Hindawi

Wireless Communications and Mobile Computing

Volume 2022, Article ID 1606317, 15 pages
[Link]

Research Article
Cluster-Based Routing Schema Using Harris Hawks
Optimization in the Vehicular Ad Hoc Networks

Behbod Kheradmand ,1 Ali Ghaffari ,2 Farhad Soleimanian Gharehchopogh ,1

and Mohammad Masdari 1
1
Department of Computer Engineering, Urmia Branch, Islamic Azad University, Urmia, Iran
2
Department of Computer Engineering, Tabriz Branch, Islamic Azad University, Tabriz, Iran

Correspondence should be addressed to Ali Ghaffari; ghaffari943@[Link]

Received 14 March 2022; Revised 12 June 2022; Accepted 2 July 2022; Published 14 July 2022

Academic Editor: B. D. Parameshachari

Copyright © 2022 Behbod Kheradmand et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.

Today, intelligent transportation systems (ITS) have received a lot of attention due to their significant impact on increasing the
safety, efficiency, and convenience of transportation. One of the main applications of ITS is vehicular ad hoc networks
(VANETs). In particular, a more flexible, reliable, real-time, and scalable routing scheme across the large urban areas is one of
the most critical issues for VANETs. Past VANET routing methods have various technical issues with VANET evolutions. On
the other hand, clustering improves the reliability and scalability of routing schemes in VANETs. In this paper, a cluster-based,
Traffic-aware and Low-Latency Routing Schema (TaLAR) is proposed for VANETs. In the proposed scheme, the Harris hawks
optimization (HHO) algorithm is used to select cluster head (CH) nodes by considering appropriate parameters such as
intracluster distance, link reliability, and relative speed of vehicles. The path between source and destination CH nodes is
identified by using the HHO algorithm; it chooses the appropriate route based on link reliability and intercluster distance.
Also, in the interconnection area, a traffic-aware and reliable route is identified by using a digital map of the streets and the
Dijkstra algorithm. The performance evaluation of the proposed scheme is analyzed in terms of packet delivery rate (PDR),
average end-to-end delay, and throughput. The output of the proposed scheme is compared with the Clustering Routing Based
on PSO (Particle Swarm Optimization) (CRBP) and Grey Wolf Optimization Based Clustering in Vehicular Ad Hoc Network
(GWOCENT) methods. Simulation results show that the proposed scheme improves PDR (22 and 19%), throughput (25 and
21%), and average end-to-end delay (23 and 18%).

1. Introduction the increasing trend of critical applications of VANET and its

dynamic nature (high mobility of vehicles), an efficient, real-
In recent years, VANETs as a special type of Mobile Ad hoc time, and reliable routing scheme is essential and plays a key
Networks (MANETs) [1] are aimed at transmitting safety role in supporting safety and improving the overall quality of
information from source to destination vehicles to improve service (QoS) in most applications of VANETs [4–9]. However,
passengers’ safety and prevent accidents. One of the critical VANETs have special limitations and characteristics, such as
issues in VANETs is the real-time and reliable information unsteady connectivity, frequent exchange of information, fast
transmission from the source to the destination vehicles to changes in network topology, and high vehicular node mobility
allow drivers to make appropriate and timely decisions that that can significantly impact data transmission schemes
increase road and passengers’ safety [2]. Recently, VANETs [10–12]. VANET data transmission schemes use V2V (vehi-
support different types of applications and services, such as cle-to-vehicle), V2I (vehicle-to-infrastructure), and hybrid
mobile vehicular cloud services, route discovery, traffic methods. In VANETs, the existing routing schemes can be clas-
monitoring, and context-sensitive infotainment [3]. Due to sified into the following subcategories: (a) broadcast-based
2 Wireless Communications and Mobile Computing

schemes are used when the vehicle node that will receive the Cellular base Cellular
message from other nodes is out of transmission range. This station Internet base station
scheme uses flooding data transmission, and every vehicle node
in the VANETs will receive the safety and security messages. RSU
This scheme improves the packet delivery rate. (b) Position- V2 V2
I

I
I

V2I

V2
based schemes are selected when the positions of the source
and destination nodes and the optimal route to the destination
vehicle node can be known using the global positioning system

VV22V
V
(GPS). This type of routing scheme does not need to maintain V
2
any routing table. (c) Topology-based schemes are based on the V
topology of VANET nodes and maintain a routing table in each

V2
vehicle node. This type of routing scheme is slower than other

V
schemes. (d) Cluster-based schemes increase scalability and
decrease the number of control messages via data aggregation
in CH nodes. CH nodes of two clusters send data to each other
and can make a proper decision on routing data packets. More-
CH node
over, several cluster-based routing schemes have been proposed
V2V communication
for VANETs [13–15]. Figure 1 shows a cluster-based routing
V2I communication
scheme in VANET. In this figure, CH gathers data from its
members, aggregates, and transmits that to another CH [16]. Cluster
Clustering in VANETs is an NP-hard problem, and metaheur-
istic algorithms are appropriate to solve this problem [17, 18]. CH to CH communication

