Dynamic routing discovery scheme for high mobility in mobile ad hoc wireless networks

International Journal of Electrical and Computer Engineering (IJECE)
Vol. 10, No. 4, August 2020, pp. 3702~3714
ISSN: 2088-8708, DOI: 10.11591/ijece.v10i4.pp3702-3714  3702
Journal homepage: http://ijece.iaescore.com/index.php/IJECE
Dynamic routing discovery scheme for high mobility in mobile
ad hoc wireless networks
Haider Alani1
, Maha Abdelhaq2
, Raed Alsaqour3
1
IT Center and System, Iraqi Ministry of Electricity, Baghdad, Iraq
2
Department of Information Technology, College of Computer and Information Sciences,
Princess Nourah bint Abdulrahman University, Saudi Arabia
3
Department of Information Technology, College of Computing and Informatics,
Saudi Electronic University, Saudi Arabia
Article Info ABSTRACT
Article history:
Received Dec 18, 2019
Revised Jan 29, 2020
Accepted Feb 1, 2020
An innovative technology that is widely used in many applications is
the Mobile Ad-hoc Network (MANET). Discovery and maintenance of
routes at MANET are important issues. Within MANET, broadcasting is
used to discover a path within on-demand routing protocols. Establishing and
maintaining a route periodically among the nodes is the challenge that
requires the transmitting of control packets across a network. This state leads
to the issue of broadcasting storms. Broadcasting control packets increase
control packets overhead and decrease network performance. In this paper,
we proposed a scheme called AODV-Velocity and Dynamic (AODV-VD)
for effective broadcast control packets. The routing protocol for the ad-hoc
on-demand distance victor (AODV) is used to implement the proposed
AODV-VD scheme. AODV-VD scheme reduces both the excessive route
discovery control packets and network overhead. Network simulator version
2.35 (NS2.35) was used to compare the proposed AODV-VD scheme to
the AODV routing protocol in terms of end-to-end latency, average
throughput, packet transmission ratio and overhead ratio.
Keywords:
AODV
Link breakage
MANET
Mobility
Routing discovery
Copyright © 2020 Institute of Advanced Engineering and Science.
All rights reserved.
Corresponding Author:
Maha Abdelhaq,
Department of Information Technology,
College of Computer and Information Sciences,
Princess Nourah bint Abdulrahman University,
84428 Riyadh, Saudi Arabia.
Email: MSAbdelhaq@pnu.edu.sa, maha.ukm@gmail.com
1. INTRODUCTION
Mobile Ad-hoc Network (MANET) is a mobile, infrastructure-free, self-configuring, interactive and
multi-hop network [1]. MANET is an emergency network with a range of mobile nodes. The nodes in
MANET move randomly within the network due to rapid and frequent topology changes in the MANET [2].
Such mobile nodes can function simultaneously as hosts and routers since they can travel anywhere in
the MANET [3]. In MANET, it is the responsibility of the routing protocol to develop a route between nodes
to transmit packets, which selects the optimum and maintains the route. When a source node wants to
transmit data, the routing protocol constructs a path between a source node and a destination node. MANET
routing protocols are divided into three classifications: positive, reactive and hybrid [4].
In MANET routing protocols, a broadcasting scheme is necessary when packets are sent between
mobile nodes to maintain network connectivity [5]. Broadcasting is usually defined as the process of
transmitting a packet to all nodes in the network from a source node. Broadcasting is more frequently used
in MANET when the source node broadcasts Route Request (RREQ) packets to check for a route to
a destination node, usually in path discovery. Broadcasting is often used in route maintenance when nodes

Int J Elec & Comp Eng ISSN: 2088-8708 
Dynamic routing discovery scheme for high mobility in mobile ad hoc wireless networks (Haider Alani)
3703
exchange Hello packets for gathering neighboring information. Intermediate nodes in MANET assist in
the broadcast process. Intermediate nodes are responsible for forwarding the packet to other nodes in
the network from the source node [6].
The flooding results in multiple redundant transmissions on the network, in which a node from
various nodes can get the payload of the same packet. This event is known as a broadcasting storm
problem [7]. The issue of broadcasting storms leads to regular contention and packet collisions, which
increases network overhead traffic. In the route maintenance phase, the broadcast storm problem occurs,
during which routes are updated by activating new route discovery requests to replace broken routes. In this
paper, we suggested a scheme called AODV–Velocity and Dynamic (AODV-VD) for efficient broadcast
control packets. The results obtained from evaluating the AODV-VD scheme are very positive and show
the effectiveness of the proposed scheme in mitigating the problem of the broadcast storm.
The remainder of the paper is structured according to this. We provide a description of the research
background and work related to it in Section 2. Section 3 describes the new scheme AODV-VD. Section 4
describes the environmental simulation and the performance metric. Section 5 addresses the findings and
evaluations and Section 6 summarizes our research and our future work.
2. BACKGROUND AND RELATED WORK
2.1. Ad-hoc on-demand distance vector routing protocol
AODV, that Perkins and Royer built [8], is a reactive routing protocol, and its multi-hop
routing protocol and discovery react on demand. In the AODV routing protocol, the strengths of
Destination Sequenced Distance Vector Routing (DSDV) and Dynamic Source Routing (DSR) protocols are
combined [8]. The source node broadcasts a route request (RREQ) packet throughout the MANET nodes in
the route discovery process of the AODV routing protocol on MANET and sets a timer to wait for the reply.
The RREQ packet contains routing information, including the IP address of originator, ID of broadcast and
sequence number of destination.
Each intermediate node receives the RREQ packet and holds two operations running the reverse
path to the source node. First, the intermediate node checks if it has previously received the RREQ packet
with the same originator IP address and broadcast ID, and then determine whether the RREQ packet should
be rejected or accepted. Second, if the RREQ packet is accepted the intermediate node will check
the destination sequence number stored in its routing table. The intermediate node uni-cast the Route Reply
(RREP) packet to the source node if the number of the sequence is greater than or equal to the number
contained in the RREQ. If no intermediate node has a sufficiently fresh path to the destination node
(new destination sequence number), then the RREQ packet must continue navigating until it reaches
the destination node. Figure 1 shows the source node (S) that transmits RREQ packets over the network to
its neighboring nodes until the RREQ packet reaches the destination node (D). Figure 2 also shows
the destination node (D) which replies to the source node with an RREP.
Figure 1. AODV broadcasts RREQ packet Figure 2. AODV replies RREP packet
In AODV, every node in the network also sends Hello packet periodically to keep its routing table at
the one-hop neighbor. A Hello packet is used to determine if the neighboring connection is still alive.
The node sends Hello packet with a time interval called hello-interval to its neighbor node to detect broken
links between the nodes. Every node sends Hello packets to its neighbors and they receive their
acknowledgment. If a node sends Hello packets to a neighbor twice and has not received a message of
acknowledgment for it, then the node initiates the broken link process.
2.2. Related work
In [9], the authors proposed a routing protocol called Mobility Aware and Dual-Phase Ad-hoc
On-demand Distance Vector with Adaptive Hello Messages (MA-DP-AODV-AHM). It focuses primarily on

 ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 10, No. 4, August 2020 : 3702 - 3714
3704
developing routes that consider the speeds and direction of motion of vehicles with regard to source vehicles,
thereby providing more efficient routes and reducing route breaks. The authors have suggested an adaptive
packet notification system, which is directly connected to the regular hello messages solution, with
the speeds of the vehicles, resulting in a substantial reduction in the amount of overhead control and network
congestion. The protocol was evaluated using QualNet simulator in terms of overhead control, end-to-end
delay and energy consumption. Results of the simulation summarized that MA-DP-AODV-AHM effectively
contributes to mitigating network instability by producing reliable, secure routes and reducing connection
failures. It has also shown its superiority over AODV and MDA-AODV.
In [10], the authors introduced a Quality of Service (QoS) algorithm in ad hoc networks.
The proposed algorithm entitled Modified Light Weight distributed the QoS Algorithm, taking into account
the position and relative velocity of the nodes and estimating the importance of linking reliability, choosing
a path with longer duration, reducing link breaks and enhancing QoS. The authors measured the algorithm's
efficiency in packet delivery ratio, packet delivery latency, average end-to-end delay overhead routing and
packets lost. The simulation results show that the Modified Light Weight distributed QoS Algorithm
enhances the QoS parameters of QoS in ad hoc networks.
In [11], the authors proposed two predictive link availability techniques performed during
the control packet routing. The first technique referred to as zone-based estimation divides each node's area
of transmission into three areas, namely inner, middle, and outer, based on the signal strength obtained.
Here, nodes lying in the inner and outer zone are prohibited from transmitting the control packet as they have
extremely high and low signal strength received, respectively. Nodes in the middle region are only permitted
inside the locality for the control packet flow. The second technique is an improvement over zone-based
estimation, in which nodes are considered for route selection in the outer zone and are referred to as
segment-based estimation. It calculates the link availability ratio of the node for each adjacent link based on
the neighbor's current position, its adjacent segment, and the sector in which the neighbor is present within
the node's transmit region. Such information assists in the transfer of control packets within the outer region.
The simulation was performed over DSR and AODV protocols in terms of control overhead, bandwidth,
number of control packets, length of the route, and number of route errors generated within the network.
The results show that knowledge of the link status during the routing phase helps to find more stable routes
with reduced overhead control and thus increases the overall performance of the network.
In [12], a neighboring information-based broadcasting scheme is being introduced to minimize
latency for ad hoc wireless networks. In the scheme, few Hello messages are interchanged to gather
information about one-hop neighbors. The information collected is used to calculate the density of
the neighbor, the ratio, and the number of exposed neighbors with one-hop, on which the probability and
delay of retransmission are modified. The way the probability and delay of rebroadcasting are described in
the neighboring knowledge-based broadcast scheme decreases the overhead transmission and effectively
restrains the aggregation of traffic. After that, a method for the velocity-based delivery of data is proposed
and applied to a neighboring knowledge-based broadcast scheme to further reduce latency, creating
neighboring knowledge and speed-based broadcast scheme. It is specified that few higher-velocity nodes are
used to retransmit the incoming message with a greater probability. The simulation determines the schemes
efficiency under different network configurations. The results show that the new scheme outperforms current
broadcast schemes in terms of overhead, and in particular, in the average end-to-end delay. The proposed
scheme lowers the overhead by 88.4 percent relative to flooding and the average end-to-end delay by
a maximum of 88.9 percent.
In [13], the authors suggested a novel discovery scheme within MANET called the Dynamic
Probabilistic Route (DPR). The DPR scheme forwarded the packets with a dynamically determined
probability known as a forwarding probability (FP) to the neighboring node. The likelihood of forwarding is
dynamically determined to depend on two factors; the first is the density of neighboring local nodes and
the second is the total number of its neighbors protected by broadcast. Using the DPR scheme, network
performance is higher than other protocols.
In [14], the authors introduced a new probabilistic broadcasting scheme for MANETs that
overcomes the constraints of existing broadcasting schemes. Through detailed simulations, it is shown that
the newly proposed schemes in different operating conditions and scenarios outperform previous broadcast
schemes. Unlike previous works, the proposed broadcast scheme strategy is based on information from
the node velocity vector to adjust the retransmission probability and categorize the reliability of the nodes
accordingly. The authors applied this velocity vector and evaluated its efficiency with respect to various
important metrics such as link stability and overhead RREQ packets. The key benefit of this scheme is to
avoid the traditional AODV's route re-discovery process, especially at high mobility nodes.
A novel scheme for Channel Adaptive Probabilistic Broadcast (CAPB) is proposed in [15], which is
implemented in the AODV routing protocol to replace pure flood-based broadcast. Under the scheme,

