Novel computational intelligence-based model for effective traffic management in intelligent transportation system

IAES International Journal of Artificial Intelligence (IJ-AI)
Vol. 13, No. 3, September 2024, pp. 2524~2539
ISSN: 2252-8938, DOI: 10.11591/ijai.v13.i3.pp2524-2539  2524
Journal homepage: http://ijai.iaescore.com
Novel computational intelligence-based model for effective
traffic management in intelligent transportation system
Impana Appaji1
, Raviraj Pandian2
1
GSSS Institute of Engineering and Technology for Women, Visvesvaraya Technological University, Mysuru, India
2
Department of Computer Science and Engineering, GSSS Institute of Engineering and Technology for Women Mysuru, Visvesvaraya
Technological University, Belagavi, India
Article Info ABSTRACT
Article history:
Received Sep 1, 2023
Revised Oct 27, 2023
Accepted Dec 2, 2023
The evolution of intelligent transportation system (ITS) is essentially meant
for upgrading the driving experience with more safety and accessibility of
various analytical information from its extensive network. However, a
significant gap is observed that doesn't cater up the complete demands of
ITS. It has been also noted that computational intelligence (CI) based
approach is slowly gaining pace in solving the transport related problems in
ITS as compared to its other counterpart existing methodologies like
artificial intelligence. The proposed manuscript introduces a novel
computational framework towards assisting in relaying routing and
navigational services using CI-based approach. A design of novel
navigational controller unit is presented for global ITS scenario towards
yielding an optimal decision of routing. The CI-based approach is
implemented by integrating fuzzy process with evolutionary searching,
learning and probability theory in most simplified form. The study also
introduces a novel concept of relaying decision as feedback from
navigational controller unit to specific vehicular node discretely unlike
existing traffic controller system with an agenda to offer faster and effective
clearance of queued vehicular nodes from target area. The study outcome
shows higher consistency in its relay with better performance from existing
study model in ITS.
Keywords:
Computational intelligence
Fuzzy
Intelligent transportation
system
Probability
Vehicular network
This is an open access article under the CC BY-SA license.
Corresponding Author:
Impana Appaji
GSSS Institute of Engineering and Technology for Women, Visvesvaraya Technological University
Mysuru, India
Email: impana.appaji@gmail.com
1. INTRODUCTION
Intelligent transportation system (ITS) induces modernized technologies in order to improve the
process management as well as enhance the performance of transport system [1]. The prime reason of its
increasing pace of adoption is due to improved mobility, supportability towards economic development,
minimizing the environmental impact, and improving efficiency and safety [2]–[5]. However, it is also
characterized by various impending challenges which attracts the attention of researchers. Some of the
impending challenges of ITS are associated with technical complexity, data privacy and security,
interoperability, higher implementation cost, public acceptance, and lack of standardization [6]–[10]. At
present, there are various implementation study model being developed targeting to address the problems
associated with ITS [11]–[15]; however, they are still in much nascent stage of development and is yet to
witness its complete full-fledge implementation from practical perspective. The standard architecture for ITS
is classified into various layers viz. physical layer, communication layer, data layer, middleware layer, and

Int J Artif Intell ISSN: 2252-8938 
Novel computational intelligence-based model for effective traffic management in ... (Impana Appaji)
2525
application layer [16]. However, majority of the existing study models are considered only few of the
layer-based operation and more work is required to be carried out.
A closer look into majority of the trends of existing research work showcase that relaying of the
navigational services in ITS is one of the underlying operations associated towards facilitating the beneficial
features of ITS. However, such form of service relaying demands heavier traffic data followed by using
certain analytical scheme to perform processing of traffic data to yield better decision towards navigational
services. There are certain approaches where internet-of-things (IoT) [17], internet-of-vehicles (IoV) [18],
big data [19], and vehicular adhoc network (VANET) based schemes [20] has come together to offer such
services towards routing. It is also noted that adoption of intelligence-based scheme is on rise towards
solving complex research problem in vehicular network. Hence, the proposed study introduces a contributory
framework where an intelligent-based scheme is harnessed towards facilitating an effective ITS in
challenging traffic condition. The contribution of proposed study model are as follows:
− The proposed study model introduces an integrated computational framework offering seamless
connectivity of vehicular network.
− A navigational controller unit is introduced in proposed scheme which considers both local and global
traffic related information to undergo decision making towards traffic management.
− The study model introduces a concept of computational intelligence (CI) using simplified fuzzy logic
and probability theory modelling towards undertaking decision of routing in urban traffic system.
− The intelligent navigational controller unit is capable of determining the optimal channel for clearing the
queued vehicle in target area ensuring minimal waiting time in urban traffic scenario.
This section discusses about various available methodologies being implemented towards revising
or upgrading the performance in ITS identified different set of research problems associated with it.
Akhtar and Moridpour [21] have presented discussion of various predictive methods using artificial
intelligence (AI) towards solving congestion problem in traffic. According to the study, clustering techniques
are the most frequently adopted for this purpose mainly focusing on three essential methods i.e., deep
learning, shallow machine learning, and probabilistic reasoning. Adoption of AI-based methodology is also
witnessed in work carried out by Benterki et al. [22] that emphasizes towards analyzing the behavior of a
vehicle using trajectory prediction and classification methods based on maneuvering. The study model has
integrated long-short term memory (LSTM) along with neural network for this purpose. Further adoption of
AI is also reported in Lytras et al. [23] where it is stated that AI schemes are most suitable for catering up
dynamic demands of ITS along with using IoT-based services. Further, several researchers have presented a
discussion about predictive techniques using AI toward analysis of flow of traffic and realizing the influence
of autonomous vehicle on urban traffic [24], [25]. Tak et al. [26] have presented further discussion associated
with detecting and tracking vehicles in ITS using deep learning. Malek et al. [27] have used deep learning
scheme in order to perform prediction of speed of vehicular node. The concept used in LSTM considering
both multi/univariate environment for assessing the accuracy of predictive outcome.
A unique scheme towards solving problem of allocating transport vehicle in ITS is presented by
Chavhan et al. [28] using emergent intelligence scheme. This scheme processes various information related
to public transport and performs scheduling and allocation of depot’s agent for catering the dynamic need of
commuters. Study towards problem of public transport system is also discussed in [29]. The authors
attempted an investigation towards realizing the fair possibility of integration existing scheme for optimal
planning of public transport system. The outcome of the study has been directed towards the need of
optimization model in ITS, which is currently not witnessed in existing schemes. Eom and Kim [30] have
discussed about the traffic related problems in intersection specifically associating with traffic signal. The
notable outcomes of this study are; i) absence of any benchmark model for controlling traffic signal
performance, ii) congestion and queue management in intersection is computationally challenging task to
resolve. Study towards investigating the significance of traffic signal is also carried out by Qadri et al. [31].
Adoption of clustering towards addressing the repositioning issue of vehicle is presented in [32]. According
to this study, the process of multi-tasking during sequential learning operation can be significantly improved
by clustering process.
Adoption of CI towards managing wireless network using sensors are reported in [33]. The
identified problem of restoration of nodes is carried out using enhanced Lagrangian relaxation model along
with bioinspired algorithm using cognitive characteristic of grey wolf. There is less occurrence of pure CI-
based model in ITS; however, the subsidiary and enabling technologies in it has been found to be
implemented by various researcher considering vehicular network. Adoption of evolutionary computing as
CI-based scheme in ITS has been discussed by Chen et al. [34] where the study outcome highlights the
complexity still exists towards configure collaborative framework using edge and cloud. Husnain et al. [35]
have used developed an efficient data transmission scheme using bioinspired optimization method
considering the cognitive characteristic of whale for performing clustering in vehicular network. Jain et al.

