Journal Pre-proof

A Hybrid Secure Routing and Monitoring Mechanism in IoT-basedWireless Sensor Networks

Deebak B D , Fadi Al-Turjman

PII: S1570-8705(19)30505-0DOI: https://doi.org/10.1016/j.adhoc.2019.102022Reference: ADHOC 102022

To appear in: Ad Hoc Networks

Received date: 29 May 2019Revised date: 5 September 2019Accepted date: 10 October 2019

Please cite this article as: Deebak B D , Fadi Al-Turjman , A Hybrid Secure Routing and Mon-itoring Mechanism in IoT-based Wireless Sensor Networks, Ad Hoc Networks (2019), doi:https://doi.org/10.1016/j.adhoc.2019.102022

This is a PDF file of an article that has undergone enhancements after acceptance, such as the additionof a cover page and metadata, and formatting for readability, but it is not yet the definitive version ofrecord. This version will undergo additional copyediting, typesetting and review before it is publishedin its final form, but we are providing this version to give early visibility of the article. Please note that,during the production process, errors may be discovered which could affect the content, and all legaldisclaimers that apply to the journal pertain.

© 2019 Published by Elsevier B.V.

1

A Hybrid Secure Routing and Monitoring Mechanism in IoT-based Wireless Sensor

Networks

Deebak B D, Fadi Al-Turjman

Vellore Institute of Technology, School of Computer Science and Engineering, Vellore

Campus

Professor Computer Engineering Dept. Antalya Bilim University, Antalya, Turkey

Abstract – Internet of Things (IoT) has advanced its pervasiveness across the globe for the

development of smart networks. It is aimed to deploy network edge that enables smart

services and computation for the IoT devices. In addition, this deployment would not only

improve the user experience but also provide service resiliency in case of any catastrophes. In

IoT applications, the edge computing exploits distributed architecture and closeness of end-

users to provide faster response and better quality of service. However, the security concern

is majorly addressed to resist the vulnerability of attacks (VoA). Existing methodologies deal

only with static wireless sensor web to deduce the intrusions in which the sensor nodes are

deployed in a uniform manner to retain the constancy. Since the sensor nodes are constantly

being in question through different transmission regions with several levels of velocities,

selection of sensor monitoring nodes or guard nodes has become a challenging job in recent

research. In addition, the adversaries are also moving from one location to another to explore

its specific chores around the network. Thus, to provide flexible security, we propose a secure

routing and monitoring protocol with multi-variant tuples using Two-Fish (TF) symmetric

key approach to discover and prevent the adversaries in the global sensor network. The

proposed approach is designed on the basis of the Authentication and Encryption Model

(ATE). Using Eligibility Weight Function (EWF), the sensor guard nodes are selected and it

is hidden with the help of complex symmetric key approach. A secure hybrid routing protocol

is chosen to be built by inheriting the properties of both Multipath Optimized Link State

Routing (OLSR) and Ad hoc On-Demand Multipath Distance Vector (AOMDV) protocols.

The result of the proposed approach is shown that it has a high percentage of monitoring

nodes in comparison with the existing routing schemes. Moreover, the proposed routing

mechanism is resilient to multiple mobile adversaries; and hence it ensures multipath

delivery.

Keywords – Internet of Things, Wireless sensor networks; Symmetric key; Authentication

and encryption model; Optimized link state routing; Ad hoc on-demand multipath distance

vector; Multipath delivery

1. Introduction

IoT is widely accepted as a 3rd

industrial revolution that embeds the computing object to

send and receive the physical data over the Internet [1,2]. It is uprising at breathtaking-pace,

initiated with 2 billion physical objects in 2006 into 200 billion by 2020 [3] i.e. growth of

200%. IoT devices / sensors generally collect and observe the temporal/spatial information to

2

manage the real-time events addressing various challenges [4,5]. IoT applications are

becoming smarter for various applications namely education, finance, energy, healthcare,

transportation and smart cities [6]. Subsequently, academia, industry and individuals are

enduring to provide security and safety i.e. for IoT devices and networks. These factors

should chiefly be concerned to avoid data catastrophe to the IoT users. For example, a smart

home system can be monitored remotely by the cyber-attackers, and smart vehicle

communication can be seized to create a source of danger among the citizens. This

catastrophe condition is highly exposed to Internet-connective objects to affect the IoT

security systems and ecosystems of complex communication networks such as social

networks, application, websites and Robo networks i.e. botnet.

In contrast, cooperating a single communication channel or component can make the

IoT-based system powerless as a part or complete network access. Dyn cyberattack gathered

the connective device to install within smart cities and gathered them as botnets i.e. Zombie

Army through middleware known as Mirai in 2016. In addition to the vulnerabilities, the IoT

system is now evolving of attack vectors in terms of diversity and complexity. Therefore,

Wireless Sensor Networks (WSN) is considered as a set of resources that impels sensor nodes

to gather data from the environment; compute the collected outputs into a formatted data, and

transmit it to the destination terminal through the wireless medium. The source of input will

be the sensed information gathered from different types of sensors such as temperature,

pressure, magnitude, level and flow sensors and so on. The open nature of the wireless

medium makes the network weak and defenseless to protect its data from adversaries when

compared to the strongly built infrastructure wired networks.

WSN [7] provides valuable communication in the battle fields or defense-oriented

applications to recover out the lights, electromagnetic signals, chemical or biological vapors,

and the enemy presence or border violations. Providing security with optimized energy in

WSNs is a hard job when nodes are in motion. Because managing the localization of sensor

nodes and moving adversaries are the decisive one in terms of protective factors. There are

dissimilar cases of attacks created by the adversaries depends on their objectives or without

any motives [8]. Any wireless sensor node equipped with sufficient hardware and software

can act as an adversary to sense the wireless channel to grab the transmitting data in an

unauthorized way. In addition, the adversaries may try to change the natural behavior of the

normal sensor nodes and compromise it to violate the activities of the wireless sensor

network and this will make the sensor nodes to downhill on their performance, throughput,

and service [9].

3

To infer these vulnerable activities or attacks, Intrusion Detection System (IDS) is used

in practice. The IDS [10-12] is mainly used in wired networks with the deployment of

hardware systems between servers or nodes to monitor the network activities. In literature,

IDS based learning mechanisms have been considered for the evaluation of traditional

network systems i.e. not specifically for IoT systems [13-18]. Agarwal et al. [13] reviewed

various data mining techniques for network anomaly detection. Buczak et al. [14] explained

the data mining and machine learning methods to show the significance of cyber analytics in

the endurance of intrusion detection and prevention. These survey papers provided substantial

references to summarise the challenges of cyber securities.