3705
the probability of retransmitting RREQ packets is regulated by the current signal interference plus noise ratio
(SINR) and neighboring density of the nodes. Compared to the other three broadcasting schemes which
include AODV, fixed probabilistic scheme AODV-P, and the dynamic noise-dependent probabilistic scheme
(DNDP), the CAPB scheme performs better in terms of overhead routing, throughput and end-to-end delay.
Energy Efficient Neighbor Coverage protocol (EENCP) was proposed by [16] to route discovery in
MANET. Uncovered neighbor and neighborhood node density are two metric factors considered under
the EENCP protocol to reduce the forwarding of the RREQ packet. The forwarding of the RREQ packet by
the node relies on its battery life and neighborhood density, which forwards when there is adequate energy
and when the node density surpasses a particular threshold. The EENCP protocol can limit rebroadcasting
with packets more than AODV and Neighbor Coverage-based Probabilistic Rebroadcast (NCPR) protocols.
A new probabilistic vector velocity scheme was developed in [14] for route discovery in MANET. Based on
the AODV routing protocol, this scheme categorizes all mobile nodes into reliable nodes and unreliable
nodes in terms of sender and receiver node velocity, and assigns a high retransmission probability for reliable
nodes and a low retransmission probability for unreliable nodes. This type of scheme helps to discover
the steadiest and most efficient routes, thereby enhancing route discovery efficiency. This scheme
demonstrates the RREQ packet's superiority in terms of both overhead and connection stability. A novel
algorithm, Reliability Factor-based Routing Protocol (RFBRP), was introduced by [17] to discover a reliable
route in MANET. It is designed based on reactive routing protocols such as AODV and DSR. It is positioned
on the reliability factor (RF) between the intermediate nodes to select a dependable route. RF is a primary
metric that limits the routing failure and the number of route discovery requests, reducing the overhead costs
of each node in the route discovery process. In terms of overhead, delivery of a packet, and end-to-end delay,
the RFBRP algorithm performs better than AODV and semi-proactive AODV (SP-AODV) protocols.
Hanji and Shettar [18] introduced an Improved AODV (I-AODV) protocol in MANET.
The route discovery in the I-AODV protocol was based on the location and energy of the nodes in
the network. The source node selects the intermediate node located in the communication range, which has
less distance than the other intermediate nodes to the destination node and has energy greater than the other
intermediate nodes to increase the lifetime of the path. The I-AODV protocol provides efficient packet
transmission, eliminates overhead and improves the lifetime of the path that leads to a stable path compared
to the AODV protocol.
Kok, et al. [19] proposed an effective ad-hoc wireless broadcasting scheme, known as Improved
Partial Dominant Pruning (IPDP), based on Partial Dominant Pruning (PDP). It uses neighborhood
information to reduce redundancy of broadcasting and ensuring that all nodes in a network are reachable to
packets. To compare their efficiency in two sample sets, IPDP and PDP were built. The first sample in static
scenarios while the second set of data in mobile scenarios examines the schemes. The IPDP is very effective
in static scenarios in reducing the redundancy of broadcast while maintaining availability to all nodes in
a network. From the mobile scenario, all variants of IPDP performed at low to high node movement
speed close PDP in terms of reachability. A new AODV (NAODV) in MANET was suggested by [20] in
the routing protocol. It resolves the AODV routing protocol issues by accounting local congestion of
the network while selecting the route. Based on network congestion, the number of hops is identified while
selecting the next node to limit the survival time of forward and reverse routes, and reduce the number of
RREQ packet retransmissions, reducing overhead costs. Mohsin, et al. [21] examine the effects of the density
and mobility nodes of different maritime traffic model based on the capabilities of three MANET routing
protocols that comprise AODV, AOMDV and DSDV. Various node densities and mobility interpret
the traffic models that will be mainly found in the marine environment. The performance assessment of
the MANET protocols is judged in terms of packet delivery ratio.
Zohra and Rahman [22] introduced the dynamic probabilistic broadcast algorithm in MANET.
This algorithm estimates the rebroadcast probability by utilizing the forwarding probability of a node.
The algorithm introduced a logical model to identify the expected number of forwarding nodes that are
essential to accomplish a broadcast in a self-pruning algorithm. It employs neighbor knowledge to limit
the unnecessary retransmissions in MANET. It exhibits superior performance compared to the static
probabilistic algorithm and the probabilistic adjust broadcast algorithm. A dynamic partitioning scheme
(DPS) dependent on nodal vehicles was suggested by Rayeni, et al. [23]. The DPS model computes partition
sizing and quantities, after which it computes a size and broadcast timetable for each partition. DPS scheme
performs reliably and time-efficiently in an emergency broadcast program across multiple hops in either low
or high-density traffic states. This method suitably applies to varied scenarios featuring scarce, mid and
high-density traffic. Simulations confirm that DPS performs outstandingly with low latencies and
high reliability. Kimura, et al. [24] presented a density-aware probabilistic routing model featuring adaptive
probabilistic forwarding operation, wherein the forwarding probabilities decline as messages are redirected.
Such density-sensitive probabilistic routing models control the speeds of circulating copies of messages in