This paper proposes a cluster-based routing scheme for Figure 1: Clustering in VANETs [19].
VANETs using the HHO algorithm. The main objective of
the proposed scheme is to improve the QoS (quality of ser-
vice) parameters such as packet delivery rate, average end-
to-end delay, and throughput. In the proposed scheme, the 2. Literature Review
HHO algorithm is used for the clustering process (CH selec-
tion) due to its less computational and timing complexities There are various research works have been made on cluster-
and high scalability features. In the interconnection area, based routing schemes for VANETs. Some researchers from
the proposed scheme obtains an adjacent matrix from the related works are discussed in this section.
map of that area. Then, identifies the shortest path between In [19], the authors proposed a grey wolf optimization-
the vehicles in the interconnection area using the Dijkstra based clustering algorithm for VANETs using the social
algorithm. behavior and hunting scheme of grey wolfs for creating opti-
Briefly, the main contributions of this paper are as mal clusters. The special characteristic of grey wolf nature
follows: makes it converge earlier, which creates the optimal number
of clusters. Simulation results show that it is an optimal and
cluster-based robust routing protocol for VANETs, which is
(i) The HHO algorithm is used for selecting appropri-
suitable for highway scenarios and can improve the PDR.
ate CH nodes considering proper parameters such
In [20], using a Tabu search algorithm, a reliable and
as link reliability, the relative speed of the vehicle,
multilevel routing scheme was introduced. Even if the topol-
and intracluster distance
ogy constantly changes, multilevel clustering-based routing
(ii) The shortest route between source and destination enables the organization and sustains the path. Furthermore,
CH nodes considering link reliability and interclus- it solves the problem of creating a local optimum trap using
ter distance is identified using the HHO algorithm the Tabu search. Tabu search uses an appropriate fitness
function to choose the solution among a set of solutions.
(iii) The Dijkstra algorithm is used for selecting a traffic- The effective parameters of this method are selecting the best
aware and reliable routing scheme in the intercon- route, including the distance between nodes, the speed of
nection area nodes, node angle, link sustainability, and link reliability.
(iv) Average end-to-end delay, packet delivery rate, and The evaluation of this method shows that PDR and average
throughput are improved because of the appropriate end-to-end delay are improved.
CH and path section using the HHO algorithm In [21], a reliable and cluster-based routing algorithm
is proposed for VANET. In this method, for a proper clus-
The rest of this article is organized as follows: Section 2 tering process, the authors consider different parameters
explains the related work in the field of routing in VANETs. such as node degree and node coverage by using an impe-
Section 3 presents the proposed TaLAR method. Section 4 rialist competitive algorithm. For selecting optimal CH
discusses the performance and comparative analysis of the nodes, this method uses Radiation-Based Function (RBF)
proposed scheme with existing methods. Finally, the conclu- and considers appropriate parameters such as free buffer
sion of this article is given in Section 5. size and speed of vehicles.
Wireless Communications and Mobile Computing 3

In [22], the authors proposed a hybrid genetic algorithm or “Z” and communication expiration time or “T.” This
(GA) to improve cluster maintenance in the Weighted K routing scheme includes two main parts: (a) determining
-medoid Clustering Algorithm (WKCA). In this model, the the forwarding vehicle node in the road based on the fuzzy
Tabu search algorithm was combined with the GA (genetic logic and (b) road selection at the road junction to choose
algorithm) to allow the search of the whole state space and the right route to reach the data to the destination node.
obtain the most suitable solution and avoid trapping the The authors have compared their proposed scheme with
local optima. This model enhances the allocation of nodes other well-known protocols such as MoZo, BRAVE, and
to clusters, which makes effective vehicle communication OFAODV. Simulation results show that this proposed
and ensures the reliability of the cluster architecture. scheme has higher performance in terms of average end-
According to the evaluation criteria, the results of this model to-end delay and control packet overhead.
are more stable and robust than the original WKCA and In [28], the authors proposed a new hybrid cluster-based
other approaches. routing algorithm using a modified K-means algorithm with
In [23], the authors presented a hybrid scatter Tabu continuous Hopfield network and maximum stable set
search (HSTS) method to allocate cluster members to the problem (KMRP). In this scheme, the maximum stable set
proper CH nodes. In this study, the WKCA (Weighted K problem combines with a continuous Hopfield network for
-medoid Clustering Algorithm) is used to form the cluster selecting optimal clusters. Then, each cluster member nodes
and to obtain the optimal solution. The Tabu search is inte- assign to CH nodes using the K-means algorithm consider-
grated into the internal scatter searching process to locate ing link reliability and distance. Finally, the cluster head
the global minimum. The simulation results show that this nodes are selected using an appropriate fitness function with
model improves network sustainability. different parameters such as the amount of free buffer space,
In [24], effective cluster-based routing scheme in the speed, and the node degree. KMRP increases throughput
VANET was presented to improve V2V performance using metric by reducing traffic congestion. In addition, the KMRP
PSO algorithm. Three basic factors in selecting clusters are method reduces the average end-to-end latency and
considered for the sustainability of clusters: alignment with increases the PDR.
nodes, more neighbors, and similar velocity and position In [29], the author proposed a cluster-based routing pro-
with other nodes. First, the CH nodes are chosen. Second, tocol for hybrid VANET-WSN communication (PRAVN)
to optimize the necessary routing, particle path, node’s speed for road safety applications. The author used improved
coding rules, iteration rules, and proper functions are water wave optimization (IWWO) algorithm for clustering
planned. Third, the routing method is used in clusters and the vehicle nodes. An important contribution of PRAVN is
among clusters. The results show that the number of nodes, the introduction of a rider optimization (RO) algorithm to
communication radius of nodes, and maximum hops choose neighbor vehicle nodes to improve the network life-
between the CH nodes and each normal node in the cluster time and lossless connections. This method by forwarding
have an important impact on the proposed scheme valid data from the sender to the receiver vehicle nodes will
performance. improve road safety.
In [25], the author presented an enhanced cluster-based In [30], the authors proposed a new improvement to
AODV for the Internet of Vehicles (AODV-CD) to get a sta- the scheme of OLSR (Optimized Link State Routing Proto-
ble and efficient clustering to simplify routing and ensure col), named CACA (Cluster-based Adept Cooperative
QoS. To achieve cluster stability, this method applies two Algorithm). In this scheme, each node identified a reliable
unique messages, i.e., the HELLO packet and the CH packet, route between source and destination using the cluster-
in the route discovery phase and the path conservation pro- based QoS algorithm. The CACA is aimed at maintaining
cess in the clustering algorithm for the AODV protocol. The long-lived paths for which the most sustainable path is
route replay (RREP) message is transmitted to nodes if a adaptively selected considering different parameters such
path is available; otherwise, the CH will send the RREQ mes- as signal strength and vehicle node mobility. It decreases
sage. Due to reduction in RREQ messages, this method control message overhead and updates the routing table.
reduces congestion and network overload. The work in [3] proposes an Innovative Cluster-Based
In [26], the authors presented a destination-aware Dual-Phase Routing Protocol Using Fog Computing and
context-based routing (DACR) method using a clustering Software-Defined Vehicular Network (ICDRP-F-SDVN).
process with soft computing for VANETs. This scheme The authors combined fog computing technology and
consists of two phases. In the first phase, they combined geo- Software-Defined Networks (SDN) and proposed a reliable
graphic and context-based clustering scheme. The object of and robust architecture that overcomes critical challenges
this clustering scheme was to prevent network congestion arising from advanced technological development and fast
and reduce clustering overhead. In the second phase, a rout- escalation of intelligent vehicle nodes. This scheme can
ing algorithm based on destination is presented for data reduce the overhead of control messages.
transmission between two cluster nodes which improves Table 1 shows the compression of the routing scheme for
the overall PDR and end-to-end delay. vehicle ad hoc network literature.
In [27], the authors proposed a fuzzy-logic-based rout- Using an intelligent routing scheme can effectively
ing scheme for VANETs to improve the QoS parameters. improve the quality of service parameters such as average
They have defined two important parameters to identify end-to-end delay, PDR, and throughput in VANETs. How-
the forwarding vehicle node: channel quality factor (CQF) ever, for critical applications of VANETs, PDR and average
4 Wireless Communications and Mobile Computing