 ISSN: 2252-8938
Int J Artif Intell, Vol. 13, No. 3, September 2024: 2524-2539
2526
[36] have integrated AI and bioinspired approach for developing a unique charging scheme for smart
vehicles. Further, fuzzy logic has been witnessed towards deploying CI-based ITS. Li et al. [37] have
developed a controller system for preventing skidding of electric vehicle using fuzzy set theory. Several
researchers have also highlighted research work using fuzzy approach towards decision making for traffic
system [38], [39]. Adoption of type-2 fuzzy approach is considered in work of Liu et al. [40] or controlling
the automated driving feature. Further, adoption of fuzzy process towards rail transport system is seen in
study model of Lu et al. [41] towards sorting out routing problem. Shi et al. [42] have developed a sliding
mode controller using fuzzy logic for adjusting the switching operation towards acceleration control in
electric vehicle. Simić et al. [43] have developed a model where fuzzy approach is used for overcoming the
traffic problem in urban in order to dispatch the goods in conventional transport system during emergency.
Adoption of fuzzy neural network is reported in work of Tang et al. [44] towards imputing the missing data
associated with traffic flow. Apart from this, adoption of neural-network based CI approach in ITS is
reported in [45]-[47]. Therefore, it is noted that there are various existing approaches of multiple variants
towards addressing the problems in ITS and evolve up with a new solution. The next section highlights the
associated problems with existing methodologies.
There are various beneficial aspects of existing research methodologies highlighted in prior section;
however, some critical shortcomings have been noted as follow:
− The consideration of problems associated with ITS is highly specific to application or form of services.
Hence, compatibility of such solution in global network of ITS is still not proven.
− The rise of AI and CI based solution is proven to be increasingly adopted by researchers; however, the
prime pitfall is associated with its adoption of highly iterative mechanism and less emphasis is offered to
overall ITS ecosystem.
− Majority of existing approaches are focused on adopting highly sophisticated solution in order to address
the specific problem. However, such solution comes at the cost of high computational burden that is yet
not validated.
− At present, there is no benchmarked study to prove any effective solution towards navigational-based
relay service in ITS when considered over global ITS scenario.
Therefore, the statement of the problem is “developing a computational model for global scenario of
ITS that can offer highly synchronized connectivity of vehicular network in order to facilitate dynamic
analytic operation with faster response time is still a challenging task”. The next section discusses the
methodology adopted to solve the issues.
2. PROPOSED SYETM
The prime purpose of this part of implementation is to deploy a CI in order to improve the data
dissemination performance in ITS services/application. The implementation of this scheme is carried out in
analytical research methodology (Figure 1). The core target of this part of implementation is also towards
accomplishing an optimization scheme which refers to minimize the dependencies from sophisticated
computational effort as well as varied resources in order to accomplish higher degree of scalable performance
in ITS.
Prior to briefing the architecture, it is essential to understand the problem formulation of proposed
study. It should be noted that prime agenda of proposed scheme is to utilize the cost-effective CI scheme for
effective navigational support towards ITS. One simpler way to do this to when the navigational controller
within the vehicle is facilitated with a discrete value of traffic command signal. However, generating traffic
command signal discretely for each set of waiting vehicles on different channels demands more information
from the global perspective. Therefore, mathematically this problem can be formulated as (1):
𝑎𝑟𝑔𝑚𝑎𝑥[𝑑𝑖𝑠(𝑡𝑐)] → 𝑣(𝑐𝑐)𝑡𝑎 (1)
In (1) of problem formulation states that it is challenging task to maximize argmax the discrete dis facilitation
of traffic command tc to each vehicle v present in current channel cc for a specific target area ta. Conventional
ITS system generates common tc for all v on specific ta without assuming the dynamic demands of direction
of journey for each vehicle. Even if this formulated problem is somehow solved, the next set of challenge
will be to ensure seamless connection between two nodes, it could be vehicle-to-vehicle (V2V) or
vehicle-to-infrastructure (V2I) or vehicle-to-anything (V2X). Without any infrastructure and considering
dynamic mobility of a vehicle, the second problem in ITS for urban use-case can be formulated as (2):
𝑎𝑟𝑔𝑚𝑎𝑥(𝑐𝑣) → [𝑣1, 𝑣2|∀𝑡𝑐𝑐 ] (2)

2527
In (2), it states that it is computational higher and iterative effort task for increasing the strength and quality
of communication vector cv between two communicating nodes v1 and v2 in presence of obtained traffic
response signal tcc. The significance of this problems is an effective tcc can be only relayed of the
navigational controller unit possess information about the same for its surrounding, which is unfortunately
not there is existing ITS scheme. Hence combining mathematical in (1) and (2), the final problem
formulation is stated as (3):
𝑡𝑐𝑐 ⇒ max [𝑒𝑥𝑖𝑡(𝑣(𝑐𝑐))] (3)
The inference in (3) will mean that final problem formulation of proposed system arrived is to
balance the relayed traffic clearance command tcc should offer maximum clearance of vehicles v in current
channel over its exit points. The exit points may lead to final destination or it may eventually connect to
another channel. The proposed system initiates by constructing a distributed ITS topology considering three
types of inputs i.e.; i) location information of vehicle, ii) direction of vehicle, and iii) velocity of vehicle. All
this information were used further for constructing an underlying graph to offer a better control of the
distributed network system in ITS. The next process is towards exploring an effective position of the vehicle
followed by effective path. The idea is to offer a faster clearance of the vehicles on any point in traffic system
that is connected by an underlying graph for information dissemination. The study considers the presence of
one hot spots in every lane segment in such a way that its transmission region encapsulates all the incoming
and outgoing vehicles within that lane segment. The study also considers that all the hot spots are
synchronized with each other connecting themselves with the distributed ITS topology which is hosted by
cloud environment in collaborative manner. On the basis of the input obtained from vehicles, a proposed
scheme also performs its decision using an objective function with CI. The outcome of the decision is then
forwarded to the vehicles in the form of navigational vectors to offer faster clearance of each lane system.
The objective function using CI is designed in such a way that it should offer consistency in its
decisive outcome irrespective of any situation of traffic. Following are further process carried out towards
accomplishing an effective optimization performance:
− A distance-based computation is carried out for assessing the effective position and path of vehicles.
− A many-value logic system is introduced in the form of a CI unit that formulates the objective function
for determining the optimal probability of road to be selected.
− The model emphasizes mainly towards minimizing the waiting period with faster clearance especially
considering the use-case of intersection point where the probability of jam is either always higher or
uncertain in ITS.
− All the information is forwarded to both requestor node as well non-requestor node in order to ensure
better transparency in information propagation in ITS. This section further presents discussion of the
strategy adopted in order to implement the proposed study model.
Figure 1. Architecture of proposed scheme
Location
Information
Direction of vehicle
Velocity of vehicle
Construct a distributed
ITS Topology
Formulate Graph
Find effective
position
Find Effective path
Construct Objective function with computational
Intelligence
Construct Integrated ITS network
Select requestor node
Select non-requestor
node
Optimize Service Relaying