 ISSN: 2088-8708
3706
high-density areas. The results of the performance assessment present the adaptive probabilistic forwarding
model in a high-density nodal region. The On-demand, Tree-based Routing Protocol (OTRP) was suggested
by Al Aamri, et al. [25] for MANET. The OTRP with AODV protocol allowed it to scale up and also lessen
overheads in MANET routing discovery, by pairing the concept of hop-to-hop routed transmissions with
the routing discovery algorithms known as Tree-based Optimized Flooding (TOF). The primary elements
which govern OTRP performance were hypothetically examined and assessed in terms of nodal branch
counts and locations, and counts of RREQ routing expirations. At various levels of nodal density and
mobility, OTRP was contrasted to AODV, OLSR, and dynamic MANET On-demand (DYMO) routing
protocols. The findings indicate that OTRP mechanisms perform outstandingly with much-lessened network
overhead and declining traffic loads.
Zhang, et al. [26] suggested Neighboring Coverage-derived Probabilistic Rebroadcasting (NCPR)
protocols for MANET. The method dynamically tallies rebroadcasting delays and rebroadcasting
probabilities. Rebroadcasting delays are used to resolve the forwarding orders and to efficiently utilize
neighboring coverage information to derive further and more exact coverage ratios. Connection factors are
employed in NCPR to determine rebroadcasting probabilities, to maintain networked connectivity, and to
decrease incidences of redundantly retransmitted traffic. Research on routing requests in routing discovery
methods contrasted NCPR with Dynamically Probabilistic Route discovery protocols (DPR), and AODV
protocols. NCPR generates less traffic rebroadcasts than flooding protocols as a consequence of fewer
redundantly rebroadcast traffic, which decreases incidences of packet collisions and network contention,
raises the proportion of packet deliveries, and diminishes average end-to-end latencies.
3. PROPOSED AODV VELOCITY AND DYNAMIC (AODV-VD) SCHEME
The design and implementation of the proposed AODV-VD algorithm is based on the literature
proposed Advance Velocity Aware Probabilistic (AVAP)-AODV scheme by [14], AODV-VD scheme makes
use of probabilistic velocity vector to initiate route discovery. Each mobile node is classified as either
a Reliable Node (RN) or an UnReliable Node (URN). This classification is based on the sender and receiver
nodes’ velocity vector. Reliable nodes are assigned a high rebroadcast probability value, while unreliable
nodes are given a low rebroadcast probability value. The scheme proposed by Kanakaris, et al. [27],
is considered a dynamic density-driven route request forwarding scheme. Moreover, this made it, possible for
each node to use its neighborhood density as a basis for forwarding the RREQ message. The neighborhood’s
node density plays a vital role in determining the rebroadcast probability [15].
The proposed AODV-VD scheme transmits an RREQ packet while discovering a route between
the nodes, AODV-VD scheme uses the node’s dynamic density and velocity vector. It aims to reduce
the unnecessary broadcasting of route control packets and identify the reliable intermediate nodes existing
between the source and destination nodes. AODV-VD scheme makes use of several adjustable parameters to
determine whether retransmission is permitted or forbidden. The information obtained from the node density
within the neighborhood can be utilized to rebroadcast RREQ, selectively. If the neighbourhood has many
nodes, then the aim would be to decrease the probability of retransmission to avoid possible redundancy.
In this case, nodes that are only one hop away are defined as being “adjacent”. Therefore, if a node sends an
RREQ packet, retransmission can only be performed by a subset of nodes in its neighborhood. The number
of nodes allowed to perform retransmission can vary dynamically depending on the neighborhood node.
Table 1 shows the parameters used in the AODV-VD scheme. The AODV-VD scheme flowchart can be seen
in Figure 3 When any node (Fi = 1,2,3,..,n) receives the RREQ packet, the proposed AODV-VD scheme can
be performed and the packet is processed as follows Figure 3.
Table 1. AODV-VD parameters specification
Parameters Description
n total number of nodes presented in the network
Fi set of nodes that have received the RREQ message
i range from 1 to n number of nodes ƞ
βi number of nodes adjacent to node i
Pi packet forwarding probability based on βi
d node adjacency
Cf control factor between (0,1) adjusted using Pi
R random number in the range [0, 100]
θth predefined angle threshold
θ Cosine angle between transmitter and receiver to
decide if the receiver is RN or URN

3707
Figure 3. AODV-VD scheme flowchart
1. When a source node (S) tries to discover a new route, its velocity vector is placed in the RREQ packet
header.
2. When this RREQ packet is obtained by any node receiver (R) within the broadcast range of S, the cosine
angle θ is calculated. It then makes a rebroadcast decision using the following process:
- If the node’s cosine angle value is θ > θth, then the receiver node is categorized as a URN. It then
assigns a low rebroadcast probability.
- If the node’s cosine angle value is θ < θth, then the receiver node is categorized as RN. Then go to
STEP 3.
3. If βi ≤ d then
Forward the RREQ packet
else
Calculate forwarding probability Pi at node Fi
Pi =
100
βi
∗ (d ∗ Cf ; For 0<Cf ≤1
4. Generation of a Random Number (R) between (0,100)
5. If R < Pi then
Forward the RREQ packet
else
Ignore and Drop the RREQ packet
6. End