Table 1: Compression of the routing schema for vehicular ad hoc network literature.

References Architecture Advantages Disadvantages

[19] V2V Throughput and PDR are improved Increases the overhead of network
[20] V2V Improves the PDR and throughput High latency
Improved energy efficiency, throughput, and end-to-
[21] V2V High overhead and less scalable
end latency
[22] V2V Improves the stability and robustness High overhead
[23] V2V Improves the network sustainability High overhead and high end-to-end delay
[24] V2V Improves the QoS parameters High overhead and less scalable
[25] V2V Improves network overhead Less scalable
Not suitable for urban environments; loop occurs in
[24] V2V Improves stability and scalability
routing process
[25] V2V Improves the average end-to-end delay High network overhead
[28] V2V Less routing overhead High time complexity
[29] V2V Less overhead, high PDR Not scalable
[30] V2V Low average end-to-end delay, high PDR High overhead
[3] V2V Low overhead, high PDR, less end-to-end delay Not scalable

end-to-end delay should be improved. The Harris hawks wherein the Euclidean distance is computed using Equation
optimization (HHO) algorithm finds the optimal path with (2) as follows:
the search and hunt mechanism. The advantages of HHO
are high accuracy, fast convergence, and easy realization distðx, yÞ
appear. Therefore, this paper uses the HHO algorithm to qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
= ðx1 − x2 + T:Sðsin α − sin βÞÞ2 + ðy1 − y2 + T:Sðcos α − sin βÞÞ2 ,
select optimal CH nodes and identifies optimized paths
among clusters and improves the quality of service ð2Þ
parameters.
where we have
3. The Proposed Scheme
L1 = x1 − x2 ,
The proposed method has the following two main phases. In
the first phase, an appropriate clustering scheme is proposed L2 = T:Sðsin α − sin βÞ,
using an HHO algorithm with a new fitness function to ð3Þ
L3 = y 1 − y 2 ,
select proper CH nodes. In the second phase, a reliable and
real-time routing scheme is proposed using the HHO algo- L4 = T:Sðcos α − sin βÞ:
rithm to find a reliable and real-time path between source
and destination CH nodes. This section explains the pro- 3.1.2. Relative Speed of Vehicles. The relative velocity V l is an
posed scheme in detail. important parameter for selecting CH nodes and can be cal-
culated using the following equation:
3.1. Clustering Phase. Variable description used in all equa-
tions is given in Table 2. pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
We assume that the vehicles are moving at the same Min f 2 = V l = S 2ð1 − sin θÞ, ð4Þ
speed and S1 = S2 = S. For appropriate CH selection, a new
fitness function is obtained using the following parameters: whereθis the angle between nodesxandy.

3.1.1. Intracluster Distance. This important parameter is the 3.1.3. Link Reliability. In VANETs, due to node mobility
average distance between each cluster head node (CH) and and high relative velocity of vehicle nodes, finding a reli-
its cluster members and can be calculated using Equation able path is a complicated task. In this regard, the link reli-
(1). If this parameter is minimum, then data transmission ability is an important metric for discovering optimum
latency is also minimum. Hence, the intracluster distance path between vehicle nodes. Therefore, the link reliability
is f 1 , which can be minimized as follows: Ri among of the vehicles can be calculated using the fol-
lowing equation:

Ij ! qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
m
∑i=1 distðni , CHi Þ ðL1 + L2 Þ2 + ðL3 + L4 Þ2
Min f 1 = 〠 , ð1Þ Ri = : ð5Þ
j=1 Ij Vl
Wireless Communications and Mobile Computing 5

Table 2: Variable description.