 ISSN: 2252-8938
2528
2.1. Exploring optimal channel
This is the first part of implementation which targets to find solution of mathematical representation
of problem (1) mainly. It also indirectly solves the issue for problem formulated as (3). The problem solution
is presented in the form of optimal channel exploration process. The primary strategy adopted in proposed
study implementation is to explore the best possible channel. The term channels will refer to set of lanes
present in considered urban vehicular network in ITS. With reference to Figure 2, this is the first
implementation module where all the vehicular nodes are considered followed by a construction of a
temporary matrix that retains information associated with current channel and traffic commands. The term
current channel will refer to current lane segment which is under observation for analysis while the term
traffic command relates to decision carried out by the centralized navigational system present in target area.
Figure 2. Mechanism for exploring optimal channel
The study considers that such controller takes the input of existing condition of traffic for current
channel as well as other connected channel in global urban traffic system, performs analysis and relays the
information about its decision making. This information is specific to vehicles queued in current channel.
The system further constructs a target area which is basically a region where all the other neighboring
channels integrate with each other. The study considers that all the decision of the navigational controller are
basically relayed specifically in the target area where the vehicles are queued and awaits further input from
the navigational controller to move ahead. The difference between proposed controller and existing controller
scheme of urban traffic is that: proposed controller scheme performs decision of clearing the vehicle based on
input from all the channels whereas existing controller system (or traffic light) yields solution only based on
current channel. Another set of difference is that proposed controller offers unique directional information to
vehicular node which may differ from another vehicular node even if they are on same channel. However,
existing controller offers similar input of directionality of mobility of all the vehicular nodes uniformly
without considering the global circumstances. Hence, proposed methodology targets for more enlarged
scenario of urban traffic while existing traffic system emphasizes only on current segment of lane.
Upon constructing a target area, the proposed scheme assesses a sequential set of conditional logic
to find out if the target area differs from certain constant. Further, the proposed scheme implements fuzzy
ruleset as a part of CI where further evolutionary search as well as probability logic is applied towards
assessing the problem space. The method further considers the cardinality of the traffic commands which can
vary from one to another channel followed by assessing another conditional logic. This is meant for
comparing the applicability of traffic commands with respect to CI ruleset. The term traffic command will
mean a set of certain commands which are usually predefined set of instructions to be relayed by the
navigational controller at the target area to specific vehicles for clearing the queue in current channel.
Start
Consider all
vehicular nodes
Construct a
temporary matrix
Construct Target
Area
cond1
Construct CI
ruleset
Consider
cardinality of
traffic commands
cond2
Allocate specific
Adjacent Channel
Formulate matrix
of optimal search
Stop
Traffic
commands
Current
channel
v
Ui
cc tc
ta
P
L
Ac
ψf
ta<const?
L>P?
T
T
F
F

2529
Further, the scheme allocates specific adjacent channel from the target area as the adopted options to clear the
queued vehicles. The outcome of this optimal search is retained in specific matrix for further utilization in
next stage of operation.
2.2. Relaying final traffic command
This part of implementation is associated with solving the formulated problem exhibited in (2) and
(3). According to both this equation, the prime solution is furnished by the usage of non-requestor node to
deal with likelihood of intermittent connection between the vehicular nodes. Not only this, but it also reduces
the effort of retransmission of any request from the vehicle towards specific routes. The problem solution is
also presented by formulating a temporary memory as well as matrix of optimal search from prior problem
solution, which offers all the alternative possibilities of routes. The mechanism for final relay of traffic
command is illustrated in Figure 3.
Figure 3. Mechanism for final relay of traffic command
The information within the optimal search is subjected to dynamic change when any specific traffic
situation on specific channel is encountered. Upon obtaining the outcome of optimal search, this part of
implementation results in further confirming the relaying of decision towards clearing the vehicles from
target area. The entire operation is carried out considering two forms of nodes i.e., requestor node and
non-requestor node. The requestor node is vehicular node which is seeking input from navigational controller
for direction towards tentative adjacent channel. The non-requestor node is vehicular node whose assigned
adjacent channel doesn’t reside in current target area and they play the role of offering information about
direction and node density for the navigational controller unit in urban traffic system. Using all the vehicular
nodes and outcome of optimal search, this part of study model implementation formulates three discrete
conditions considering both requestor and non-requestor node. A matrix for cumulative decision is formed
using clearance response signal and traffic command from the navigational unit. The term clearance response
signal is basically an analyzed outcome from navigational unit on the basis of various environment
parameters of vehicles and channel. It should be noted that traffic commands just retain a predefined set of
instruction to be relayed. The system further constructs a temporary buffer for target area followed by further
conditional assessment considering residual observation in order to compute waiting period of signal. This
upon updating results in finally relayed traffic information which is followed by the queued vehicles by
adopting the generated adjacent channels. Further elaboration is carried out in next section in form of
algorithms.
Start
Consider all
vehicular nodes
Consider Optimal
Search Outcome
Cond3
Construct matrix
of cumulative
decision
Construct
temmporary
buffer
Cond4
Cond5
Assign subsidiary
matrix to traffic
command
Compute waiting
period of signal
Obtain relayed
traffic information
Stop
Traffic command
Clearance
Response Signal
Residual
Observation
Target Area
alloc
update
v ψf
tc
tcc
σ
buff(ta)
Ψf(k)≠Ψf(ta-1)
k≠2
ta
k-1
α(γ)
tc(ψf) ß [h1 h2]
tc
T
T
F
F

 ISSN: 2252-8938
2530
3. ALGORITHM IMPLEMENTATION
This section presents discussion about the system design in the form of an algorithm that is
implemented towards accomplishing the defined objective for managing traffic system. The core objective of
the implemented algorithm is to facilitate a better form of decision-making using CI in vehicular network.
The complete algorithm implementation is carried out considering the common form of practical
environment on every urban traffic. The prime task of the algorithm is towards selection of an optimal
channel followed by reducing the bottleneck possibility in uncertain and dense vehicular nodes. The
discussion of the algorithmic steps are as follows:
Algorithm for exploring optimal channel
Input: v, cc, tc
Output: ψf
Start
1. For i=1:v
2. init Ui=[cc, tc]
3. ta=f(Ui)
4. For cond1=True
5. P=f1(ta, tuc)
6. For j=1:L
7. L(j)=argmax(L)
8. If cond2
9. ψfßAc
10. End
11. End
12. End
13. End
End
The discussion of the algorithm is as following: the proposed system takes the input of v (vehicular
nodes), cc (current channel), and tc (traffic command) which after processing yields an outcome of ψf
(selected channel). The implementation of the proposed algorithm is essentially meant for all the vehicular
nodes v and hence its algorithmic processing steps is required to consider it first (Line-1). The algorithm then
formulates a temporary matrix Ui which retains information about initialized value of current channel cc and
traffic command tc (Line-2). In the subsequent step of algorithm, an explicit function f(x) is constructed
which takes the input argument of newly generated temporary matrix Ui (Line-3) in order to generate a target
area ta. It should be noted that complete algorithm works on an evaluation based on the input obtained by the
system and hence there is also a possibility of lag in arriving to this outcome of computation of ta due to
nature of wireless network and uncertain mobility of the vehicular node v. However, the variable current
channel as well as variable traffic command is considered to be static for certain cut-off duration of time on a
specific channel and this is the reason that adoption of these two variables (cc, tc) will not have much impact
of dynamicity on the current computation of ta. Further, it should be noted that there are possibilities of
multiple number of target area ta at the end of each channel; therefore, the proposed scheme simplifies the
computation by considering an arbitrary number of vehicular nodes passing through specific target area ta.
The consecutive step of the algorithm considers assessing a primary conditional logic cond1 which evaluates
if the number of target area ta is less than certain reference value of it (Line-4). The reference value is density
of vehicular node present in specific channel approaching the target area ta. For an example, consider the nd is
node density in channel-1 approaching target area, therefore, according to this condition cond1, the evaluation
carried out is as (4):
ta > constant. nd (4)
It is quite obvious that number of ta on specific channel should be always less than nd; however, it
could possibly fluctuate between quite lower and saturated value of nd. In such circumstances, the target area
is either required to offer lower priority to its nd in its respective channel (ta>>nd) or else it is required to offer
higher priority (ta<nd). The value of constant is residing between [0 1] of probabilistic model. If the number
of target area ta is found to be less than reference value than it signifies the state of bottleneck situation which
is sorted out by introducing another explicit function f1(x) (Line-5). It will also mean that if conditional logic
cond1 is found to be true (Line-4) than the traffic system is required to prioritize the specific channel
connected with the target area ta. This process of clearing is carried out by transmitting specific form of
traffic command tc by further updating it to obtained updated traffic command tuc to specific ta (Line-5). The
prime role of the newly formulated function f1(x) is to perform a computation of specific decision controlling
the clearance of the vehicular nodes queued in target area ta. This mechanism of clearance of the vehicular
nodes is carried by considering the updated information of traffic command tuc from other channels too to