 ISSN: 2088-8708
3708
The proposed AODV-VD scheme for MANET aims to improve the network performance and
resolve the issue of the frequent link breakage. The scheme selects the reliable node during the route
discovery process to avoid the link break and eliminate redundant retransmission to achieve the lowest value
of congestion, therefore, reduce the overhead in the network [28].
4. SIMULATION ENVIRONMENT
We used NS-2.35 to measure the AODV-VD scheme's performance. The nodes are moving
according to the random waypoint mobility model. In the random waypoint mobility model, nodes move
freely and randomly without border restriction in this mobility model. CBR traffic is generated by
the application layer at the node. For every node, the transmission range is 250 m. The pause time for
the node is constant at 0 to allow node movement at all times. Network scenario node speed was selected at
5, 10, 20, 30, 40, and 50 m / s, and all scenarios set the number of nodes at 50. All nodes are set in a 1000 m
square area of about 1000 m. Table 2 displays the parameters in the simulation.
Table 2. Simulation parameters
Parameters Value Unit
Simulation time 300 S
Network area size 1000 × 1000 M
Number of nodes 20-100 Nodes
Data packet size 512 Bytes
Bandwidth 2 Mbps
Nodes speed 5 – 50 m/s
Pause time 0 S
Data traffic CBR
Performance metrics are used in the comparative study to assess the performance of the proposed
AODV-VD scheme. The following criteria are used for evaluating the proposed AODV-VD scheme
in scenarios.
a. Average end-to-end (E2E) delay
The average end-to-end delay metric is the average time it takes to successfully relay the data packet
from the source through the network to destination. This delay includes many smaller network delays,
including all potential delays caused by buffering latency during route discovery, queuing at router interface
queue, delays in MAC retransmission, propagation and transmission time. The average data packet delay for
the E2E can be determined using the following formula:
𝐴𝑣𝑔 𝐸2𝐸 𝑑𝑒𝑙𝑎𝑦 =
∑ (𝑅 𝑖−𝑆 𝑖)𝑛
𝑖−1
𝑛
(1)
where Ri is the total packet received, Si is the total packet sent, and n is the number of data packets.
b. Average throughput (Avg-Throughput)
The average throughput metric is the average of received successful data packets to the total
simulation time period. The total throughput is expressed in kilobits per second (kbps), and measures the
routing protocol's efficiency and reliability in receiving data packets through destinations. The formula used
to calculate the average throughput is as follows:
𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑇ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 =
∑ 𝑃𝑎𝑐𝑘𝑒𝑡𝑠 𝑟𝑒𝑐𝑒𝑖𝑣𝑒𝑑 𝑏𝑦 𝑑𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜𝑛𝑠
𝑠𝑡𝑜𝑝 𝑡𝑖𝑚𝑒−𝑠𝑡𝑎𝑟𝑡 𝑡𝑖𝑚𝑒
∗
8
1000
(2)
c. Packet delivery ratio (PDR)
The PDR metric shows the total number of received data packets by destinations divided by the total
number of data packets sent by the sources. This metric presents how a protocol successfully delivers packets
from the source to the destination. A high packet delivery ratio indicates good results, which represent the
wholeness and correctness of the routing protocol. The packet delivery ratio is computed using the following
formula:
𝑃𝐷𝑅 =
∑ 𝑝𝑎𝑐𝑘𝑒𝑡𝑠 𝑟𝑒𝑐𝑒𝑖𝑣𝑒𝑑 𝑏𝑦 𝑑𝑒𝑠𝑡𝑖𝑛𝑎𝑡𝑖𝑜𝑛𝑠
∑ 𝑝𝑎𝑐𝑘𝑒𝑡𝑠 𝑠𝑒𝑛𝑡 𝑏𝑦 𝑠𝑜𝑢𝑟𝑐𝑒𝑠
∗ 100 (3)