Variable name Definitions

s The speed of a vehicle
Vl The relative velocity with angle ɵ
Ri The amount of reliability for a given link is calculated according to Equation (3)
Distðx, yÞ The distance between two vehicles at the same speed and its calculation will be in Equation (1)
Ij Number of intracluster nodes of each cluster
m Number of clusters
 
Dist ni , CH j Distance between the ith vehicle node and the jth CH
FFðvÞ Vehicle’s fitness function
n Size of population
ι Location vector hawks
δ Random vector
q Random number
E0 Initial escape energy
Ep Escape energy
Y Next to running step
LFðDÞ Levy flight
MI Maximum number of iterations
Ai Packets received by node i
Zi Sent packets by node i
ES Size of data packets
DP Simulation stop time
DT Simulation start time
Ex Number of executions

Therefore, two vehicle nodes with higher link reliability As we know that all the above-mentioned parameters
is a more preferable choice as a CH. It enhances the packet have different units, therefore, min–max normalization
delivery rate (PDR) and reduces the packet loss. Hence, our function is applied to each parameter using the following
other objective in terms of link reliability is f 3 , which can equation:
be minimized using the following equation:
m f i − f min
1 F ðx Þ = , ð8Þ
Minimize f 3 = 〠 : ð6Þ f max − f min
R
i=1 i

where f i is the value of the function, f min is minimum value,

All of the above-mentioned parameters of fitness func-
f max is maximum value, and FðxÞ is the normalized value
tion must be converted into a single fitness function.
between 0 and 1.
Therefore, the weighted sum approach is applied and the
After selecting the CH nodes, each CH node advertises
fitness function can be obtained using the following equa-
itself as a candidate node to the neighbor nodes. The neigh-
tion:
bor vehicle nodes are allocated to the CH nodes with less
distance.
Minimize fitness = ω1 × f 1 + ω2 × f 2 + ω3 × f 3 , ð7Þ
3.2. Routing Process Using HHO. The process of finding the
where ω1 , ω2 , and ω3 are the weights assigned to each shortest path from the source vehicle to the destination vehi-
objective. The value of these weight parameters is between cle is done using HHO [31]. The purpose of using the HHO
0 and 1, and ω1 + ω2 + ω3 = 1: The values of these param- method is to find the optimal and reliable path based on link
eters are as 0.3, 0.45, and 0.25, respectively. The link reli- reliability and interclusters distance. Hawks have to select
ability is considered as high priority to select CH node, paths that are more reliable and shorter than other paths.
and so the value of ω2 is 0.45. The fitness function of selecting the optimal next CH node
6 Wireless Communications and Mobile Computing

Inputs: Size n of the population and MI is the total number of repetitions

Outputs: The position of prey and its fitness function
Step 1:
Initialize parameters of each vehicle node such as position, direction, and the velocity;
make a mesh topology across nodes/vertices;
Initialize Population v1, v2, . . . , vn
Setting Vp (t) as the best path for the prey
Step 2:
for t=1 to MI, repeat the following steps: // the size of the maximum iterations
Compute ff(vi)// Calculate the fitness function
Update Vp (t) if there is a better solution
for ðv i = 1 to nÞ // vi is hawks
Compute Ei=2×rand (), δ=2× (1-rand ()) and Ep based on Eq. (12)
if ðjEpj ≥ 1Þ then compute Vðt + 1Þ //Exploration phase
if ðjEpj < 1Þ then // Exploitation phase
if (ɩ ≥0.5 && |Ep| ≥0.5)
compute Vðt + 1Þ based on Eq. (13)
elseif (ɩ ≥0.5 && |Ep| <0.5)
compute Vðt + 1Þ based on Eq. (14)
elseif (ɩ <0.5 && |Ep| ≥0.5)
compute Vðt + 1Þ based on Eq. (17)
elseif (ɩ <0.5 && |Ep| <0.5)
compute Vðt + 1Þ based on Eq. (18)
Current iteration = current iteration+1
end for
update Vp
end

Algorithm 1: Reliable routing scheme using the HHO algorithm.

l12 respectively. Min–max normalization function is applied to

j1 j2
each parameter of Equation (9) using Equation (8). The
value of φ is 0.6, and this weighting parameter is considered
l13 Street for taking link reliability as higher priority in selecting next
CH node as forwarding node.
The steps of the proposed HHO routing algorithm are as
j3 j4 follows.
l34
3.2.1. Exploration and Transformation Phases. In this phase,
Figure 2: Part of the city map with marked intersections. the Harris hawks are often placed in special positions, wait,
search, and detection of prey based on the following equa-
j1 j2 j3 j4 tion [31]:
j1 0 l12 ∞ l14 (
V rand ðt Þ − ι1 ðV rand ðt Þ − 2ι2 V ðt ÞÞq > 0:5,
j2 l21 ∞ l24 X ðt + 1 Þ =
0  
j3 l31 ∞ 0 l34 V ip ðt Þ − V i ðt Þ − ι3 ðLB + ι4 ðUB − LBÞÞq < 0:5:
j4 ∞ l42 l43 0 ð10Þ

Figure 3: Matrix of the proximity of a part of the city map with the Here, t and t + 1 show the current and the next iteration,
marked intersections.
respectively. In addition, ι1 , ι2 , ι3 , and ι4 show the position
vectors of the hawks, as q is the random number in the range
for data transmission can be obtained from the following of [0,1]. The average position of the hawks can be calculated
equation: using the following equation [31]:
  1
ff ðvi Þ = φ × Ri + ð1 − φÞ × dist CHi , CH j , ð9Þ V i ðt Þ = , ð11Þ
n

where Ri , distðCHi , CH j Þ, and φ are link reliability, distance where V i ðtÞ and n indicate the positions of each vehicle in
between two neighbor CH nodes, and weighting parameter, iteration t and the overall number of vehicles. One of the
Wireless Communications and Mobile Computing 7