2531
formulate better global decision making. This process is carried out by setting up fuzzy rules using f1(x)
which are stored in matrix P (Line-5). Further, the proposed scheme considers the variable L as the adjacent
channel connected to the target area ta as the candidate solution (Line-6). Hence, better decision making is
only possible when the proposed scheme considers the maximum arguments of adjacent channels L (Line-7).
Further, the proposed scheme constructs secondary conditional logic cond2 where it assesses if the
value of adjacent channels L is more than that of fuzzy ruleset P (Line-8). This condition will mean that
formulated fuzzy ruleset P is successful in confirming the best channel out of all the available adjacent
channel L from its specific target area ta. It should be further noted that the variable L consists of various set
of traffic commands tc that is responsible for controlling the clearance of the specific vehicles queued in
specific channel. All this information is stored in a buffer of adjacent channel i.e., Ac which is that allocated
to final outcome of selected channel ψf (Line-9). Therefore, this algorithm represents a most simplified
mechanism of deploying CI in order to confirm a decision of optimal search of channel in urban traffic
system.
The next part of the algorithm implementation is the continuation of the prior algorithm that
confirms the presence of optimal channel. This algorithm is more emphasized towards offering more
reliability in case of dynamic event of present in an urban traffic system. A closer look at the prior
implemented algorithm will show that it is focused on accomplishing local optimality considering current
channel while the second algorithm is now focused on accomplishing equivalent state of optimality for global
channel system. The idea is to offer better sustainability as well as consistency in processing of the
algorithms as well as ensuring a formation of a well-connected and highly organized network of urban traffic
system. The algorithmic steps are as follows:
Algorithm for final relayed traffic command
Input: v (vehicular nodes), ψf (Selected channel)
Output: tc (relayed traffic command)
Start
1. For i=1: v
2. For j=1: ψf
3. If c ond3=True
4. Construct σ=[tc, tcc, ψf]
5. σ(i, j)ßbuf(ta)
6. If cond4=True
7. If Cond5=True
8. tc(ψf)ß[h1 h2]
9. compute α(γ)
10. tc(ψf)ß [h1 h2]
11. End
12. End
13. End
14. End
15. End
End
The discussion of the algorithm is as follows: the proposed algorithm takes the input values of v
(vehicular nodes) and ψf (Selected channel), which after processing yields an outcome of tc (relayed traffic
command). In order to keep the updated information of any form of an incoming vehicle on specific channel,
it is essential for this algorithm to consider vehicular nodes v as the primary input (Line-1). Further, the
outcome obtained from prior algorithm i.e., selected channel ψf is further considered to narrow down the
search towards the optimality state of navigation for each vehicular node (Line-2). For simplification of the
computation in proposed scheme, the algorithm is required to initialize practical values of target area ta. The
range of values that can be allocated for ta is 2>ta>4 considering the normal scenario of any urban traffic
system. In the subsequent steps of operation, the algorithm performs a conditional check cond3 where the
system assesses if the value of clearance response signal tcc is more than the reference score of clearance λ.
This condition will practically imply the higher possibility of clearing a greater number of queued vehicles in
target area ta. Upon confirming the true case of this condition cond3 (Line-3), the proposed algorithm
constructs a matrix σ which consists of prior traffic command tc, clearance response signal tcc, and selected
channel ψf. (Line-4). The prime reason for undertaking this step of algorithm are as follows: i) adoption of
prior traffic command tc assists in evaluating the success rate of prior response from signal for specific target
area ta, ii) adoption of clearance response signal tcc assists in clarifying the total number of vehicles being
permitted to exit from the queued channel, while iii) adoption of selected channel ψf. assists in offering a
candidate channel to be considered for cleared vehicles to move. This operation is carried out by constructing
a temporary buffer buff considering the input argument of target area ta that is further assigned to matrix of
cumulative decision i.e., σ (Line-5). This step of operation will eventually mean that the proposed search area
towards optimal solution is retained in a temporary buffer where further decision is carried out in faster and

 ISSN: 2252-8938
2532
more reliable way. In the subsequent process, the proposed algorithm constructs another conditional logic
cond4 which assesses the degree of equivalency for; i) selected channel ψf with respect to k-number of
observations of channel and, ii) selected channel ψf. with respect to other target area ta (Line-6). If the
assessment doesn’t find any form of equivalency than it further formulates another conditional logic cond5 to
checks k≠2 which will mean that number of observations for individual channel should be greater than
minimum 2 adjacent channel atleast for acquiring more information of residual target area ta and channels
associated with it for better decision making (Line-7). If the above-mentioned condition cond5 is found to be
true than the algorithm constructs two subsidiary matrices of selected channel as h1 and h2 and allocated them
to matrix of traffic command tc with respect to prior selected channel ψf for upgrading (Line-8). According to
this concept, the proposed scheme configures the score of tc corresponding to (k-1) as h1 while the score of
the tc corresponding to k round is assigned h2. For simplification, the proposed scheme considers h1=0.5 and
h2=0. Further, the algorithm computes duration of queue in current traffic γ with respect to the clock time α
associated with each vehicular node (Line-9). Further, the update operation is carried out for relaying the
traffic command tc on the basis of current node density nd (Line-10). The difference in Line-8 and Line-10 is
that i) execution of Line-8 calls for constructing a traffic command on the basis of k-observation and target
area ta while ii) execution of Line-10 calls for only current and other target area ta. This completes the overall
algorithmic processing where the final outcome of relayed traffic command is obtained in last step (Line-10).
A closer look at the contribution of both the connected algorithms are as follows: i) the algorithm
has less iteration and more inclusion of progressive steps resulting in faster execution with reduced
computational burden, ii) both the algorithm tends to perform computation using updated traffic information
without saturating the existing memory owing to usage of more temporary buffers, whose values are disposed
off once utilized. The next section discusses about the accomplished outcomes of above algorithm
implementation.
4. RESULTS AND DISCUSSION
From the prior section, it is noted that proposed scheme is mainly associated with implementing a
novel and simplified CI-based framework in order to perform optimal traffic management. Hence, the
evaluation is carried out by developing an initial ITS simulation framework with presence of various
attributes e.g., vehicular nodes, hotspots, navigational controller unit, channels and its connected adjacent
channels, traffic command related from navigational controller unit, and target area. The adopted simulation
parameters are highlighted in Table 1.
Table 1. Simulation environment
Parameter Value
Total vehicular node 800
Capacity of one channel 50
Throughput 300 Mbps
Vehicle wireless standard IEEE 802.11
Number of adjoining channels in target area 4
Total iteration 1000
Simulation Area 1200 x 1200 m2
Traffic type Urban
Different from existing assessment of vehicular network using map-based approaches in urban
traffic system, the framework of the proposed scheme performs on-the spot computation in respective target
area, uses computation intelligence followed by generation of traffic command for clearing the vehicles in
target area by navigational controller unit. Multiple unconventional performance metric has been used owing
to the novelty of the proposed study model. The core idea of this assessment is to understand the
effectiveness of relaying traffic command for clearing maximum number of vehicles in channels.
The simulation visuals are shown in Figure 4 for one specific case study towards single target area
which is represented by one intersection point of a road in ITS. The simulation outcomes show number of
vehicles for each specific lanes that are either forwarded with waiting signal or forwarding a clearance signal.
The unique observation is that irrespective of the presence of many numbers of vehicles, the navigational
controller selects specific and discrete number of vehicles in each channel to be cleared unlike existing traffic
management where all the vehicles are cleared. Hence, this is a significant novelty introduced by proposed
scheme which identifies the demands of vehicles on the basis of waiting time and then undertakes the
decision of either letting them to wait or to move.