Table 1 Important Notation Used

Notation Description

IoT Internet of Things

VoA Vulnerability of Attacks

TF Two-Fish

ATE Authentication and Encryption Model

EWF Eligibility Weight Function

OLSR Optimized Link State Routing (OLSR) and

AOMDV Ad hoc On-Demand Multipath Distance Vector

DSDV Destination Sequenced Distance Vector

TARCS Topology Change Aware-Based Routing Protocol

WSN Wireless Sensor Networks

IDS Intrusion Detection System

DIDS Distributed Intrusion Detection System

BS Base Station

CMS Concealed Monitor Set

QoS Quality of Services

DSR Dynamic Routing Protocol

MANET Mobile Ad hoc Network

VCG Vickrey, Clarke, and Groves

MPR Multipoint Relays

ANSN Advertised Neighbor Sequence Number

ANIA Advertised Neighbor Information Address

INITREQ Initial Request

INITRES Initial Response

MaRES Mask Request

MaREQ Mask Response

HMAC Hash-based Message Authentication Code

PSQN Packet Sequence Number

MAC Medium Access Control

RCP Routing Control Packets

PIT Protocol Interface Table

SRT Sandwiched Routing Table

CBR Constant Bit Rate

UDP User Datagram Protocol

4

DCF Distributed Coordination Function

KMP Knuth–Morris–Pratt

Even though, Fadlullah et al. [17] focused on deep learning mechanism to study the

traffic-control systems. Hodo et al. [18] presented taxonomy of deep and swallow networks

to survey on intrusion detection and prevention systems. In addition, Wang and Jones [15]

reviewed data mining, machine learning, deep learning, and bid-data to evaluate the criteria

such as data streaming, processing, reduction and feature characteristics. Mishra et al. [16]

compared and examined the limitation and constraints of machine learning techniques to

analyze intrusion detection. Table 1 shows the important notation used in this paper.

1.1. Research Motivation

The existing survey targets IDS and IoT to identity two basic state-of-the-arts such as an

overview of IoT, IDS, and classification of taxonomy [19,20]. In addition, they proposed a

detailed survey to compare the different detection and prevention system i.e. for IoT to

analyze the parameters such as detection approaches, validation strategies, and placement

schemes. Though, BenKhelifa et al. et al. [19] focused on the advancements of detection and

prevention practices in IoT. They surveyed the recent state-of-the-art approaches with a

special reference on IoT architecture. In random Ad hoc sensor networks, the sensor nodes

are presented in a distributed manner without any centralized equipment and it is widely

known to be a Distributed Intrusion Detection System (DIDS) [21]. Since then, several DIDS

schemes [22-24] have been presented to discover and forestall the attacks in random sensor

networks. The purpose of the schemes was to select watchdogs or monitor nodes to protect

the wireless sensor network. Hither, to detect the attacks, the safeguard nodes are selected

from neighbor nodes.

For data transmission and monitoring purpose, the sensor node has two selective

approaches, namely Hierarchical or Clustered and Flat or Random. In the cluster approach

[25,26], the sensor nodes are used to form as clusters; and thus it will elect its cluster head for

each cluster using any leader election algorithm. By then, the cluster head will act as a

monitor node to the local cluster and it will help to detect the availability of the unauthorized

traffics or intruder nodes in the cluster. Eventually, it will share the detection reports between

the monitor nodes or other cluster heads to deliver to a base station (BS). Then, it will start

the fault recovery process at the necessary links or nodes through the BS. In each round time

execution, the cluster head will be selected by the election scheme which will later create a

computation overhead to the sensor networks. Moreover, each cluster will be monitored by a

5

single cluster head that may cause a single point failure. In the random approach [27], few

wireless sensor nodes are randomly selected based on neighbor locations or hops.

The selected nodes are eligible to monitor the nearest sensor nodes within the accessible

transmission range. This random approach is suitable to elect the nodes in Ad hoc sensor

networks. In both of the approach, the selection of safeguarding nodes is open in the wireless

medium; and thus the security scheme is being challenged to ensure the channel privacy. The

recent approaches use the key management technique to provide security on the monitor node

selection process which still remains susceptible to intelligent attacks. The adversary with

sufficient hardware and software modules of cryptanalysis technique can easily revoke the

secret keys and the original data by sensing the unsecured channel. By finding the locations

of sensor nodes, the intelligent attacker can harm the data communication or the node

performance [28].

1.2. Research Contribution

The proposed scheme would provide a suitable solution against these issues by making a

secure routing and monitoring mechanism against global adversaries in Ad hoc sensor

networks using random node selection approach. This work has the following contribution:

1. Proposes a hybrid routing scheme using two routing protocols: 1. Optimized Link

State Multipath Routing (Proactive); and 2. Ad hoc On-demand Multipath Distance

Vector Routing (Reactive) discussed in Section 4.2;

2. This is the hybrid routing and monitoring mechanism with an ability to act as

proactive (link-state) and reactive (distance-vector) depend on time with the help of

Modified Two-Fish algorithm discussed in Section 4.4;

3. It is used to select the optimal paths for data transmission nodes as well as monitor

(guard) nodes with the help of several optimization algorithms to attain Concealed

Monitor Set (CMS) and secure data transmission Section 4.4 and Section 4.5; and

4. This proposed protocol finds the optimal routing paths depending upon the network

scenario, nodes mobility, frequent link breaks, consecutive update rate, network

functionality, obviously sensor nodes energy levels and so on discussed in Section

4.7.

Generally, the routing protocols are used to find the optimal path between source and

destination nodes to transmit the data; but it does not select the monitor nodes in multi-path

environments. In the above scope, this hybrid routing and monitoring approaches have been

developed and the purpose is to provide better security on the multiple channels. The

selection of monitoring nodes is done efficiently at the time of the routing process. By using

6

these combined routing functionalities and sensor monitor selection approaches, the multiple

channels are protected using two fish symmetric cryptosystem with unpredictable key shares

derivation and the data is transmitted between Ad hoc sensor nodes efficiently; even though

the Ad hoc sensor nodes are identified by the multiple global adversaries at the meantime.

The remaining sections of the paper are organized as follows: Section 2 discusses the

literature works related to ad hoc routing protocols. Section 3 formulates problem statement,

network design, adversary model, routing and monitoring, algorithms and channel-dependent

key shares for IoT-based WSNs. Section 4 presents the architecture of the proposed system

and routing algorithms. Section 5 demonstrates the simulation results of the proposed system

in comparison with other routing protocols. Section 6 concludes the research work.