Table 3: Simulation parameters. The above model calculates the next moving stage (Y) of
the hawks. Hawks also dive to attack prey according to the
Parameters Value following equation:
OS Linux Ubuntu
Population size (particles) 100 Z = Y + δ × LFðDÞ, ð16Þ
Maximum iterations 150
Communication range 10 m-30 m where LF represents the levy flight function and δ is the ran-
Maximum speed of the nodes 40 km/h dom vector for size 1 × D; at this point, the position is
MAC protocol 802.11p updated as follows [31]:
Time 2000 second
Number of vehicle nodes 500 (
Y if F ðY Þ < F ðV ðt ÞÞ,
Packet size 64 bytes V ðt + 1 Þ = ð17Þ
Transaction size 75 bytes Z if F ðZ Þ < F ðV ðt ÞÞ:
Number of executions 5
ω1 , ω 2 , ω 3 0.3, 0.45, 0.25
(4) Hard Besiege with Progressive Rapid Dives.
φ 0.6 (
Y if F ðY Þ < F ðV ðt ÞÞ,
VV ðt + 1Þ = ð18Þ
Z if F ðZ Þ < F ðV ðt ÞÞ:

important stages of HHO is the transition from the explora- From here, Y and Z are estimated using the following
tion phase to the exploitation phase; this change is expected equation [31]:
between the different simulated exploitative behaviors based
on the evading energy factor Ep of the prey (V ip ), as it  
Y = V p ðt Þ − EδV p ðt Þ − V i ðt ÞZ = Y + δ × LFðDÞ: ð19Þ
reduces dramatically during the evading behavior. The
energy of the prey can be calculated using the following
equation [31]: Algorithm 1 shows the routing process of the proposed
method.
 
t
Ep = 2E0 1 − , ð12Þ
T 3.3. Traffic-Aware Routing in Interconnection Area Using
HHO. Each vehicle can obtain a graph adjacent using a dig-
where E0 , T, and Ep show the initial escape energy, total ital street map that models the city map. Whenever a vehicle
number of iterations, and the escape energy, respectively. node wants to send a packet to another node, it must first
add two source and destination vertices to the matrix. It then
3.2.2. Exploitation Phase calculates the shortest route using the Dijkstra algorithm
and sends the packet to the destination vehicle. The marked
(1) Soft Besiege. The HHO soft besiege behavior can be cal- intersections of the city map in the circles jn and the street
culated using the following equation [31]: sections with length lij are given in Figure 2. In the graph
  of this map, the intersections with the vertices and the streets
V ðt + 1Þ = ΔV ðt Þ − EδV p ðt Þ − V ðt ÞΔV ðt Þ = V p ðt Þ − V ðtÞ, are shown as edges. Also, the corresponding proximity
ð13Þ matrix is shown in Figure 3 [32].
Using the Dijkstra algorithm on the graph, which shows
the city map, the list of intersections is calculated. The
where δ is the jump intensity of escaping process that can be weight graph of each edge is determined by connecting that
demonstrated as δ = 2 × ð1 − ι5 Þ. segment of the street. The weight of each edge is propor-
tional not only to its length but also to the network connec-
(2) Hard Besiege. In this case, the current locations are tion, and the matrix values are redefined in Figure 3 as
obtained using the following equation [31]: lij /Y ij , where 0 < Y ij < 1:
A weak connection is indicated by the value of Y ij due to
V ðt + 1Þ = V p ðt Þ − EjΔV ðt Þj: ð14Þ
low traffic. The vehicles update Y ij values. The next street is
chosen randomly by considering the number of vehicles.
(3) Soft Besiege with Progressive Rapid Dives. We assumed The vehicle node sets the nearest neighbor for everyone
that hawks could compute (decide) their next run based on based on the next intersection [32]. The connection to pass
the following equation [31]: through the street section between the intersections of ji
  and j j and subsequently Y ij and Y ji will be updated using
Y = V CH ðt Þ − EδV p ðt Þ − V ðt Þ: ð15Þ the HHO algorithm when the intersections of ji and j j are
8 Wireless Communications and Mobile Computing

100

95

90

Packet delivery rate (%)

85

80

75

70

65
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(a) Transmission range = 10
100

95

90
Packet delivery rate (%)

85

80

75

70

65
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(b) Transmission range = 20

Figure 4: Continued.
Wireless Communications and Mobile Computing 9

100

95

90

Packet delivery rate (%)

85

80

75

70

65
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(c) Transmission range = 30

Figure 4: PDR vs. a number of vehicles.

Table 4: PDR (%) vs. number of vehicles (transmission range = 30).

where delayij , min delayij , and B are the time spent, the
Number of vehicles CRBP GWOCNET TaLAR minimum delay recorded by the vehicle for that street,
50 68 74 82 and a fixed value, respectively. Since the minimum delayij
100 71 76 84 is less than or equal to delayij , the variable ð2/πÞ ∗ arctan
150 73 77 85
ðmindelayij /delayij Þ is between zero and 0.5. Therefore,
the required values for variable B will be less than 0.5,
200 74 78 87
and for variable ΔY ij less than 1. Also, the initial value for
250 76 80 89
the Y ij variable will be less than 1. Each time Y ij is renewed,
300 78 82 93
so the value of ðlij /Y ij Þ decreases. At constant intervals of
350 79 83 95
each vehicle, Y ij reduces all streets using the following
400 81 84 96
equation [32].
450 82 86 97
500 83 88 98  
Y ij = max 0:1, α × Y ij , ð22Þ

registered in the two consecutive connection points based on where 0 < α < 1. After a period of time, we reduce the Y ij
the following equation [32].
pheromone with Equation (22). So, we multiply the previ-
! ous pheromone Y ij by α.
Y ij + ΔY ij Therefore, the route chosen by the vehicle of the source
Y ij = Y ji = , ð20Þ will be appropriate to the traffic conditions of that area. Also,
1 + ΔY ij
each vehicle node regularly broadcasts its ID and location on
its radio range in that area. Each vehicle sets its timer so that
where the variable ΔY ij is computed based on the following fewer monitors are needed.
equation [32]:
4. Performance Evaluation
!
2 min delayij The proposed TaLAR approach had been simulated, and its
ΔY ij = B + × arctan , ð21Þ performance was evaluated in Network Simulator version 2
π delayij
(NS-2) on Linux Ubuntu 18.04 LTS environment. This
10 Wireless Communications and Mobile Computing