2533
Figure 5 showcase the original simulation outcome with respect to response consistency in
Figure 5(a), time for feedback relaying in Figure 5(b), duration of traffic command in Figure 5(c), and
counter of clearance and number of vehicles cleared in Figure 5(d). From the direct outcome, it can be stated
that all the performance parameters are to exhibit a better uniformity with respect to test case of 4 channels
under assessment environment. Although, the outcome has been obtained instantaneously for a single target
area, it is observed that irrespective of any time of vehicle journey in dynamic scenario of ITS, the
navigational controller always maintains similar form of consistency in its traffic management. The next
section further elaborates these 4 performance parameters and offers collaborate inference of results obtained
in global environment.
Figure 4. Visuals of Simulation in one single intersection of road (target area)
(a) (b)
(c) (d)
Figure 5. Graphical results of simulation for (a) response consistency, (b) time for feedback relaying,
(c) duration of traffic command and, (d) counter of clearance and number of vehicles cleared

 ISSN: 2252-8938
2534
4.1. Assessment of response consistency
The term response consistency will mean the proportion of the vehicular nodes being managed by 4
test channels under observation without exhibiting fluctuation inspite of variable navigational demands
requested by each vehicular node. The observation is carried out on 4 sets of iterations (iteration-1: 250,
iteration-2:500, iteration-3:750, and iteration-4:1000). The rationale behind adopting this performance metric
is to understand the capability of handling vehicular clearance system by navigational controller unit on
respective channels. Higher fluctuating over increased iteration will only mean less coordinated navigational
unit while less fluctuating number of vehicles on each channel will represent highly synchronized
navigational controller unit.
The outcome exhibited in Figure 6 showcase that channel-1 with a less node density maintains the
range of 10-11 vehicular nodes over increasing iteration. Similarly, channel-2, channel-3, and channel -4 are
found to maintain a range of 10-14, 12-13, and 10-14 vehicular nodes respectively over increased iteration.
The overall difference in proportion of fluctuation in vehicular nodes are only 1.5%. Therefore, the graphical
outcome states that proposed scheme can offer better consistency of response of traffic command to different
vehicular nodes residing on different channels with different node density. Hence, better exhibit of system
performance consistency can be observed.
Figure 6. Assessment response consistency
4.2. Assessment of time for feedback relaying
The performance parameter time for feedback relaying represents overall time required by the
navigational controller unit to relay the decision of navigational feedback (requested by the vehicular node)
on each individual channels under observation. The reason for adopting this performance metric is to
evaluate the difference of time consumed for relaying computed traffic command to different requestor node
in different channels with varied traffic characteristic. The system can be called as efficient only if; i) there is
lesser time consumption for relaying the feedback of traffic command, and ii) there are nearly equivalent
trend of these time attributed over all the channels in ITS.
The outcome in Figure 7 shows that time spent for relayed the feedback for channel-1, channel-2,
channel-3, and channel-4 are in range of 7-9 seconds, 6-8 seconds, 6-8 seconds, and 7-8 seconds respectively.
This outcome offers two inferences viz; i) the mean difference in time required for feedback relaying for all
the channels under observation is approximately 1.75%, which represent a satisfactory consistency in
relaying performance, and ii) overall 2 seconds of maximum difference is found for increased iteration for a
single channel. This outcome states that proposed CI-based navigational controller unit can judge equally all
the request from varied sources and yet consumed less processing time to relay the actual feedback to
different vehicles in ITS. This outcome is also in supportability of higher synchronization of varied services
in ITS with faster feedback.
4.3. Assessment of duration of traffic command
The term duration of traffic command represents the duration of activeness for relayed traffic
command towards requested vehicles on each channel under observation. The conventional traffic commands
are only three viz. red light, green light, and yellow light; however, it is quite inevitable that futuristic

2535
ITS-based services will call for introducing multiple form of pre-defined traffic commands which could be
either more or less than conventional traffic commands. Similar to conventional traffic command, the
proposed scheme also offers a standard time of activeness of relayed traffic command. Upon expiration of
this current traffic command duration, the next computed traffic command takes the place of prior one to be
relayed. However, this duration of traffic command cannot be same for all the channel at single point of time.
Owing to variable traffic condition, such duration will also vary. Hence, the worst possibility with
sub-optimal outcome will arise if there is heavy fluctuating duration of activeness of traffic command in
different channels at same time. While better optimal solution can be ensured if all the channels experience
nearly same duration of activeness of this traffic command. Hence, the motive of this performance metric is
to assess the variation observed in duration of traffic command in each channel under observation.
From the outcome in Figure 8, it is noted that channel-1 witness’s duration of 0.7-1 seconds, while
Channel-2 witnesses a duration f 1.6-1.8 seconds. Further, channel-3 and channel-4 witnesses a duration of
1.5-1.9 seconds and 1.5-1.8 seconds. A closer look into all the trends shows that overall variation among all
the 4 channels is approximately 4% while the minimum and maximum duration observed is between 0.7-1.8
seconds. This score justifies a very reduced duration of one active traffic command, which also infers
following: i) proposed scheme always maintain a significantly variable duration of active traffic command in
each channels but not exceeding more than 4 seconds even when the iteration is increased to maximum, and
ii) there is a faster switching mechanism for all the traffic commands exhibited by this lower duration score
representing proposed scheme does equivalent justification for non-requestor node too along with requestor
node in priority. Hence, it will eventually mean that proposed scheme processes the query of requestor by
navigational controller unit very discretely without affecting the interest of navigation for non-requestor node
in specific channel.
Figure 7. Assessment of time for feedback relaying
Figure 8. Assessment of duration of traffic command