2. Related Works

Generally speaking, Wireless Sensor Networks (WSNs‟) are widely used to provide

responsive information in a wide range of applications. Data concealment over the open

communication medium is offered to protect the sensor nodes and network data, and thus it

has now been picked out randomly or hierarchically for the detection of adversaries. Several

routing protocols have been proposed for mobile ad-hoc networks [28]. Each routing protocol

has a particular application scenario and characteristics. As an instance, Optimized Link State

Routing (OLSR) [29] protocol is well suited to high-dense mobile ad-hoc networks and

Destination Sequenced Distance Vector (DSDV) [30] protocol is desirable for small-scaling

ad-hoc networks. When the application scenario is difficult and the network topology varies

quickly, a single routing protocol cannot guarantee the network mobility‟s to provide better

performance. Accurate environmental perception and appropriate approaches have developed

an adaptive protocol for a better quality of services (QoS). Radhika Ranjan Roy [31]

explained the mobility model to analyze communication metrics. Fan et al. [32] represented

several mobility parameters such as geographic limitation, non-temporal and spatial

dependence to capture the important characteristics. Hong et al. [33] described the network

mobility‟s and topologies to analyze network speed and performance. A dynamic routing

mechanism extended from the dynamic routing protocol (DSR) [28] that selects relative static

nodes to apply aeronautical ad hoc networks [34].

Zheng et al. [35] presented the mobility and load aware routing to combine mobile and

load sensing that finds the multi-point relay to mitigate the average end-to-end delay.

Moreover, it is really applicable to unmanned aerial networks with high-speed and

unbalanced workloads. Bamis et al. [36] proposed a framework model that defines three-layer

7

mobility services to experience mobility and routing services. Moreover, it is well suited for

high-mobility networks with non-dense nodes. Yu et al. [37] merged the DSR routing with

ant-colony optimization to stabilize the signaling power in order to detect network

congestion. Swidana et al. [38] provided the basic idea to prevent the high-mobility node in

order to discover the routes. Khalaf et al. [39] introduced a two-probability model to improve

the perception speed that improves the network performance of ad-hoc on-demand routing

[40]. Brahmbhatt et al. [41] presented a reliable routing that selects a powerful node to

stabilize the signaling strength in order to solve the routing problem in multipath routing

networks. Alejandro Proan˜o and LoukasLazos [42] have focused on selective jamming

attacks which can be launched by real-time packet classification at the physical layer.

They have combined the symmetric key cryptographic schemes with physical-layer

attributes to provide a strong hiding commitment scheme. Also, they have compared the

security features and resource overheads with hiding techniques based on cryptographic

puzzles and AONT based hiding scheme. The selective jamming attacks are treated as an

internal model, i.e. the adversary has the internal knowledge of protocol specifications and

networks secrets. These adversaries will live only for a short period of time but focusing on

launching a jamming attack on highly important messages. Abderrezak Rachedi and Hend

Baklouti [43] proposed a smart monitoring mechanism for wireless sensor network based on

IEEE 802.15.4 MAC beacon-enabled technology. Existing watchdog mechanisms are

applicable for Mobile Ad hoc Network (MANET) and not suitable for WSN. This analytical

model aims at improving the quality of the monitoring process as well as maintaining low

energy consumption. The mu-Dog mechanism monitors the change of behavior in the normal

activities of the network, especially the non-cooperative nodes that take part in the routing

process. Parameters such as node density, size, and distance between monitored and

monitoring nodes are evaluated to assess the performance of mu-Dog mechanism and it does

a better job when compared to watchdog mechanism.

Moreover this, energy consumption is one of the main constraints of mobile sensor

networks. To keep the network lifetime high, the node with the highest remaining resource

can be elected as leader for the network. Noman Mohammed et al. [44] have identified two

critical problems in choosing the leader of mobile ad hoc networks in the presence of selfish

nodes for intrusion detection. First, is the node lying with the highest remaining resource will

pretend selfishly and try to avoid being elected as the leader and the second problem is the

optimal collection of these leaders will lead to performance overhead. These problems are

fixed by mechanism design theory. This theory will encourage the nodes in the network to

8

honestly participate in the leader election process. The nodes which actively participate are

encouraged with a certain amount of incentives for based on Vickrey, Clarke, and Groves

(VCG) Model. The optimal election of leader problem is solved by local election algorithms

and also ensures low cost. Through Cluster Dependent Leader Election and Cluster

Independent Leader Election the authors have justified the effectiveness of the proposed

schemes. The task of identifying and isolating a compromised node in wireless sensor

networks is a tedious process which leads to several security breaches such as confidentiality,

integrity and various other issues to the data gathered by the network.

The nodes in the network are monitored to identify the changes in the behavior of the

network. Nodes that are neighbors to each other are generally used for the monitoring

operation. A secure cluster formation algorithm [45] would be efficient to establish the

formation of trusted clusters with pre-distribution keys. This scheme provides reputation and

trust for a node to determine another node has been compromised or not and to make

necessary positive actions through negative information sharing and independent trust-based

decision making. Reported information is verified even in the presence of colluding nodes by

using a simple location verification algorithm based on received signal strength property. Tao

Shu et al. [46] have proposed a secured routing protocol with intrusion detection based on

WSN clustering architecture. They have predicted an energy model for nodes which are used

to identify attacks in the election phase of cluster head opted for key management technique

to ensure the security of nodes at the stage of cluster formation. Also, the flow prediction

model uses to prevent the routing-related attacks. Routing mechanisms [47] plays a major

role in wireless sensor networks. Once an adversary captures the routing protocol algorithm it

can easily traverse and cause threats to all the data packets sent over the route. To avoid this,

multipath routing [48] can be suggested to prevent these types of attacks. The data packets

are transmitted through different paths over time, which makes it difficult for the adversary to

track the sequence of packets. Besides randomness, simulation results also prove that

multipath routing is highly diffuse and energy-efficient.

A secure routing protocol does the work of sending the message to the intended

receivers. Yet the dynamic [49] environment needs an effective Intrusion Detection System

(IDS) to keep track of malicious activities of the network. Watchdog mechanisms [50] are

only suitable enough to report the abnormal behavior of the network and do not indulge in

immediate corrective actions. Hence a possible solution would be energy efficient

homogeneous and heterogeneous IDS [51,52] which meets the dynamic environment

requirements and takes control over the entire network lifetime. This work proposed a hybrid

9

routing scheme using two routing protocols [53,54,55]: 1. Optimized Link State Routing

(Proactive); and 2. Ad hoc On-demand Multipath Distance Vector Routing (Reactive). These

protocols are used to select the optimal paths for data transmitting nodes as well as monitor

nodes with the help of several layers of optimized algorithms to attain CMS and secure data

transmission. Hing et at. [56] presented routing protocol for fly-ad hoc networks to improve

the quality of services such as packet delivery ratio, end-to-end delay, throughput rate.

However, it could not be adapted to a group and pursue mobility model [57]. Of late, various

existing works [58-63] have studied the fundamental elements of IoT that address the critical

issues related to architecture, system model, and application types. However, they were

unsuccessful to investigate different types of communication protocols and standards in order

to ensure the system security.