7500

7000

6500

6000

Throughput (kbps)
5500

5000

4500

4000

3500

3000

2500
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(a) Transmission range = 10
7500

7000

6500

6000
Throughput (kbps)

5500

5000

4500

4000

3500

3000

2500
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(b) Transmission range = 20

Figure 5: Continued.
Wireless Communications and Mobile Computing 11

8500

7500

6500

Throughput (kbs)
5500

4500

3500

2500
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(c) Transmission range = 30

Figure 5: Throughput vs. a number of vehicles.

sent from the source vehicle and can be calculated through

Table 5: Throughput (Kbps) vs. a number of vehicles the following equation [33]:
(transmission range = 30).
1 ∑n Ai
Number of vehicles CRBP GWOCNET TaLAR PDR = × i=1 × 100%: ð23Þ
50 3019 4109 5720 Ex ∑ni=1 Z i
100 3200 4388 5960
150 3500 4518 6200 Throughput: this important metric is the ratio of the total
size of data packets delivered to the
200 3798 4699 6489
receiver to the total simulation process time and can be
250 3980 4811 6889 obtained from the following equation [34]:
300 4019 4998 6998
350 4298 5140 7412 1 ∑n A ∗ Es 8
Throughput = × i=1 i × : ð24Þ
400 4490 5238 7609 Ex Dp − D T 1000
450 4709 5487 7818
500 4816 5691 7909 Average end-to-end delay: this metric is the average time
interval between receiving and sending times for a packet
from a transmitter to receiver node and can be calculated
through the following equation:
section indicates the performance evaluation of the proposed
scheme to validate its efficiency. The simulation results of 
p
the proposed scheme were compared with both methods ∑ j=1 T rec j − T send j
(CRBP [24] and GWOCENT [19]). Delay = , ð25Þ
p

4.1. Performance Metrics. The proposed scheme and other where T rec j , T send j , and p are receiving time, sending time of
two methods were compared based on three metrics PDR, a packet j, and the total number of packets, respectively.
throughput, and average end-to-end delay.
PDR: the total number of data packets that arrived at the 4.2. Simulation Results. The basic simulation parameters used
receiver vehicle is divided by the total number of packets in the performance evaluation phase are listed in Table 3.
12 Wireless Communications and Mobile Computing

5000

4500

4000

Average end-to-end delay (ms)

3500

3000

2500

2000

1500

1000

500
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(a) Transmission range = 10
5000

4500

4000
Average end-to-end delay (ms)

3500

3000

2500

2000

1500

1000

500
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(b) Transmission range = 20

Figure 6: Continued.
Wireless Communications and Mobile Computing 13

5000

4500

4000

Average end-to-end delay (ms)

3500

3000

2500

2000

1500

1000

500
50 100 150 200 250 300 350 400 450 500

Number of nodes
CRBP
GWOCNET
TaLAR
(c) Transmission range = 30

Figure 6: Average end-to-end delay vs. a number of vehicles.

networks. Stable and reliable routes will increase the packet

Table 6: Average end-to-end delay (ms) vs. a number of vehicles
delivery rate. The use of these important parameters and
(transmission range = 30).
the HHO algorithm contributes to the selection of a reliable,
Number of vehicles CRBP GWOCNET TaLAR noncongested short route to increase PDR. In Figure 4(a), in
50 4189 3508 2221
the transfer range equal to 10, the proposed TaLAR method
performs better than the CRPB and GWOCNET methods
100 4006 3478 2000
and has a higher packet delivery rate. In Figures 4(b) and
150 4004 3218 1729 4(c), this superiority is maintained as the transmission range
200 3980 3005 1680 increases. In the CRPB method, only clusters based on PSO
250 3877 2879 1356 are selected to send data, and practically, no other strategy
300 3800 2690 1133 for transmitting data is considered. In the GWOCNET
350 3600 2432 1011 method, only clusters are selected based on the grey wolf
400 3532 2396 728
algorithm and only with the distance criterion.
Throughput: Figure 5 and Table 5 show the throughput
450 3419 2140 617
versus the number of vehicle nodes for three other schemes:
500 3219 2004 592 TaLAR, CRPB, and GWOCNET. The results of the simula-
tion of TaLAR show that the proposed method in this
criterion is better than the other two methods. In the sce-
PDR: PDR has decreased due to path breakdown, con- nario with a transfer range of 10 in Figure 5(a), the proposed
gestion, and short lifespan link. As shown in Figure 4 and method has a much higher throughput than the CRPB and
Table 4, the PDR rate of the proposed TaLAR method is GWOCNET. This is because the proposed scheme deter-
higher than CRPB and GWOCNET in all three scenarios mines the optimal CH nodes based on several significant
by changing the transmission range. The reason is that the parameters such as link reliability, the relative speed of
proposed TaLAR method considers appropriate fitness func- vehicles, and intracluster distance. The proposed method
tion with several significant parameters such as link reliabil- determines the optimal route based on important QoS
ity, relative speed, and intracluster distance for clustering parameters such as link reliability using the HHO algorithm.
and routing process. Selecting CH nodes using the HHO In general, more vehicle nodes in the VANETs increase the
algorithm with proper fitness function based on link reliabil- throughput metric. This is because a vehicle gets more for-
ity and relative vehicle speed will increase the stability of warder vehicle nodes with high reliability to forward its data.
14 Wireless Communications and Mobile Computing