 ISSN: 2252-8938
2536
4.4. Assessment of duration of traffic command
The term vehicle clearance system as a performance metric represent two attributes mainly i.e.:
i) counter for clearance, and ii) number of vehicles being cleared. The first attribute counter of clearance will
represent total number of clearances as a traffic command exhibited in one set of iteration (i.e., 250 rounds)
by the navigational controller unit at same point of time. The second attribute number of vehicle will
represent total number of requestor node being cleared upon receiving the clearance feedback from
navigational controller unit out. The core idea of this performance attribute is to assess the effectiveness of
clearing the queued vehicles from each target area. The assessment of vehicle clearance system is illustrated
in Figure 9.
Figures 9(a) and 9(b) represents the outcome of counter for clearance and number of vehicles being
cleared. The outcome shows that channel-1, channel-2, channel-3, and channel-4 has witnessed 36, 44, 34,
and 39 number of traffic commands for clearing the queued vehicles at same point of time (Figure 9(a)).
Further, at the end of observation round, it is noted that channel-1, channel-2, channel-3, and channel-4 has
cleared 440, 700, 500, and 720 vehicles successfully. This part of analysis was carried out with small number
of vehicles first (v=100) , which was later exponentially increased by increasing the capacity of channels by
introducing a greater number of connecting channels. The idea is to assess the performance of the overall
system when exposed to a large channel of networks in ITS. The outcome exhibits that proposed navigational
controller unit is quite efficient in processing the query and give final feedback for clearance without much
fluctuation representing a good synchronous links among all navigational controller units. Hence, one the
basis of this outcome, it can be stated that proposed scheme is capable of identifying and realizing the traffic
condition for large scenario, perform distributed computation, and generates the feedback of traffic command
on the basis of node density, and number of queued vehicles.
(a) (b)
Figure 9. Assessment of vehicle clearance system for (a) counter for clearance and (b) number of vehicles
cleared
Apart from the discussed study outcomes, the outcome of the proposed scheme is further compared
with the state-of-the art methods. Table 2 highlights the outcomes which states that proposed scheme
eventually has a beneficial aspect in contrast to different variants of approaches towards improving ITS. One
of the important aspects to understand here is that almost all the existing approaches have considered various
set of problems associated with ITS where the solution obtained are claimed to be optimal. However, when
the same existing framework is considered for global ITS scenario, there are highlights of various pitfalls.
The prime reason is adoption of methodology and structuring them towards addressing the specific set of
problem which doesn’t include complete ITS scenario. It is to be noted that a complete ITS scenario must
involve seamless connectivity and coverage with all the involved entities.
Therefore, from Table 2, it can be noticed that existing AI-based schemes are mainly associated with
higher computational complexity due to their highly iterative operation as well as higher dependencies to
resources. Hence, such AI-based existing methodologies may suit well for non-emergency applications in
ITS but not for dynamic and tie-critical application in ITS. The existing analytical learning model has better
benefits compared to AI till now as they can be successfully proven to be used in scheduling and route
planning which is one essential requirement in ITS in order to cater up many applications in ITS. However,
they are too specific of services and cannot be deployed on overall ITS scenario. The CI-based scheme is still
in starting level of implementation. Although some positive outcomes have been recorded using evolutionary

2537
approach, fuzzy approach, and neural network approach, but its shares nearly the similar pitfalls like that of
existing approaches using AI. On the other hand, proposed scheme introduces a very simplified,
non-iterative, and highly progressive computational framework which considers complete ITS scenario. It is
capable of performing analytical operation on the basis of all connected navigational controller unit. Further,
it is capable of relaying the unique computed feedback outcomes for clearing the vehicles uniquely unlike
any conventional traffic model. Apart from this, there is a higher degree of consistency of outcomes observed
in proposed model when exposed to low to peak traffic condition. Apart from this, it is to be noted that an
effective traffic management can be also carried out by using artificial based approach which can facilitate
power of CI to extract all hidden traits of the problem. Appaji and Raviraj,[48] have used deep neural
network for facilitating better form of vehicle communication system and it’s proven to offer low resource
consumption, low computational burden, and higher overall efficiency. Hence, the proposed model can be a
complimentary optimization scheme further on our prior model towards better traffic management.
Table 2. Comparison with state of art methods
Approaches Computational
complexity
Benefits Limitation
AI
[21]–[23], [26], [27]
High Higher accuracy − Solution not applicable for global ITS
− Highly Iterative
− Accuracy depends on trained dataset
− Not resource efficient
Analytical learning method
[28]–[32]
Medium-high ‒ Effective traffic scheduling
‒ Effective planning of routes
‒ Simplified clustering
− Determination of optimal solution
still not applicable for complete ITS
− Service specific solution
CI
Evolutionary- [34]–[36]
Fuzzy- [37], [41]–[44]
Neural network- [45]–[47]
High ‒ Capable of solving multiple problem
‒ Effective in modelling higher
dimension of transport system
‒ Can be used for both hardware and
software-based approach
‒ Capable of modelling complex problem
− Non-considering of complete ITS
ecosystem
− Sub-optimal outcome due to lack of
constraint consideration
Proposed Low-medium ‒ Considers complete ITS scenario
‒ Solves routing and analytical problems
‒ Maintain consistency in output in
variable traffic condition
‒ Novel feedback relay system specific to
vehicles
Applicable only on urban traffic system
5. CONCLUSION
The current study model has introduced a novel and simplified computational framework which is
capable of offering a faster and reliable navigational and routing services in ITS. Following are the novelty of
the proposed scheme: i) the proposed scheme introduce a navigational controller unit which offers decision
feedback of routing specifically for a discrete requestor vehicular node, ii) the complete analysis of the
decision feedback of routing is carried out by considering the global scenario of channels in ITS unlike
existing scheme which only considers the local routes, iii) the complete process is designed using CI-based
scheme which doesn’t include excessive iterative operation and is highly progressive in its operation, iv) the
most significant novelty of this model is that it is capable of offering the different decision feedback of
routing to different vehicular nodes maintaining higher level of consistency for all the channels. The future
work can be carried out further to optimize the search towards bottleneck points in traffic channels.
REFERENCES
[1] G. Dimitrakopoulos, L. Uden, and I. Varlamis, “Intelligent transport systems and smart mobility,” in The Future of Intelligent
Transport Systems, Elsevier, 2020, pp. 199–205. doi: 10.1016/b978-0-12-818281-9.00018-8.
[2] C. Bıyık and T. Yigitcanlar, “Intelligent transport systems in Turkish urban environments: a comprehensive review,”
International Journal of Knowledge-Based Development, vol. 11, no. 4, 2020, doi: 10.1504/ijkbd.2020.10037852.
[3] S. C. Rajkumar and L. J. Deborah, “An improved public transportation system for effective usage of vehicles in intelligent
transportation system,” International Journal of Communication Systems, vol. 34, no. 13, 2021, doi: 10.1002/dac.4910.
[4] F. Alonso, M. Faus, M. T. Tormo, and S. A. Useche, “Could technology and intelligent transport systems help improve mobility
in an emerging country? Challenges, opportunities, gaps and other evidence from the caribbean,” Applied Sciences, vol. 12, no. 9,
2022, doi: 10.3390/app12094759.
[5] Z. Lv and W. Shang, “Impacts of intelligent transportation systems on energy conservation and emission reduction of transport
systems: A comprehensive review,” Green Technologies and Sustainability, vol. 1, no. 1, 2023, doi: 10.1016/j.grets.2022.100002.
[6] K. Shaaban, M. Elamin, and M. Alsoub, “Intelligent transportation systems in a developing country: benefits and challenges of
implementation,” Transportation Research Procedia, vol. 55, pp. 1373–1380, 2021, doi: 10.1016/j.trpro.2021.07.122.