Table 1 Summarize the contribution of existing works in relation with IoT-based cloud

computing

Existing

Work

Year of

Publication IoT

Multipath

Routing

Encryption

Model Key Contributions

Mutlag

et al. [58]

2019

o Utilize three important key factors to

manage the computation and communication resources effectively

i.e. for cloud-based healthcare

systems.

o Review some existing works to

realize the limitation and drawbacks

of system framework.

Kumari

et al. [59]

2018

o Signify various addressing issues such

as bandwidth, energy and privacy

related to healthcare system using fog

computing.

o Design three-layer architecture to

realize the key factors of real-time

application systems.

Garcia

et al. [60]

2018

o Present a fog computing framework

to speed-up the computation process

i.e. mobile clients

o Apply this framework to reduce the

computation process by four times.

Farahani

et al. [61]

2018

o Deal with a systematic framework to

provide user friendly eco-systems.

o Discuss the security issues related to

IoT-based network devices.

Ahmadi

et al. [62]

2018

o Show the design architecture for IoT-

based healthcare systems.

o Investigate the design model to

analyze the critical issues and

challenges of the system.

Baker 2017 o Discuss the key elements of

10

et al.

[63]

communication methods in IoT-based

healthcare systems

o Aim to review the cloud computing

systems to realize the issues data

storage.

3. Research Background

This section discusses the problem statement, network design, adversary model, routing

and monitoring, algorithms and channel-dependent key shares for IoT-based WSNs.

3.1. Problem Statement

From Fig.1, it is observed that a layer of proposed mechanism is employed with three

significant activities such as detection, prevention and isolation of malicious nodes. In order

to gather and analyze the network report, various communication protocols namely OLSR,

DSDV and TARCS are employed. Moreover, to evaluate the trustworthiness of the nodes, a

network pattern using KMP is generated that promiscuously monitor the on-demand routing

protocols to examine the network links. On the account of network broadcasting, the

communication channels are protected using MAC or TF. The encrypted channels may

closely observe the nodes to estimate the resource requirements that passively monitor the

trustworthiness of neighbor nodes. In order to assess the behavior of routing/path selection in

ad hoc network, the proposed hybrid routing scheme is extensively studied i.e. with DSDV,

OLSR, TARCS and AOMDV. Assume that in large scale or dense WSNs, number of

sources may send the sensed information as the data packets to a number of receivers at

time interval . There is number of moving attackers may progress themselves around the

entire network to sense the open channel or to compromise the sensor nodes. At any time,

they can inject their attacks like any runtime attacks like Packet Dropping (Black Hole), IP

Spoofing or Worm Hole to any channel or any node. To compromise the network or node, the

adversaries try to gather the confidential information or any security parameters i.e. secret

key from the wireless channel in an unauthorized way. Fig.1 illustrates the layer view of the

proposed system

11

Detection, Malicious Node Isolation, Prediction

Network Report

Gathering and AnalysisOLSR, DSDV, TARCS

On-Demand Monitoring

Reporting on Link

(AOMDV)

KMP Pattern Analysis

Multi-Channel Encryption – MAC / TF – 128 Bits

Path Selection /

Routing

OLSR, AOMDV,

DSDV, TARCS

Node Selection – Concealed Monitor Sets (CMS)

Fig.1 Layer view of the proposed system

3.2. Network Design

Let us assume the WSNs with uniform mobile sensor nodes having a high-end

microcontroller, multiple sensing diodes and one or more transceivers according to their

needs with minimal in size, memory, and energy level. Since the sensor nodes have limited

storage capacity, low communication cost and battery lifetime, a suitable IoT module i.e.

6LoWPAN is preferred to handle the crypto-system computation and execution process e.g.

BitLocker. Among those nodes, there are set of adversaries can have their chances to

compromise the network with respect to their knowledge and efficiency. This is assumed as a

random network without a base station or any clusters. All the nodes are configured with an

equal priority of functions. These node are elected and created as the random sets of allocated

weights with respect to their locality, energy level, velocity, mode of work (monitors, routers

or idle) etc. The next section will describe the details. Fig.2 illustrates the initial network

setup (OLSR) that has two source nodes , three destinations , and

Multipoint Relays (MPRs) with one-hop and two-hop neighbor. Multiple

receivers use to get the data through forwarding sensor nodes, which continuously change

from different transmitting sensors.

12

M7

M6

N4

N5

N2

S1

N1

N3

M1 M2

D3

S2

M5

D1

M4 M3

D2

S – Source

D – Destination

N – Node

M – Multi-Point Relay

Fig.2 Initial network setup

In a random graph , there are edges and vertices to create a subgraph that

provides the coverage scalability with respect to the following conditions.

{ } { } { } (1)

Where, is the set containing number of sensor nodes, { }, is the subset

containing sensor nodes provided for multipath data transmission, { };

is the subset containing sensor nodes considered for the initial evaluation process using

the mechanism of guard node election (the purpose of the initial evaluation process is to

separate the sensor guard nodes from other sensor nodes based on the hop and link-based

searching, where { }; is the subset containing the sensor nodes,

such as covered and protected as the guard nodes, where { } and is the time

interval in sec; and is the subset containing sensor nodes to represent the idle or

disconnected or damaged state; . The above conditions are taken into the account to

elect sensor guard nodes that are available from the set of n sensor nodes of the designed

network (illustrated in Fig.2). It is mainly used to identify the set of idle sensor nodes or

available nodes from multiple channels that are to act as the guard nodes in the later process.

13

This guard node selection is intended to reduce the tension on individual sensor nodes that

select the sufficient random guard nodes to travel around the network. In Fig.2, each sensor

node can either act as a data forwarding node or as a guard node to meet the required

condition; though it is not having any specific deployment or predefined sensor guard node to

incur the vulnerabilities in the open medium.

After the selection of nodes for different subsets, those are connected via one to one

vertices to the edges , like { }; { } temporarily for a session.

The wireless connection has been identified by three types of links, first, symmetric links

where the links are enabled by the secure routing protocol [64,65,66] are same for both

transmission and reply; second, asymmetric links where the links are enabled by the protocol

for transmission are not same as the receiving path; third, lost links are indicating path

breaks, node failure due to any type of adversary activities or environmental problems.

Initially, the sensor nodes and their neighbors are elected proactively by optimized and

secured link-state routing approach to avoid the adversaries at the starting point of data

communication. Then the reactive mechanism is called for managing ad hoc nature of the

sensor nodes. With respect to the link type and availability of the sensor nodes, the neighbors

are identified and maintained for secure data transmission and also for guard task separately.

By using this hybrid secure routing [57] and monitoring protocol the overhead happens in

every sensor nodes are divided as much as possible and maintain sufficient guard nodes on

multipath transmission dynamically.

