cycle on (a) Road-1, (b) Road-2, (c) Road-3,
and (d) Road-4 in light traffic condition.
as shown in Fig. 3a. This is comparatively higher, when signaling method as compared in Fig. 3. However, for this
compared to traditional Green signaling method with duration comparison, we change the traffic environment to heavy traffic
of 15 Seconds per road on Road-1. Even the considerable num- condition instead of light traffic condition. In Fig. 4a, it
ber of vehicles are reduced when traditional Green signaling is clearly seen that the proposed Green signaling method
method with duration of 10 Seconds per road is implemented. with the total duration of 80 Seconds per cycle significantly
Secondly, we consider the number of vehicles on Road-1 when outperforms when compared to traditional Green signaling
the duration varies with the total duration of 80 Seconds per method with duration of 20 Seconds per road. This better
cycle in the proposed Green signaling method as shown in Fig. performance could even be seen in all other proposed Green
3a. This is comparatively higher, when compared to proposed signaling methods with the total duration of 60 Seconds per
Green signaling method with total duration of 60 Seconds per cycle and 40 Seconds per cycle when compared to traditional
cycle on Road-1. Even the considerable number of vehicles are Green signaling methods with 15 Seconds per road and 10
available when proposed Green signaling method with total Seconds per road respectively. Similarly, the same pattern of
duration of 40 Seconds per cycle. number of vehicles on each road such as Road-2, Road-3,
Similarly, same comparisons for Road-2, Road-3, and Road- and Road-4 can also be observed in every number of cycle as
4 are shown in Fig. 3b, Fig. 3c, and Fig. 3d respectively. shown in Fig. 4b, Fig. 4c, and Fig. 4d respectively.
Particularly, in Fig. 3d, the number of vehicles for all signaling Finally, a significant benefit could be realized when high-
methods for all the number of cycles on Road-4 is compara- way traffic systems are operated using our proposed Green
tively higher. This is because of the condition of the road as signaling method, particularly, with the total time duration of
more vehicle flow can normally be occurred in Road-4. 40 Seconds per cycle when compared to the traditional Green
In Fig. 4, similar comparison has been performed for signaling method with the time duration of 20 Seconds per
traditional Green signaling method against proposed Green road.
231
Authorized licensed use limited to: Zhejiang University. Downloaded on August 01,2024 at 15:30:13 UTC from IEEE Xplore. Restrictions apply.
Traditional Green signalling method - 20 Seconds per road Traditional Green signalling method - 20 Seconds per road
45 Proposed Green signalling method - Total duration of 80 Seconds per cycle 45 Proposed Green signalling method - Total duration of 80 Seconds per cycle
40 Traditional Green signalling method - 15 Seconds 40 Traditional Green signalling method - 15 Seconds
Number of Vehicles
Number of Vehicles
Proposed Green signalling method - Total duration of 60 Seconds per cycle Proposed Green signalling method - Total duration of 60 Seconds per cycle
35 Traditional Green signalling method - 10 Seconds 35 Traditional Green signalling method - 10 Seconds
30 Proposed Green signalling method - Total duration of 40 Seconds per cycle 30 Proposed Green signalling method - Total duration of 40 Seconds per cycle
25 25
20 20
15 15
10 10
5 5
0 0
10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100
Number of Cycles Number of Cycles
(a) (b)
Traditional Green signalling method - 20 Seconds per road Traditional Green signalling method - 20 Seconds per road
45 Proposed Green signalling method - Total duration of 80 Seconds per cycle 45 Proposed Green signalling method - Total duration of 80 Seconds per cycle
Traditional Green signalling method - 15 Seconds Traditional Green signalling method - 15 Seconds
40 40 Proposed Green signalling method - Total duration of 60 Seconds per cycle
Number of Vehicles
Number of Vehicles
Proposed Green signalling method - Total duration of 60 Seconds per cycle Traditional Green signalling method - 10 Seconds
35 Traditional Green signalling method - 10 Seconds 35
Proposed Green signalling method - Total duration of 40 Seconds per cycle
30 Proposed Green signalling method - Total duration of 40 Seconds per cycle 30
25 25
20 20
15 15
10 10
5 5
0 0
10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100
Number of Cycles Number of Cycles
(c) (d)
Fig. 4: Comparison between Traditional Green signaling methods with 20, 15, and 10 Seconds per road against Proposed
Green signaling methods with total time duration 80, 60, and 40 Seconds per cycle on (a) Road-1, (b) Road-2, (c) Road-3,
and (d) Road-4 in heavy traffic condition.
V. C ONCLUSIONS R EFERENCES
With the continuous growth of road infrastructure and the [1] Y. Zhai, Yujia, Yan Wan, and Xiaoxiao Wang, “Optimization of traffic
congestion management in smart cities under bidirectional long and short-
widespread use of automobiles, the management of traffic and term memory model,” Journal of Advanced Transportation, pp. 1-8, Apr.
transportation networks presents a growing challenge. Traffic 2022.
signal systems are commonly deployed in urban areas as a [2] U. K. Lilhore, et al., “Design and implementation of an ML and IoT
based adaptive traffic-management system for smart cities,” Sensors-22,
strategy to effectively manage and control the movement of no. 8, Apr. 2022.
vehicles, with the aim of mitigating the negative impacts asso- [3] M. Saleem, et al., “Smart cities: Fusion-based intelligent traffic con-
ciated with traffic congestion. Therefore, in our research work, gestion control system for vehicular networks using machine learning
techniques,” Egyptian Informatics Journal, vol. 23, no. 3 pp. 417-426,
initially, we have proposed a model with an algorithm that Sep. 2022.
utilizes a simulation-based approach to automatically adjust [4] Y. Alsaawy, et al., “A comprehensive and effective framework for traffic
traffic light schedules in real-time, considering variations in congestion problem based on the integration of IoT and data analytics,”
Applied Sciences, vol. 12, no. 4, pp. 2043, Feb. 2022.
traffic density in each road. We also examined the traditional [5] M. Humayun, et al., “Smart Traffic Management System for Metropoli-
Green signaling method, exploring its implementation with tan Cities of Kingdom Using Cutting Edge Technologies,” Journal of
different total time duration of 80 Seconds, 60 Seconds, and Advanced Transportation, no. 3, pp. 1-13, Sep. 2022.
[6] P. González Aliste, Pablo, Iván Derpich, and Mario López, “Reducing
40 Seconds. Most importantly, we could be able to achieve urban traffic congestion via charging price,” Sustainability, vol. 15, no.
the significant advantage when the highway traffic system 3, pp. 2086, Jan. 2023.
is implemented using our proposed Green signaling method, [7] M. Anjaneyulu and M. Kubendiran, “Short-Term Traffic Congestion
Prediction Using Hybrid Deep Learning Technique,” Sustainability, vol.
which operates with a total cycle duration of 40 Seconds 15, no. 1, pp. 74, Dec. 2022.
whereas the traditional Green signaling method performed [8] C. Cenedese, et al., “Incentive-based electric vehicle charging for man-
lower with the fixed time duration of 20 Seconds per road. As aging bottleneck congestion,” European Journal of Control, vol. 68, Nov.
2022.
a result, this approach would definitely reduce traffic delays
and congestion in the transportation management system.
232
Authorized licensed use limited to: Zhejiang University. Downloaded on August 01,2024 at 15:30:13 UTC from IEEE Xplore. Restrictions apply.