 ISSN: 2252-8938
2538
[7] J. Kołodziej, C. Hopmann, G. Coppa, D. Grzonka, and A. Widłak, “Intelligent transportation systems – models, challenges,
security aspects,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture
Notes in Bioinformatics), Springer International Publishing, 2022, pp. 56–82. doi: 10.1007/978-3-031-04036-8_3.
[8] Q. E. Ali, N. Ahmad, A. H. Malik, G. Ali, and W. U. Rehman, “Issues, challenges, and research opportunities in intelligent
transport system for security and privacy,” Applied Sciences, vol. 8, no. 10, 2018, doi: 10.3390/app8101964.
[9] V. Sharma, L. Kumar, and S. Sergeyev, “Recent developments and challenges in intelligent transportation systems (ITS)—a survey,”
in Intelligent Computing and Communication Systems, Springer Singapore, 2021, pp. 37–44. doi: 10.1007/978-981-16-1295-4_4.
[10] E. J. Tomaszewska, “Barriers related to the implementation of intelligent transport systems in cities-the polish local government’s
perspective,” Engineering Management in Production and Services, vol. 13, no. 4, pp. 131–147, 2021, doi: 10.2478/emj-2021-0036.
[11] I. Damaj, S. K. Al Khatib, T. Naous, W. Lawand, Z. Z. Abdelrazzak, and H. T. Mouftah, “Intelligent transportation systems: a
survey on modern hardware devices for the era of machine learning,” Journal of King Saud University - Computer and
Information Sciences, vol. 34, no. 8, pp. 5921–5942, Sep. 2022, doi: 10.1016/j.jksuci.2021.07.020.
[12] S. Galea, D. Seychell, and M. Bugeja, “A survey of intelligent transportation systems based modern object detectors under night-
time conditions,” in Proceedings of the 3rd International Conference on Intelligent Sustainable Systems, ICISS 2020, IEEE, 2020,
pp. 265–270. doi: 10.1109/ICISS49785.2020.9316076.
[13] Zulkarnain and T. D. Putri, “Intelligent transportation systems (ITS): a systematic review using a natural language processing
(NLP) approach,” Heliyon, vol. 7, no. 12, 2021, doi: 10.1016/j.heliyon.2021.e08615.
[14] J. N. Njoku, C. I. Nwakanma, G. C. Amaizu, and D. S. Kim, “Prospects and challenges of metaverse application in data-driven
intelligent transportation systems,” IET Intelligent Transport Systems, vol. 17, no. 1, pp. 1–21, 2023, doi: 10.1049/itr2.12252.
[15] T. Hassan, A. El-Mowafy, and K. Wang, “A review of system integration and current integrity monitoring methods for positioning in
intelligent transport systems,” IET Intelligent Transport Systems, vol. 15, no. 1, pp. 43–60, Nov. 2021, doi: 10.1049/itr2.12003.
[16] K. Gao, S. Huang, J. Xie, N. N. Xiong, and R. Du, “A review of research on intersection control based on connected vehicles and
data-driven intelligent approaches,” Electronics, vol. 9, no. 6, 2020, doi: 10.3390/electronics9060885.
[17] E. B. Priyanka, C. Maheswari, and S. Thangavel, “A smart-integrated IoT module for intelligent transportation in oil industry,”
International Journal of Numerical Modelling: Electronic Networks, Devices and Fields, vol. 34, no. 3, 2021, doi: 10.1002/jnm.2731.
[18] N. M. Elfatih et al., “Internet of vehicle’s resource management in 5G networks using AI technologies: current status and trends,”
IET Communications, vol. 16, no. 5, pp. 400–420, 2022, doi: 10.1049/cmu2.12315.
[19] L. Zhu, F. R. Yu, Y. Wang, B. Ning, and T. Tang, “Big data analytics in intelligent transportation systems: a survey,” IEEE
Transactions on Intelligent Transportation Systems, vol. 20, no. 1, pp. 383–398, 2019, doi: 10.1109/TITS.2018.2815678.
[20] S. Kandasamy and S. Mangai, “A smart transportation system in VANET based on vehicle geographical tracking and balanced
routing protocol,” International Journal of Communication Systems, vol. 34, no. 4, 2021, doi: 10.1002/dac.4714.
[21] M. Akhtar and S. Moridpour, “A review of traffic congestion prediction using artificial intelligence,” Journal of Advanced
Transportation, vol. 2021, pp. 1–18, 2021, doi: 10.1155/2021/8878011.
[22] A. Benterki, M. Boukhnifer, V. Judalet, and C. Maaoui, “Artificial intelligence for vehicle behavior anticipation: hybrid approach based on
maneuver classification and trajectory prediction,” IEEE Access, vol. 8, pp. 56992–57002, 2020, doi: 10.1109/ACCESS.2020.2982170.
[23] M. D. Lytras, K. T. Chui, and R. W. Liu, “Moving towards intelligent transportation via artificial intelligence and internet-of-
things,” Sensors, vol. 20, no. 23, pp. 1–4, 2020, doi: 10.3390/s20236945.
[24] S. A. Sayed, Y. A. -Hamid, and H. A. Hefny, “Artificial intelligence-based traffic flow prediction: a comprehensive review,”
Journal of Electrical Systems and Information Technology, vol. 10, no. 1, 2023, doi: 10.1186/s43067-023-00081-6.
[25] A. Severino, S. Curto, S. Barberi, F. Arena, and G. Pau, “Autonomous vehicles: an analysis both on their distinctiveness and the
potential impact on urban transport systems,” Applied Sciences, vol. 11, no. 8, Apr. 2021, doi: 10.3390/app11083604.
[26] S. Tak, J. D. Lee, J. Song, and S. Kim, “Development of AI-based vehicle detection and tracking system for C-ITS application,”
Journal of Advanced Transportation, vol. 2021, pp. 78311–78319, 2021, doi: 10.1155/2021/4438861.
[27] Y. N. Malek, M. Najib, M. Bakhouya, and M. Essaaidi, “Multivariate deep learning approach for electric vehicle speed
forecasting,” Big Data Mining and Analytics, vol. 4, no. 1, pp. 56–64, 2021, doi: 10.26599/BDMA.2020.9020027.
[28] S. Chavhan, D. Gupta, B. N. Chandana, R. K. Chidambaram, A. Khanna, and J. J. P. C. Rodrigues, “A novel emergent
intelligence technique for public transport vehicle allocation problem in a dynamic transportation system,” IEEE Transactions on
Intelligent Transportation Systems, vol. 22, no. 8, pp. 5389–5402, 2021, doi: 10.1109/TITS.2020.3011198.
[29] C. Iliopoulou and K. Kepaptsoglou, “Combining ITS and optimization in public transportation planning: state of the art and future
research paths,” European Transport Research Review, vol. 11, no. 1, 2019, doi: 10.1186/s12544-019-0365-5.
[30] M. Eom and B. I. Kim, “The traffic signal control problem for intersections: a review,” European Transport Research Review,
vol. 12, no. 1, Sep. 2020, doi: 10.1186/s12544-020-00440-8.
[31] S. S. S. M. Qadri, M. A. Gökçe, and E. Öner, “State-of-art review of traffic signal control methods: challenges and opportunities,”
European Transport Research Review, vol. 12, no. 1, 2020, doi: 10.1186/s12544-020-00439-1.
[32] S. Lee, D. E. Lim, Y. Kang, and H. J. Kim, “Clustered multi-task sequence-to-sequence learning for autonomous vehicle
repositioning,” IEEE Access, vol. 9, pp. 14504–14515, 2021, doi: 10.1109/ACCESS.2021.3051763.
[33] S. Mohammadi and G. Farahani, “Computational intelligence-based connectivity restoration in wireless sensor and actor networks,”
Eurasip Journal on Wireless Communications and Networking, vol. 2020, no. 1, 2020, doi: 10.1186/s13638-020-01831-0.
[34] Z. G. Chen, Z. H. Zhan, S. Kwong, and J. Zhang, “Evolutionary computation for intelligent transportation in smart cities: a
survey,” IEEE Computational Intelligence Magazine, vol. 17, no. 2, pp. 83–102, 2022, doi: 10.1109/MCI.2022.3155330.
[35] G. Husnain, S. Anwar, G. Sikander, A. Ali, and S. Lim, “A bio-inspired cluster optimization schema for efficient routing in
vehicular ad hoc networks (VANETs),” Energies, vol. 16, no. 3, 2023, doi: 10.3390/en16031456.
[36] R. Jain et al., “Internet of things-based smart vehicles design of bio-inspired algorithms using artificial intelligence charging
system,” Nonlinear Engineering, vol. 11, no. 1, pp. 582–589, 2022, doi: 10.1515/nleng-2022-0242.
[37] C. Li, H. Zhao, K. Huang, and Y. H. Chen, “Optimal design for anti-skid control of electric vehicles by fuzzy approach,” Chinese
Journal of Mechanical Engineering (English Edition), vol. 34, no. 1, 2021, doi: 10.1186/s10033-021-00642-8.
[38] S. Li, K. Shu, C. Chen, and D. Cao, “Planning and decision-making for connected autonomous vehicles at road intersections: a
review,” Chinese Journal of Mechanical Engineering (English Edition), vol. 34, no. 1, 2021, doi: 10.1186/s10033-021-00639-3.
[39] V. Servizi, F. C. Pereira, M. K. Anderson, and O. A. Nielsen, “Transport behavior-mining from smartphones: a review,”
European Transport Research Review, vol. 13, no. 1, Nov. 2021, doi: 10.1186/s12544-021-00516-z.
[40] Y. Liu, Q. Xu, H. Guo, and H. Zhang, “A type-2 fuzzy approach to driver-automation shared driving lane keeping control of
semi-autonomous vehicles under imprecise premise variable,” Chinese Journal of Mechanical Engineering (English Edition), vol.
35, no. 1, 2022, doi: 10.1186/s10033-022-00706-3.