Any node of any subset will be joined or eliminated to or from the subsets with respect to

their needed activities. The subset { } will be created and connected at the time of

Guard node based protection. The remaining nodes are considered as isolated in any way. At

the end of this first evaluation, k sensor nodes are taken for the second step of effective

identity verification and the third of Two-Fish symmetric key encryption using channel-based

key shares. The nodes, which were selected in this step, maybe the adversaries. To avoid this,

H-MAC has been used to ensure the monitor/guard node identity.

3.3. Adversary Model

Assume that a wireless ad hoc network composes of mobile nodes with limited

transmission region. They establish direct communication with their neighbor or adjacent

nodes over a wireless channel. In addition, they are more cooperative with each other to

achieve multi-hop routing. Most importantly, the identity of each node is cryptographically

validated in line with the conformance of network systems. Therefore, the nodes conforming

to the system protocol are assumed to be a legitimate node, whereas the node deviating from

14

the set conformity is assumed to be adversarial. The adversary tries to control the number of

malicious nodes, which may be either internal or external to gain network access. It is usually

provided with identical cryptography primitives to discover the legitimate nodes.

Since a previous legitimate node can be fully compromised to act as an internal or

external adversary, there may be a chance to create a single-adversarial mode to generate

multiple network-nodes in order to utilize the identities of compromised node and its related

cryptographic keys. This adversarial mode may randomly be variance with the defined

routing protocols. In particular, the protocol may even replay, modify or drop any

transmission message. Since the adversary has a limited computation cost, it cannot break the

cryptography primitives. If it really beneficial to rise in data theft, then the adversary may

prolong the period of routing execution time as long as possible. Therefore, the malicious

node or the adversary activity may coordinate with a compromised node to mount the data

collusion attack. In addition, there may be a chance to communicate with far distance nodes

to mount either tunnel or wormhole attack in order to chock up the network access.

There are several nodes acting as adversaries with necessary hardware privilege,

knowledge about packet data format, and data transmission protocols on the network. They

may have the facilities to do cryptanalysis process on cryptosystem. Those adversaries can

change their locations autonomously around the network, which will help them to

compromise the neighbor nodes or to eavesdrop the channel information. These attackers are

either internal attackers (compromised) or external attackers to provide packet dropping, IP

malfunctions or Wormholes.

4. Proposed Security Architecture

Fig.3 illustrates the architecture of the proposed system. The sensed real-time

environmental factors are converted into a collection of bits and transmitted to other

sensor nodes with respect to time interval . In these WSNs, the single or multiple senders

may transmit their data to one or more destinations. As understood above this is dense WSN,

there are many proactive and reactive routing protocols to select the optimum paths to route

the packets. But the particular protocol is not good for all the aspects like throughput, route

discovery, link error identification and security scalabilities in congested wireless sensor

networks. Reactive protocols effectively manage surviving routes acquires minimum

overhead but these are not suitable for the dense network. Due to this congested and

continuously updating situation, reactive protocols consume maximum delay to find and

update the routes.

15

Monitoring

Module

Node Selection

Algorithm

KMP Pattern

Matching

Local /

Global Buffer

Security Module

Authentication

Module

Channel

Encryption

||

Keyed Linear

Congruent

PRF

Keyed Linear

Congruent

PRF

AOMDV

OLSR,

DSDV,

TARCS

Routing Table

Reactive – DV

Routing Table

Proactive -LS

Routing Module Interface Table

TIMER

FLAG

TIMER

FLAG

Channel

Dependent

Key Shares

Generation

Two Fish

256 BITS

UMA

128 Bits

Fig.3 Architecture of the proposed system

In contrast, proactive protocols always have knowledge of the entire network and

periodically update the routes. This is good for large-sized network but takes enormous

memory and computational overhead. To solve these issues a hybrid routing protocol has

been designed using the functions and properties of both Multipath Optimized Link State

Routing (OLSR) / Destination Sequenced Distance Vector (DSDV) / Topology Change

Aware-Based Routing Protocol (TARCS) i.e. Proactive and Ad hoc On-Demand Multipath

Distance Vector (AOMDV) i.e. Reactive. At the same time, this fusion protocol would be

elected properly, however random concealment in monitoring the sensor nodes in a secured

way is to analyze the network traffic and the adversary involvements. In this scenario, the

routing has been done using multipath fashion which is unavoidable.

4.1. Algorithms

This subsection discusses the algorithms such as route identification rreq, initial selection

phase, secure monitors selection for multiple channels, on-demand route maintenance, and

secure transmission, aggregated monitor report generation and updates, aggregated monitor

16

report generation and updates and channel-dependent key shares to analyze the

communication metrics of routing protocols.

4.2. Algorithm 1 - Initial Setup and Neighbors Identification using MPRs

(SRC: ADDRESS; DST: ADDRESS; MSG: MESSAGE)

The algorithm illustrated in Table 2 has the initial route request broadcasting steps. (as

shown in Fig. 4). Source sends hello packet to identify the neighbors through Multipoint

relays (MPRs) on the basis of the OLSR approach. Before the formation of MPRs, the link

codes are identified to gather the modes of links and neighbors.

Table 2 Initial Setup

Hello Message: Initial setup

( )

{

Step1: Source broadcasts Hello Packets at time ;

Step2: One hop Neighbors „ „ receive Hello Packets ;

Step3 of checks Link Type of ; // Bidirectional

if // Check Link Code

{ Form of ; // Multipoint Relays

Step4: retransmits Hello Packets to neighbors ; // Two hop

Step5: Do the same Step 3 for of ;find seq. num;

Step6: Find the path to destinations ;

Step7: Redo (Periodical) ;

} }

OLSR Routing Table (Network Advertisement/ Overall Updates)

Node S1Destinations

{Monitors}

Sequence

Number

One Hop

Neighbors

Two Hop

NeighborsMPR Timer

Node S2Destinations

{Monitors}

Sequence

Number

One Hop

Neighbors

Two Hop

NeighborsMPR Timer

Fig. 4 OLSR Routing Table

Table 3 algorithm forms one and two-hop neighbor sets with minimal redundant requests

and responses. Also, the module given below would update OLSR link-state topology table

through the verification of numbers of requests and responses in sequence. Fig. 4 shows the

OLSR routing table fields for two sources and the destinations.

Table 3 Neighbors Identification using MPRs

( )

{

Step1: Where ; - Current Time ; -Interval

If ( ) {

17

// Advertised Neighbor ;

Neighbors identify and Check && ;

// Check Sequence Numbers

If ( )

{ Create Table };

Else

{ Drop ; }

}

}

Step2: Redo at

4.3. Algorithm 2 - Initial Selection Phase

Table 4 Initial Selection

{

Step 1: If ( ) {

Step 2: If ( || not in ) // Avoiding Redundancy

{

Set broadcast id=1;

S broadcast INITREQ message to qualified neighbors;

Wait for Response ;

}

Else

Update Table;

} }

Table 4 algorithm provides the initial selection phase of neighbors, which will be acting

as monitor nodes. The neighbors and paths are separately identified in their presence link

with less redundancy level of routing conversation. The Source sends INITREQ (Initial

Request) to the neighbors away from one hop distance and two-hop distance at random

interval then it will wait for INITRES (Initial Response) from that sensor nodes. The ad hoc

sensor network may have multiple sources and destinations.