However, this significant improvement and the result are the Conflicts of Interest
idea of using the HHO with several significant parameters
for the clustering and routing process. This combination The authors declare that they have no conflicts of interest.
determines reliable and robust CH nodes and routes
between source and destination vehicle nodes.
Average end-to-end delay: Figure 6 and Table 6 show that References
the proposed TaLAR method performs better in terms of
[1] S. Ghasemnezhad and A. Ghaffari, “Fuzzy logic based reliable
average end-to-end delay than the CRPB and GWOCNET and real-time routing protocol for mobile ad hoc networks,”
methods and has less latency in all three scenarios. The Wireless Personal Communications, vol. 98, no. 1, pp. 593–
CRPB method is considered an only node and link stability 611, 2018.
in the network to reduce latency. In the GWOCNET method [2] N. Phull, P. Singh, M. Shabaz, and F. Sammy, “Enhancing
for selecting CH nodes, only the distance criterion is consid- vehicular ad hoc networks’ dynamic behavior by integrating
ered. In the proposed method, the use of the HHO algorithm game theory and machine learning techniques for reliable
with considering several significant parameters such as link and stable routing,” Security and Communication Networks,
reliability, intracluster distance, and the relative speed of vol. 2022, Article ID 4108231, 11 pages, 2022.
vehicles avails reliable and real-time path from source to [3] K. A. Darabkh, B. Z. Alkhader, K. Ala'F, F. Jubair, and
destination vehicle, which improves the end-to-end delay. M. Abdel-Majeed, “ICDRP-F-SDVN: an innovative cluster-
Therefore, for these reasons, in the proposed TaLAR based dual-phase routing protocol using fog computing and
method, with increasing the data transfer rate and the num- software-defined vehicular network,” Vehicular Communica-
ber of vehicles, the average delay compared to other methods tions, vol. 34, article 100453, 2022.
has decreased significantly. In the proposed scheme, consid- [4] F. Abbas and P. Fan, “Clustering-based reliable low-latency
ering the link reliability in selecting the next CH node for routing scheme using ACO method for vehicular networks,”
data transmission decreases the packet loss and hence, the Vehicular Communications, vol. 12, pp. 66–74, 2018.
retransmission of lost packets will be decreased significantly. [5] L. Rivoirard, M. Wahl, P. Sondi, M. Berbineau, and D. Gruyer,
This reduction of retransmission decreases network conges- “Chain-branch-leaf: a clustering scheme for vehicular net-
works using only V2V communications,” Ad Hoc Networks,
tion and improves the average end-to-end delay.
vol. 68, pp. 70–84, 2018.
[6] A. Lakas, M. E. A. Fekair, A. Korichi, and N. Lagraa, “A multi-
constrained QoS-compliant routing scheme for highway-
5. Conclusion based vehicular networks,” Wireless Communications and
Mobile Computing, vol. 2019, Article ID 4521859, 18 pages,
VANET requirements are having more reliable, scalable, and 2019.
well-connected routes. This paper has proposed a cluster- [7] H. Bagherlou and A. Ghaffari, “A routing protocol for vehicu-
based reliable and real-time routing scheme (TaLAR) for lar ad hoc networks using simulated annealing algorithm and
VANETs using the HHO algorithm. In the proposed neural networks,” The Journal of Supercomputing, vol. 74,
scheme, a new clustering scheme was presented considering no. 6, pp. 2528–2552, 2018.
several significant parameters such as the relative velocity of [8] M. Sindhwani, R. Singh, A. Sachdeva, and C. Singh, “Improvi-
vehicle nodes, intracluster distance, and link reliability. The sation of optimization technique and AODV routing protocol
proposed method has the following two main phases routing in VANET,” Materials Today: Proceedings, vol. 49, pp. 3457–
scheme for road and interconnection scenarios. The main 3461, 2022.
goal of the first phase was to find a reliable and real-time [9] G. Husnain and S. Anwar, “An intelligent probabilistic whale
path during the route discovery process between source optimization algorithm (i-WOA) for clustering in vehicular
and destination CH nodes. Also, the goal of the second ad hoc networks,” International Journal of Wireless Informa-
phase was to identify the shortest path between vehicle tion Networks, vol. 29, no. 2, pp. 143–156, 2022.
nodes using a digital street map and the Dijkstra algorithm. [10] J. Li, B. N. Silva, M. Diyan, Z. Cao, and K. Han, “A clustering
The proposed scheme is scalable and applicable for V2V and based routing algorithm in IoT aware wireless mesh net-
V2I and hybrid architectures of VANETs. The results of works,” Sustainable Cities and Society, vol. 40, pp. 657–666,
2018.
simulation in NS-2 showed that the TaLAR method has
much better performance than CRBP and GWOCENT [11] N. Chaib, O. S. Oubbati, M. L. Bensaad, A. Lakas, P. Lorenz,
and A. Jamalipour, “BRT: bus-based routing technique in
methods in terms of PDR (22 and 19%), throughput (25
urban vehicular networks,” IEEE Transactions on Intelligent
and 21%), and average end-to-end delay (23 and 18%), Transportation Systems, vol. 21, no. 11, pp. 4550–4562,
respectively. 2019.
In the future, we will propose an SDN-based architec- [12] P. K. Pandey, V. Kansal, and A. Swaroop, “OCSR: overlapped
ture for the proposed scheme to make it more robust and cluster-based scalable routing approach for vehicular ad hoc
reliable. networks (VANETs),” Wireless Communications and Mobile
Computing, vol. 2022, Article ID 1426815, 15 pages, 2022.
[13] D. Zhang, M. Zhang, F. Ding, S. E. Li, and K. Li, “A stability-
Data Availability based clustering scheme for vehicular networks,” in 2020 IEEE
3rd International Conference on Electronics Technology
No data were used to support this study. (ICET), pp. 809–813, Chengdu, China, 2020.
Wireless Communications and Mobile Computing 15