3709
d. Packet overhead ratio (POR)
Routing overhead ratio metric is the ratio of the total number of routing packets sent to the total
number of routing packets sent and the data packets sent. This metric gives an idea about the extra bandwidth
the overhead uses to handle data traffic. The overhead routing is calculated using formulae below:
𝑅𝑂𝑅 =
𝑁𝑜 𝑜𝑓 𝑟𝑜𝑢𝑡𝑖𝑛𝑔 𝑝𝑎𝑐𝑘𝑒𝑡𝑠
𝑁𝑜 𝑜𝑓 𝑟𝑜𝑢𝑡𝑖𝑛𝑔 𝑝𝑎𝑐𝑘𝑒𝑡𝑠+𝑁𝑜 𝑜𝑓 𝑑𝑎𝑡𝑎 𝑝𝑎𝑐𝑘𝑒𝑡𝑠 𝑠𝑒𝑛𝑡
∗ 100 (4)
e. Average of collision rate (ACR)
The ACR signifies the total number of RREQ packets that have been dropped and which did not
successfully arrive at the nodes during the simulation period. The formula below shows how the ACR is
calculated:
𝐴𝐶𝑅 (𝑝𝑎𝑐𝑘𝑒𝑡/𝑠𝑒𝑐) =
∑ 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑅𝑅𝐸𝑄 𝑝𝑎𝑐𝑘𝑒𝑡 𝑓𝑎𝑖𝑙𝑒𝑑
𝑠𝑖𝑚𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒
(5)
f. Link stability
Link stability pertains to the total number of broken routes that are found during data transmission
between nodes during the simulation time. Route break measurements are done despite varying network
densities and node speeds. Calculation of the link stability follows the following formula:
Link stability (link/route) = total number of route breakage (6)
5. RESULTS AND DISCUSSIONS
5.1. Results based on the number of nodes
The network density is considered to be a significant and decisive parameter. The number of nodes
under different network densities was set to 20, 40, 60, 80, and 100 during the simulation, and for all
the network parameters, every node was set to a maximum speed of 20 m/s. Figure 4 shows the average
throughput versus the number of nodes. Figure 4 shows there is a small decrease in the average throughput of
the proposed AODV-VD scheme when the number of nodes is 40 compared to the modified AVAP-AODV
and the original AODV protocols. In a dense network, the network becomes more stable. Therefore, there is
an increase in the average throughput in a dense network and convergent ratio in different protocols. Figure 4
also shows that the AODV-VD has experienced a standard increase in average throughput with the decrease
in intensity of dropping nodes. This can be due to the random deployment of nodes that change following its
number in a network.
Figure 5 demonstrates the varying average delay of E2E by contrasting the current AODV-VD
scheme with the initial AODV protocols and AVAP-AODV. The figure shows a considerable difference
between those protocols in the average E2E delay. Compared to the original AODV and AVAP-AODV
protocols, the average E2E latency in the proposed AODV-VD is reduced. For example, in a dense network
with 100 nodes, the average E2E delay for AODV, AVAP-AODV, and AODV-VD is 0.11898s, 0.079727s,
and 0.0640505s, respectively. This might be the cause behind the decrease in packet drop over the network.
This occurs because the dense network, the routes between source and destination nodes is a stable route.
And the stable route yields to lesser route breakages and lesser re-establishment that take less delay time and
an increase in the average throughput, as observed in Figure 4.
The POR results shown in Figure 6 imply that the proposed AODV-VD scheme has enabled
a reduction of POR necessary to allocate the routes in various sources and destinations, followed by
the AVAP-AODV and original AODV protocols. The control factor in the feeding network with prior
knowledge of the node adjacency dropped in different nodes has played a significant role in the reduction of
the POR. As observed in Figure 6 in the high-density area, there is an increase in POR because of having
many routes leading to the generation of more control packets. AODV-VD technique, in the proposed
scheme, generates fewer control packets.
Figure 7 illustrates the effects of the different protocols on the ACR comparison. The results
indicate that when moving from a low-density region to a high-density area, the ACR increases dramatically.
This is because as the number of nodes increases, so does the number of possible forwarded nodes. When
the number of nodes increased progressively, the proposed AODV-VD scheme led to less collision.
The result has established the efficiency of dynamic density-driven route requests to decrease the collision by
supporting a peer-to-peer connection between terminals. But AVAP-AODV and the original AODV
appeared to pose approximately similar ACR, which is greater than the proposed AODV-VD. Additionally,
allowing the lowest amount of ACR in the network would result in less power consumption.

 ISSN: 2088-8708
3710
Figure 8 shows that the proposed AODV-VD led to a reduced number of link breakages compared
to the AVAP-AODV and the AODV protocols. The number of broken links decreases when the number of
nodes increases because the network has the propensity to be stable in a dense network. As observed in
Figure 8, the AODV-VD scheme has lesser broken links in different nodes densities, due to the technique
used to select the reliable neighbor node. This study clearly shows that offering a protocol to reduce
the probability of route breakage would enable enhancement of the route efficiency, which helps to consume
less power and generates fewer control packets used in the network.
Figure 4. Avg throughput vs number of nodes Figure 5. Avg E2E delay vs number of nodes
Figure 6. POR vs number of nodes Figure 7. ACR vs number of nodes
Figure 8. Link stability vs the number of nodes

3711
5.2. Results based on the speed of nodes
Figure 9 shows the varying average throughput by comparing the proposed AODV-VD scheme with
the original AODV and the AVAP-AODV protocols. Figure 9 shows that the proposed scheme has
appreciably outpaced both AVAP-AODV and the earlier AODV, as evidenced in the data, particularly as
node speed ratios rise progressively. This is due to the proposed scheme selecting the reliable node and lesser
controls packet leading to the transfer of more data packets.
Figure 10 illustrates the average E2E delay of packet data transfers for the routing discovery
protocol, according to the various node speeds. As node speed increases, the average E2E delay of packet
data transfers increases proportionately. This occurs because as node speed increases along the routes
between source and destination nodes, recurrent breakages and re-establishment in the link occur. Results for
the AODV-VD scheme show that it yields considerably fewer delays in comparison to AVAP-AODV and
the earlier AODV.
Figure 9. Avg throughput vs speed of nodes Figure 10. Avg E2E delay vs speed of nodes
Figure 11 shows the three protocols in comparison according to the POR and the different speeds.
The acquired results indicated a significant disparity in POR, showing its lowest peaks in those cases where
the proposed AODV-VD scheme was running at different speed ranges. Compared with that of
AVAP-AODV and the earlier AODV, the new scheme provided fewer POR. This is due to the AODV-VD
scheme's ability to control local area densities while RREQ is selectively rebroadcast.
Figure 12 shows comparative results for the ACR of packet transfers according to the node speeds.
The results acquired reveal that the earlier AODV protocol generated higher rates of collisions, followed by
the AVAP-AODV protocol. The least value for ACR was recorded when the proposed AODV-VD scheme
was executed at various speeds. This is attributable to the efficiency of the AODV-VD mechanism in
determining the state for decreasing the nodes in the neighborhood, especially in the event of high probability
conditions. The lower number of collisions in the network helps to consume less power.
Figure 11. POR vs speed of nodes Figure 12. ACR vs speed of nodes