Table 5 Secure Monitoring and Selection for Multiple Channels

{

Step 1: S receives INITRES from neighbors;

Step 2: -Neighbors;

where { } Step 3: If ( ) // Neighbor updates

{

; // Forwarding Nodes table updates

}

Step 4: S sends MADV to set of neighbors Ma;

where ∑

18

Step 5: S computes one-time MAC,

Step 6: S waits for MaRES after sending MaREQ;

Step 7: If( 1)

{

finds, The Eligibility Weight Function (EWF),

∑

Where, Energy spent from sensor node;

Total energy assigned in sensor node;

Data transfer rate (Bit/Sec);

Velocity of sensor nodes in mobility;

}

Step 8: If // Compares with reference value;;

{add to List(M); Set =1;

;// Monitor count;;

Update MTT; If and only if ( )

//builds valid count and topology table;

where( ) // Maintaining at least two monitors for each sensor nodes;;

Selected Monitoring nodes; Total number of nodes; Number of

channels;

eliminated nodes on setup phase,

and ; Non –ve integer;

Number of hops to reach the destination.

} }

Step 9: Sends message to neighbors.

Step 10: Update in ;

Step 11: Set Timer.

4.4. Algorithm 3 - Secure Monitoring and Selection for Multiple Channels

This section describes the selection procedure of CMS from the identified neighbors

randomly. The nodes are reserved to monitor the data transmission on multiple channels

while they are continuously in movement. Table 5 algorithm selects forwarding nodes on

different channels with their neighbors and forms CMS.

OSLR/AOMDV/TARCS- INITREQ

S-NIDSequence

NumberHop Count Q-NID Packet Type Neighbor Timer

Fig.5 Initial Request Message

OSLR/AOMDV/TARCS- INITRES

Reply Node's ID

(Q-Nid)

Sequence

NumberS-NID

Link Type

[Symmetric or

Asymmetric]

Neighbor Timer

19

Fig.6 Initial Response Message

Fig. 5 and Fig. 6 show the INITREQ and INITRES of packet structures executed by

OLSR protocol to identify the feasible sensor nodes for monitoring duty. The next algorithm

gives the node selection details.

MADV - OSLR/AOMDV/TARCS- MaREQ

S-NID Ma-ID Hop Count_Packet

TypeMACONbr Timer

Sequence

NumberFig.7

Monitor Selection Request Message

MRES - OSLR/AOMDV/TARCS- MaRES

Ma-NID S-ID Hop Count_Packet

TypeMACO1Nbr Timer

Sequence

NumberES||Bt||V

Fig.8.Monitor Selection Response Message

Monitoring Assignment / Rejection Message - AOMDV

NID V Bit CID FID Val(s)Type MACOSequence

NumberSID_/Key

Fig.9 Monitors Assignment / Rejection Message

Fig. 7, Fig. 8 and Fig. 9 illustrate that it has different Monitor selection requests and

responses with final commitment message which will be given to selected sensor monitor

nodes.

4.5. Algorithm 4: On-Demand Route Maintenance and Secure Transmission

Once OLSR finishes the CMS formation, then the timer goes off which will leads for

AOMDV to activate the further on-demand monitoring strategy. At this time, AOMDV has to

retrieve the details of nodes and neighbors from the OLSR through Protocol Interface Table.

This continues to build monitor‟s routing path which will be maintained in AOMDV routing

table considered as Sandwich Routing Table. Table 6 algorithm describes the on-demand

monitoring and route maintenance on different channels.

Table 6 On-Demand Route Maintenance

Step 1: {

{

20

Step 2: generates RREQM ; Get RESM;

Step 3: Builds route list (Multi-channels);

Build Concealed Monitor Sets (CMS);

Step 4: executes PF, (KMP) [ ]

// Pattern Matching

// Check Link Error;;

{

{ // Unsecure node or Channel (ID)

;

; Create ;

// MN-Malicious Node

[where ];

}

else

{ computes { };

Here , -Maximum Rounds;

At , While {Compute { };}

}

else

{Link error; ( through );

}}

AODV OP WAIT/STOP

( updates , ; Flush out AOMDV RTable)

}

Step 5: Set Timer. ; (after seconds) }

4.6. Algorithm 5: Aggregated Monitor Report Generation

Table 7 Monitoring Report Generation

Step 1: ( )

{

Compute ; Update search for route updates;

Continue transmission.

}

Step 2: Repeat the Algorithms.

Table 7 algorithm selects the sensor node as a monitor that executes pattern-matching

function to check the node's behavior, traffic patterns of the transmitting nodes and the

authenticity obviously using HMAC (Hash-based Message Authentication Code) and the

data has been encrypted using Two fish algorithm using the channel-dependent key shares,

discussed in Section 4.8.

4.7. Algorithm 6: Monitoring Report Updates

21

To do the fast monitor report updates at the entire nodes, OLSR starts again once

AOMDV stops the on-demand monitoring and routing. OLSR gets the updates from the

protocol interface table and spreads to all nodes with minimal delay and redundancy.

Table 8 Monitoring Report Updates

Step 1: ( )

{

Compute ; Update search for route updates;

Continue transmission.

}

Step 2: Repeat the Algorithms.

Protocol Interface Table – PLT

Node S1Sequence

Number

Next Hop/

Hop Count

M

Report

Neighbors

List

R

FlagDestination Monitors Timer

Node S2Sequence

Number

Next Hop/

Hop Count

M

Report

Neighbors

List

R

FlagDestination Monitors Timer

Fig.10. Protocol Interface Table

Table 8 algorithm continues periodically and reactively depends upon the nature of the

nodes and network conditions. The next section gives channel-dependent key shares

generation for MARS, RC6, Serpent and Twofish algorithm. Fig.10 and Fig.11 illustrates the

protocol interface table and the AOMDV sandwich routing table for both monitor nodes and

the normal data transmission nodes for multiple channels with proper channel IDs. The

following section provides the collection of monitor reports about the impact of adversaries

on the channels.