2539
[41] Y. Lu, M. Lang, Y. Sun, and S. Li, “A fuzzy intercontinental road-rail multimodal routing model with time and train capacity
uncertainty and fuzzy programming approaches,” IEEE Access, vol. 8, pp. 27532–27548, 2020, doi: 10.1109/ACCESS.2020.2971027.
[42] Q. Shi, M. Wang, Z. He, C. Yao, Y. Wei, and L. He, “A fuzzy-based sliding mode control approach for acceleration slip
regulation of battery electric vehicle,” Chinese Journal of Mechanical Engineering (English Edition), vol. 35, no. 1, 2022, doi:
10.1186/s10033-022-00729-w.
[43] V. Simić, D. Lazarević, and M. Dobrodolac, “Picture fuzzy WASPAS method for selecting last-mile delivery mode: a case study
of Belgrade,” European Transport Research Review, vol. 13, no. 1, 2021, doi: 10.1186/s12544-021-00501-6.
[44] J. Tang, X. Zhang, W. Yin, Y. Zou, and Y. Wang, “Missing data imputation for traffic flow based on combination of fuzzy neural
network and rough set theory,” Journal of Intelligent Transportation Systems: Technology, Planning, and Operations, vol. 25, no.
5, pp. 439–454, 2021, doi: 10.1080/15472450.2020.1713772.
[45] A. K. Haghighat, V. R. -Mouli, P. Chakraborty, Y. Esfandiari, S. Arabi, and A. Sharma, “Applications of deep learning in
intelligent transportation systems,” Journal of Big Data Analytics in Transportation, vol. 2, no. 2, pp. 115–145, 2020, doi:
10.1007/s42421-020-00020-1.
[46] R. Ojala, J. Vepsalainen, J. Hanhirova, V. Hirvisalo, and K. Tammi, “Novel convolutional neural network-based roadside unit for
accurate pedestrian localisation,” IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 9, pp. 3756–3765, Sep.
2020, doi: 10.1109/TITS.2019.2932802.
[47] F. Zhou, Q. Yang, T. Zhong, D. Chen, and N. Zhang, “Variational graph neural networks for road traffic prediction in intelligent
transportation systems,” IEEE Transactions on Industrial Informatics, vol. 17, no. 4, pp. 2802–2812, Apr. 2021, doi:
10.1109/TII.2020.3009280.
[48] I. Appaji and P. Raviraj, “Framework for simulation of vehicular communication using LSTM-based graph attention networks,”
Indian Journal Of Science And Technology, vol. 16, no. 16, pp. 1230–1240, Apr. 2023, doi: 10.17485/ijst/v16i16.1777.
BIOGRAPHIES OF AUTHORS
Impana Appaji holds an M.Tech. in Software Engineering, post-graduation from
SJCE Mysuru, affiliated to Visvesvaraya Technological University, Belagavi, and Karnataka,
India in 2014. She also received B.E. Degree in Computer Science from GSSSIETW, Mysuru,
and affiliated to Visvesvaraya Technological University, Belagavi, Karnataka, India in 2009.
She has 13 years of teaching experience. She is currently working as senior analyst at infosys.
She has published 12 research papers in international journals and conferences from 2009 to
2023. She can be contacted at email: impana.appaji@gmail.com.
Raviraj Pandian completed his doctorate degree in computer science and
engineering in the area of image processing. He holds a position of professor & head in the
Department of Computer Science & Engineering, GSSS Institute of Engineering &
Technology for Women, Mysore. Additionally, he also serves as an IQAC-Director at GSSS
Institute of Engineering & Technology for Women, Mysore. He has 16 years of teaching and
research experience. He has published more than 90 papers in international journals and
conferences and 8 book chapters. He serves as an editorial board member and reviewer for
more than 15 international journals. He is also a life member of professional bodies like
ISTE, CSI etc. He has guided the Ph.D. research scholars in the areas of image processing,
pervasive & cloud computing, bio-inspired algorithms, and robotics. He has received the
project grant Rs.5 Lakhs from the VGST, Govt. of Karnataka for the “underwater robotic fish
for surveillance and pollution monitoring”. He has served as a chair person and keynote
speakers of many national and international conferences. He has been served as a board of
studies member, subject expert, Ph.D. thesis evaluator for the various universities and
autonomous institutions. He has reviewed lot of articles in international conference
proceedings and journals. He has been received various awards such as ‘Rhastriya Gaurav
Award-2015’, ‘Shri P.K. das Memorial Best Faculty Award-2012’, ‘Young Achiever Award-
2016’, ‘Best Circuit Faculty Finalist Award-2017’ for his outstanding academic
achievements and contributions. He can be contacted at email: raviraj@gsss.edu.in.

Novel computational intelligence-based model for effective traffic management in intelligent transportation system

More Related Content

Similar to Novel computational intelligence-based model for effective traffic management in intelligent transportation system (20)

More from IAESIJAI (20)

Recently uploaded (20)

Novel computational intelligence-based model for effective traffic management in intelligent transportation system