 ISSN: 2088-8708
3712
Figure 13 illustrates the comparative results for linkage stability, which represents the number of
route breakages according to the various node speeds, which ensued in the course of the simulations. As node
speeds increase, the number of route breakages correspondingly increases, since, during high-speed
operation, nodes run paths to route breakage during their occurrences. Notably, the proposed AODV-VD
scheme generates fewer instances of route breakage in comparison to the two other protocols.
Figure 13. Link stability vs speed of nodes
6. CONCLUSION AND FUTURE WORK
This paper is focused on proposing the newer dynamically probabilistic route discovery scheme for
MANET. The scheme aims to improve network performance and resolve the problem of frequent link
breakage. The scheme selects the reliable node of the route discovery process to avoid the link break and
eliminate redundant retransmission to achieve the lowest value of congestion then, therefore, reduce
the overhead in the network. AODV-VD scheme initializes probabilistic density values when nodes receive
RREQ packets corresponding to globalized and localized information on node neighbor densities.
Exponential math functions are utilized in the technique to resolve probability values for every node,
dynamically. Thus, the scheme can be useful for reducing retransmission activity involving redundant
packets. The proposed AODV-VD scheme was compared with the original AODV and AVAP-AODV
protocols. The protocols were compared based on their performance to two sets of disparate node speeds and
various numbers of nodes. The first set evaluates the protocols in various numbers of nodes, the number of
nodes under different network densities was set to 20, 40, 60, 80, and 100 during the simulation, and for all
the network parameters, every node was set to a maximum speed of 20 m/s. The second set evaluates
the protocols according to the corresponding nodal speeds of 5 m/s, 10 m/s, 20 m/s, 30 m/s, 40 m/s, and
50 m/s during the simulation, the network density was set 50 nodes. The simulation outcomes indicate that
the AODV-VD scheme performs outstandingly. In this research, the proposed scheme performance was
assessed with the aid of current AODV protocols under NS2.35. It would be appropriate for additional
reactive protocol types such as DSR to be included in an operational assessment of the proposed schemes
performance. Additionally, the performance of the proposed scheme can be analysed with the use of
VANET. Finally, the random way mobility model was used to evaluate the performance of the AODV-VD
scheme. Several different mobility models can be used to evaluate the performance of the AODV-VD
scheme in a different model that simulates another environment.
ACKNOWLEDGMENT
This research was funded by the Deanship of Scientific Research at Princess Nourah bint
Abdulrahman University through the Fast-track Research Funding Program.
REFERENCES
[1] C. Reddy, "Node activity based trust and reputation estimation approach for secure and QoS routing in MANET,"
International Journal of Electrical & Computer Engineering (IJECE), vol. 9, no. 6, pp. 5340-5350, 2019.