22

AOMDV – Sandwiched Routing Table

Node S1Sequence

Number

Next Hop/

Hop Count

Neighbors

ListDestination

Monitors

Routing List

Channel Key

(Symmetric)

Node S2Sequence

Number

Next Hop/

Hop Count

Neighbors

ListDestination

Monitors

Routing List

Channel Key

(Symmetric)

Channel

ID_1

Dest

(D)_Hops_H Next Hop_1Monitored Prior Hop_1

Timestamp_

T1

Channel

ID_2

Dest

(D)_Hops_H Next Hop_2Monitored Prior Hop_2

Timestamp_

T2

……… ……… ……… ……………… ……… ………

E Bit

E Bit

………

Channel

ID_N

Dest

(D)_Hops_H

Next

Hop_NMonitored Prior Hop_N

Timestamp_

TNE Bit

Fig. 11 Sandwiched Routing Table

4.8. Channel Dependent Key Shares

To create the absolute random keys here the Keyed Linear congruent Pseudo-Random

Function (KLPRF) is used.

is the set of randomly chosen multi-variants from the created in-node tuples, these

are denoted as: { }, where PSQN-

Packet Sequence Number; -Node's Timer (1-hop or 2-hop neighbors); -Association

Timer of pairs; - Tuple Timer of Link availability (Symmetric type or Asymmetric type);

-Dupe Timer of Nodes; -Timer of Message ID; -Timestamp of Message

validity; and -Tuple Timer of Multi-Point Relays:

, where A - Linear congruent generator; B - Increment; q-Modulus;K0- Seed; Pc-

key for PRF; Ki-Session Round key for two fish.

Fig.12 shows channel-dependent variable key shares derived from multi-variants taken

from a set of created tuples. These variants are used to create a different set of complex key

shares for multi-path channel communication through sensor nodes. The key encryption and

decryption blocks i.e. 128/192/256 bits are used to analyze the security level of cryptographic

applications. Advanced Encryption Standard [67] provides good security strength and

performance to fulfill the security strength of smart-card due to less ROM and RAM

23

requirements. MARS, RC6, Serpent, and Twofish [68] use the concept of a symmetric-key

block cipher to encrypt the electronic data in cryptographic applications.

XOR

M

XOR

M

XOR

M

K0 Ki-1

A1

B1

PC1 PC2

B2

A2

PCi

Bi

Ai

K1 K2 Ki

Fig.12 Channel dependent Symmetric Key Shares Derivation

5. Results and Discussion

This section provides the details of the simulation of the proposed system. The

algorithms are implemented in Network Simulator Version 2.34 [69] with the integration of

OTCL (Object Tool Command Language). It is a powerful tool to simulate the mobile ad hoc

networks that provide a low-level insightful operation to examine the network topology

including sensor nodes, network link, application protocols, and queuing. In this research, the

routing protocols are carefully designed to test the performance and functionality of the

network. In the case of framework analysis containing of detection, prediction and isolation,

an analytical approach is prudently handled to investigate the traffic generation pattern and

series of routing mechanisms defined in Section 3.1. However, the other ns-versions could

not manage or negotiate a traffic pattern of end-to-end connectivity to control the network

load. For that reason, NS-2.34 is preferred, where a secure hybrid routing protocol inherited

with the properties of both Multipath Optimized Link State Routing (OLSR) and Ad hoc On-

Demand Multipath Distance Vector (AOMDV) protocols has been deployed. This routing

24

protocol tactfully manages to monitor the activities of sensor nodes. Table 9 defines the

important simulation parameters. It uses IEEE 802.11b as a MAC layer that sets the

transmission range to be 140 m. Each node travels with a communication speed ranging from

0 to 5 . Due to high mobility, the communication nodes are clustered. However, the

routing protocols including existing and proposed could not be suitable for a high dynamic

network. The routing protocols use a transport layer protocol i.e. User Datagram Protocol

(UDP) with a constant bit rate (CBR) of 512 / 1024 Bytes. To analyze the communication

metrics such as rate of monitor and detection ratio, the simulation has been performed i.e. for

. It has a cluster of communication nodes in the range of , where the

malicious nodes are randomly deployed to analyze the mobility and data collision. Most

importantly, to analyze the network data, the geographical network area can be set as

, , , and . Providing high

mobility condition in a diversified network, a size of is chosen to probe the

communication metrics namely rate of monitor, detection ratio of wormhole attack, and

detection ratio of IP spoofing attack. In addition, it can be observed that the malicious activity

is set to be around to test the attacks namely wormhole and IP spoofing.

Table 9 Details of simulation parameters

Constraints Values

Network Size

Number of Nodes 200 Nos.

Initial Energy 10 Joules

Simulation Time 100 Seconds

Security Module HMAC , TF-256 Bits, Key Generation

Routing Protocol DSDV/OLSR/AOMDV/TARCS/Secure Hybrid Routing

Channel Model Two Ray Ground Propagation

Data Transmission Mode Constant Bit Rate (CBR)

Antenna Omni-Directional

Data Packet Size 512 / 1024 Bytes

Data Transmission Rate 16 Kbps

MAC Protocol IEEE 802.11b i.e. Distributed Coordination Function (DCF)

Range of Transmission 140

5.1. Geographical design

25

Fig. 13 Number of Sensors versus Rate of Monitor Nodes (Speed 0.5 m/s)

Ad-hoc sensors are initially deployed at different regions with a total area of

. There are 200 ad-sensor nodes are created which are roaming randomly

from one location to another. This network design is taken in the format of random disk

graph with few edges and vertices to find the set of monitors and other nodes

dynamically. The adversaries are also configured with different capabilities to inject the

attacks (packet dropping, wormhole, IP spoofing) at any links or nodes in motion. Table 9

illustrates the details of the simulation parameters.

Fig. 14 Number of Sensors versus Rate of Monitor Nodes (Speed 1.0 m/s)

0

0.1

0.2

0.3

0.4

0.5

0.6

20 40 60 80 100 120 140 160 180 200

Ra

te o

f M

on

ito

rs (

%)

Number of Sensor Nodes

OLSR [29]

DSDV [30]

AOMDV [56]

TARCS [57]

Secure Hybrid Routing

0

0.15

0.3

0.45

0.6

0.75

0.9

20 40 60 80 100 120 140 160 180 200

Ra

te o

f M

on

ito

rs (

%)

Number of Sensor Nodes

OLSR [29]

DSDV [30]

AOMDV [56]

TARCS [57]

Secure Hybrid Routing

Speed 0.5 m/s

Speed 1.0 m/s

26

Fig.13 and Fig.14 provide the details of a total number of sensor monitor nodes and the

rate of selection of monitor sensor nodes against total available nodes in networks. This

shows that the rate of selected monitor nodes i.e. for secure hybrid routing maintains at

and at any situation which secures the routing nodes on multi-

channel with maximum availability while the nodes are moving with the velocity of 0.5m/s

and 1.0m/s respectively. Moreover, it is observed that the proposed secure hybrid routing

protocol stabilizes the routing mechanism to select a secure trustworthy node. Due to less

energy consumption and successful delivery factor, the proposed secure routing protocol

achieves better security efficiency and node stability in comparison with other routing

protocols [29-30, 56-57].