[14] Y. Yuan, R. He, B. Ai et al., “A 3D geometry-based THz chan- [28] K. Kandali, L. Bennis, and H. Bennis, “A new hybrid routing
nel model for 6G ultra massive MIMO systems,” IEEE Trans- protocol using a modified K-means clustering algorithm and
actions on Vehicular Technology, vol. 71, no. 3, pp. 2251– continuous Hopfield network for VANET,” IEEE Access,
2266, 2022. vol. 9, pp. 47169–47183, 2021.
[15] N. Moussa and A. E. B. El Alaoui, “DACOR: a distributed [29] M. Raja, “PRAVN: perspective on road safety adopted routing
ACO-based routing protocol for mitigating the hot spot prob- protocol for hybrid VANET-WSN communication using bal-
lem in fog-enabled WSN architecture,” International Journal anced clustering and optimal neighborhood selection,” Soft
of Communication Systems, vol. 35, article e5008, 2022. Computing, vol. 25, no. 5, pp. 4053–4072, 2021.
[16] S. Sefati, M. Abdi, and A. Ghaffari, “Cluster-based data trans- [30] N. M. Al-Kharasani, Z. A. Zukarnain, S. K. Subramaniam, and
mission scheme in wireless sensor networks using black hole Z. M. Hanapi, “An adaptive relay selection scheme for enhanc-
and ant colony algorithms,” International Journal of Commu- ing network stability in VANETs,” IEEE Access, vol. 8,
nication Systems, vol. 34, no. 9, article e4768, 2021. pp. 128757–128765, 2020.
[17] A. Kamalinia and A. Ghaffari, “Hybrid task scheduling [31] A. A. Heidari, S. Mirjalili, H. Faris, I. Aljarah, M. Mafarja, and
method for cloud computing by genetic and DE algorithms,” H. Chen, “Harris hawks optimization: algorithm and applica-
Wireless Personal Communications, vol. 97, no. 4, pp. 6301– tions,” Future Generation Computer Systems, vol. 97,
6323, 2017. pp. 849–872, 2019.
[18] S. J. Jazebi and A. Ghaffari, “RISA: routing scheme for Internet [32] F. Goudarzi, H. Asgari, and H. S. Al-Raweshidy, “Traffic-aware
of Things using shuffled frog leaping optimization algorithm,” VANET routing for city environments—a protocol based on
Journal of Ambient Intelligence and Humanized Computing, ant colony optimization,” IEEE Systems Journal, vol. 13,
vol. 11, no. 10, pp. 4273–4283, 2020. no. 1, pp. 571–581, 2019.
[19] M. Fahad, F. Aadil, Z. U. Rehman et al., “Grey wolf optimiza- [33] L. Azari and A. Ghaffari, “Proposing a novel method based on
tion based clustering algorithm for vehicular ad-hoc net- network-coding for optimizing error recovery in wireless sen-
works,” Computers & Electrical Engineering, vol. 70, pp. 853– sor networks,” Indian Journal of Science and Technology, vol. 8,
870, 2018. no. 9, pp. 859–867, 2015.
[20] E. Moridi and H. Barati, “RMRPTS: a reliable multi-level rout- [34] H. D. Nikokheslat and A. Ghaffari, “Protocol for controlling
ing protocol with tabu search in VANET,” Telecommunication congestion in wireless sensor networks,” Wireless Personal
Systems, vol. 65, no. 1, pp. 127–137, 2017. Communications, vol. 95, no. 3, pp. 3233–3251, 2017.
[21] M. Mohammadnezhad and A. Ghaffari, “Hybrid routing
scheme using imperialist competitive algorithm and RBF neu-
ral networks for VANETs,” Wireless Networks, vol. 25, no. 5,
pp. 2831–2849, 2019.
[22] R. Hajlaoui, E. Alsolami, T. Moulahi, and H. Guyennet, “Con-
struction of a stable vehicular ad hoc network based on hybrid
genetic algorithm,” Telecommunication Systems, vol. 71, no. 3,
pp. 433–445, 2019.
[23] R. Hajlaoui, T. Moulahi, and H. Guyennet, “A scatter search
based heuristic for reliable clustering in vehicular ad hoc net-
works,” in IFIP international conference on artificial intelli-
gence applications and innovations, L. Iliadis, I.
Maglogiannis, and V. Plagianakos, Eds., vol. 519 of Artificial
Intelligence Applications and Innovations, pp. 507–519,
Springer, Cham, 2018.
[24] X. Bao, H. Li, G. Zhao, L. Chang, J. Zhou, and Y. Li, “Efficient
clustering V2V routing based on PSO in VANETs,” Measure-
ment, vol. 152, article 107306, 2020.
[25] S. Ebadinezhad, “Design and analysis of an improved AODV
protocol based on clustering approach for Internet of Vehicles
(AODV-CD),” International Journal of Electronics and Tele-
communications, vol. 67, 2021.
[26] K. Aravindhan and C. S. G. Dhas, “Destination-aware context-
based routing protocol with hybrid soft computing cluster
algorithm for VANET,” Soft Computing, vol. 23, no. 8,
pp. 2499–2507, 2019.
[27] A. Debnath, H. Basumatary, M. Dhar, M. K. Debbarma, and
B. K. Bhattacharyya, “Fuzzy logic-based VANET routing
method to increase the QoS by considering the dynamic
nature of vehicles,” Computing, vol. 103, no. 7, pp. 1391–
1415, 2021.