3713
[2] M.S. Abdelhaq, R.A. Alsaqour, M. Al-Hubaishi, T. Alahdal, and M. Uddin, "The Impact of Resource Consumption
Attack on Mobile Ad-hoc Network Routing," IJ Network Security, vol. 16, pp. 376-381, 2014.
[3] A. Srivastava, A. Mishra, B. Upadhyay, and A.K. Yadav, "Survey and overview of Mobile Ad-Hoc Network
routing protocols," in 2014 International Conference on Advances in Engineering and Technology Research
(ICAETR), pp. 1-6, 2014.
[4] R. Mylsamy and S. Sankaranarayanan, "A Preference-Based Protocol for Trust and Head Selection for Cluster-
Based MANET," Wireless Personal Communications, vol. 86, pp. 1611-1627, 2016.
[5] M. B. Yassein, M. B. Khalaf, and A. Y. Al-Dubai, "A new probabilistic broadcasting scheme for mobile ad hoc on-
demand distance vector (AODV) routed networks," The Journal of Supercomputing, vol. 53, pp. 196-211, 2010.
[6] D. K. Panda and R. K. Dash, "Reliability Evaluation and Analysis of Mobile Ad Hoc Networks," International
Journal of Electrical & Computer Engineering (IJECE), vol. 7, no. 1, pp. 479-485, 2017.
[7] Y.-C. Tseng, S.-Y. Ni, Y.-S. Chen, and J.-P. Sheu, "The broadcast storm problem in a mobile ad hoc network,"
Wireless networks, vol. 8, pp. 153-167, 2002.
[8] C. Perkins, E. Belding-Royer, S. Das, "Ad hoc on-demand distance vector (AODV) routing," RFC Editor, 2003.
[9] K. A. Darabkh, M. S. Judeh, H. B. Salameh, and S. Althunibat, "Mobility aware and dual phase AODV protocol
with adaptive hello messages over vehicular ad hoc networks," AEU-International Journal of Electronics and
Communications, vol. 94, pp. 277-292, 2018.
[10] S. H. Hosseini Nazhad Ghazani, "Modified Light Weight distributed QoS Algorithm for wide area ad hoc
networks," International Journal of Communication Systems, vol. 31, pp. e3562, 2018.
[11] S. R. Malwe, N. Taneja, and G. Biswas, "Enhancement of DSR and AODV Protocols Using Link Availability
Prediction," Wireless Personal Communications, vol. 97, pp. 4451-4466, 2017.
[12] D. Lu and S. Dong, "A neighbor knowledge and velocity-based broadcast scheme for wireless ad hoc networks,"
International Journal of Distributed Sensor Networks, vol. 13, no. 11, pp. 1-14, 2017.
[13] K. Shanmugam, K. Subburathinam, and A. Velayuthampalayam Palanisamy, "A Dynamic Probabilistic Based
Broadcasting Scheme for MANETs," The Scientific World Journal, vol. 2016, 2016.
[14] M. B. Khalaf, A. Y. Al-Dubai, and G. Min, "New efficient velocity-aware probabilistic route discovery schemes for
high mobility Ad hoc networks," Journal of Computer and System Sciences, vol. 81, pp. 97-109, 2015.
[15] H. Y. Adarbah, S. Ahmad, B. Arafeh, and A. Duffy, "Efficient broadcasting for route discovery in mobile ad-hoc
networks," in Proceedings of the International Symposium on Performance Evaluation of Computer and
Telecommunication Systems, pp. 1-7, 2015.
[16] R. R. Ravi and V. Jayanthi, "Energy Efficient Neighbor Coverage Protocol for Reducing Rebroadcast in MANET,"
Procedia Computer Science, vol. 47, pp. 417-423, 2015.
[17] S. M. Khan, R. Nilavalan, and A. F. Sallama, "A Novel Approach for Reliable Route Discovery in Mobile Ad-Hoc
Network," Wireless Personal Communications, vol. 83, pp. 1519-1529, 2015.
[18] B. R. Hanji and R. Shettar, "Improved AODV with restricted route discovery area," in 2015 International
Conference on Computer Communication and Informatics (ICCCI), pp. 1-6, 2015.
[19] G.-X. Kok, C.-O. Chow, and H. Ishii, "Reducing Broadcast Redundancy in Wireless Ad-Hoc Networks with
Implicit Coordination Among Forwarding Nodes," Wireless Personal Communications, vol. 81, pp. 253-278, 2015.
[20] Z. Zhao, M. Gao, M. Hou, and N. Zhang, "Design of Redundant New Ad-Hoc On-Demand Distance Vector
(NAODV) Routing Protocol Based on Congestion and Survival Control," Wireless Personal Communications,
vol. 85, pp. 2657-2668, 2015.
[21] R. J. Mohsin, J. Woods, and M. Q. Shawkat, "Density and mobility impact on MANET routing protocols in
a maritime environment," in Science and Information Conference (SAI) 2015, pp. 1046-1051, 2015.
[22] F. T. Zohra and A. Rahman, "Mathematical analysis of self-pruning and a new dynamic probabilistic broadcast for
MANETs," in 2015 International Conference on Networking Systems and Security (NSysS), pp. 1-9, 2015.
[23] M.S. Rayeni, A. Hafid, and P.K. Sahu, "Dynamic spatial partition density-based emergency message dissemination
in VANETs," Vehicular Communications, vol. 2, pp. 208-222, 2015.
[24] T. Kimura, T. Jonouchi, T. Matsuda, and T. Takine, "Density-aware store-carry-forward routing with adaptive
forwarding probability control," in 2015 IEEE International Conference on Consumer Electronics-Taiwan
(ICCE-TW), pp. 474-475, 2015.
[25] H. AlAamri, M. Abolhasan, D. Franklin, and J. Lipman, "Optimised relay selection for route discovery in reactive
routing," Ad Hoc Networks, vol. 11, pp. 70-88, 2013.
[26] X.M. Zhang, E.B. Wang, J.J. Xia, and D.K. Sung, "A neighbor coverage-based probabilistic rebroadcast for
reducing routing overhead in mobile ad hoc networks," Mobile Computing, IEEE Transactions on, vol. 12,
pp. 424-433, 2013.
[27] V. Kanakaris, D. Ndzi, and K. Ovaliadis, "Improving AODV performance using dynamic density driven route
request forwarding," arXiv preprint arXi, 1107.3630, 2011.
[28] S. Sharma, D. Jindal, and R. Agarwal, "Analysing Mobile Random Early Detection for Congestion Control in
Mobile Ad-hoc Network," International Journal of Electrical & Computer Engineering (IJECE), vol. 8, no. 3,
pp. 1305-1314, 2018.

 ISSN: 2088-8708
3714
BIOGRAPHIES OF AUTHORS
Haider Alani is a Programmer at the IT Center in the Ministry of Electricity of the Republic of
Iraq, Baghdad, Iraq. He received his BSc degree in Computer Science from Alrafidain
University College (RUC) Iraq, in 2004. MSc degree in Computer Science from Universiti
Kebangsaan Malaysia (UKM) in 2016. His research interests include mobile ad hoc network
(MANET) and vehicular ad hoc network (VANET).
Maha Abdelhaq is an Assistant Professor at College of Computer and Information Sciences,
Princess Nourah bint Abdulrahman University, Saudi Arabia. She received her PhD from the
Faculty of Information Science and Technology, National University of Malaysia, Malaysia in
2014; MSc in securing wireless communications from the University of Jordan, Jordan in 2009;
BSc in Computer Science from the University of Jordan, Jordan in 2006. She is a member in
IEEE, ACM and International Association of Engineers (IAENG). Her research interests include
vehicular networks, MANET routing protocols, artificial immune systems, network security and
intelligent computational.
Raed Alsaqour is an Associate Professor at College of Computation and Informatics, Saudi
Electronic University, Riyadh Branch, Kingdom of Saudi Arabia. He received his PhD in
Wireless Communication Systems from National University of Malaysia, Malaysia in 2008;
MSc in Distributed Systems from University Putra Malaysia, Malaysia in 2003; BSc in
Computer Science from Mutah University, Jordan in 1997. He is a member in IEEE, ACM and
IAENG. His research interests include wireless ad hoc and vehicular networks, routing protocols,
simulation and network performance evaluation.

Dynamic routing discovery scheme for high mobility in mobile ad hoc wireless networks

More Related Content

What's hot (20)

Similar to Dynamic routing discovery scheme for high mobility in mobile ad hoc wireless networks (20)

More from IJECEIAES (20)

Recently uploaded (20)

Dynamic routing discovery scheme for high mobility in mobile ad hoc wireless networks