Fig.15 and Fig.16 show the attack detection ratio for wormhole and IP spoofing attacks. The

attackers try to compromise the channels or the nodes to overhear the data transmission at the

MAC and routing layer. The plot gives a sense of security against increasing mobile sensor

attackers. It shows the results of attack detection ratio over the selected sensor monitor nodes

for various attacks in network links and nodes. In addition, it provides the results of routing

overhead during network route formation over mobility aspects. While comparing the

examination results, the existing protocols [29-30, 56-57] were compared with the proposed

secure routing scheme. This result illustrates that the initial routing overhead of the proposed

approach is quietly managed, though the network is congested or nodes move randomly.

Fig.15 Detection Ratio of Wormhole Attack

0

15

30

45

60

75

90

2 4 6 8 10 12 14 16 18 20

Dete

cti

on

Rati

o (

%)

Number of Warmhome Attackers

OLSR [29]

DSDV [30]

AOMDV [56]

TARCS [57]

Secure Hybrid

Routing

27

Table 10 states the compared results of different cryptosystems used for this approach.

Due to the maximum security and random ability, Two-Fish has been chosen with H-MAC

for providing multi-channel security in this network. At the same time key generation has

been done with the taken node wise random tuples attributes. The proposed approach

identifies and selects the sensor nodes for routing and monitoring functions which will not

allow tolerating one type of node on others' way. The data can be routed through the optimal

path using multi-path routing by the authenticated forwarding nodes and in a strongly

encrypted format via multiple channels. The initial selection of monitoring nodes and path

establishment requires few additional routing conversations. That has to be identified in the

increasing routing overhead than other routing approaches. But, it decreases gradually once

the sensor nodes are selected for routing and monitoring for the first time using OLSR

protocol with minimal delay. Then, AOMDV maintains the route liveliness after the initial

selection phase continuously through the period of OLSR expires.

Fig.16 Detection Ratio of IP Spoofing Attack

Moreover, the ideal or the sensor nodes which are having the least data transmission rate

or minimal velocity are taken for the monitor node selection process from the appropriate

paths. This would not harm the sensor nodes‟ overhead or the network overhead dramatically

in any way. To realize the significance, the proposed hybrid routing compared with MARS,

RC6, Serpent, and Twofish to provide best security and randomness in tiny sensor node

processors. The features of block cipher algorithms are described in Table 10.

0

10

20

30

40

50

60

70

80

90

2 4 6 8 10 12 14 16 18 20

Dete

cti

on

Rati

o (

%)

Number of IP Spoofing Attack

OLSR [29]

DSDV [30]

AOMDV [56]

TARCS [57]

Secure HybridRouting

28

Table 10 Crypto System Performance

Systems Level of

insecurity S-Boxes Rounds Key

Time

/blocks

MARS 8/14 Constant 10 128/192/256 350/425/475

Serpent 11/32 Constant 32 128/192/256 330/380/450

Twofish 6/16

Key

Dependent-

Random

16 128/192/256 360/430/490

RC6 15/20 Constant 20 128/192/256 475/530/575

6. Conclusion

The proposed hybrid routing and monitoring mechanism have been designed and

implemented with dynamically selected sensor monitor nodes in ad hoc sensor networks to

improve secure data transmission. To offer flexible security, a secure routing and monitoring

protocol was proposed with multi-variant tuples using symmetric key approaches such as

MARS, RC6, Serpent and Twofish. This proposed approach discovered and prevented the

adversaries in the global sensor network. The proposed approach designed on the basis of

Authentication and Encryption Model (ATE) that uses Eligibility Weight Function (EWF) to

select the sensor guard nodes with the help of complex symmetric key approach. With the

help of MARS, RC6, Serpent, and Twofish approach, the proposed hybrid approach use node

selection algorithm that chooses the sensor monitor node to improve with strong security

measures. This mechanism allows the ad hoc sensor network to escalate the secure data

transmission. Using Eligibility Weight Function (EWF), the sensor guard nodes are selected

to minimize the effect of malicious activities. An extensive simulation shows that the

proposed hybrid secure routing achieves a better rate of monitor and detection ration in

comparison with other existing protocols [29-30, 56-57].

In the future, a real-time testbed will be set up to implement the hybrid routing and

monitoring mechanism in order to validate the examination results. Moreover, it will be

applied in IoT-based WSNs to authenticate the secrecy of application environment e.g. smart

cities.

Conflict of Interest

29

We have participated in (a) conception and design, or analysis and interpretation of the data;

(b) drafting the article or revising it critically for important intellectual content; and (c)

approval of the final version.

This manuscript has not been submitted to, nor is under review at, another journal or other

publishing venue.

We do have affiliations with organizations with direct or indirect financial interest in the

subject matter discussed in the manuscript:

References

1. J. Rifkin, “The Zero Marginal Cost Society: The Internet of Things, the Collaborative Commons,

and the Eclipse of Capitalism: Book,” Apr. 2014

2. A. Grau, “The Internet of Secure Things What is Really Needed to Secure the Internet of Things?

Icon Labs,” Mar. 2014. [Online]. Available: http://www.iconlabs.com/prod/internet-

securethings-%E2%80%93-what-really-needed-secure-internet-things

3. U. N. IDC, Intel, “A Guide to the Internet of Things Infographic,” Feb. 2015. [Online].Available:

https://www.intel.com/content/www/us/en/internet-of-things/infographics/guide-to-iot.html

4. F. Al-Turjman, H. Zahmatkesh, "An Overview of Security and Privacy in Smart Cities‟ IoT

Communications", Wiley Transactions on Emerging Telecommunications Technologies, 2019.

DOI. 10.1002/ett.3677

5. F. Al-Turjman, “Intelligence and Security in Big 5G-oriented IoNT: An Overview”, Elsevier

Future Generation Computer Systems, vol. 102, no. 1, pp. 357-368, 2020.

6. O. Vermesan and P. Friess, “Internet of Things Applications - From Research and Innovation to

Market Deployment Book,” River Publishers, Jun. 2014. [Online]. Available:

http://www.internet-of-thingsresearch.eu/pdf/IERC Cluster Book 2014 Ch.3 SRIA WEB.pdf

7. I.F. Akyildiz et al., “Wireless sensor networks: A survey”, Computer Networks 38 (4) (2002)

393–422.

8. Yun Zhou et al., “Securing Wireless Sensor Networks: A survey”, IEEE Communication

Surveys, Volume 10, No.3, 2008.

9. S.H. Jokhio et al., “Node capture attack detection and defence in wireless sensor networks,

Published in IET Wireless Sensor Systems”, 8 August 2011.

10. Abror Abduvaliyev et al., “On the Vital Areas of Intrusion Detection Systems in Wireless Sensor

Networks”, IEEE Communications Surveys & Tutorials, Vol. 15, No. 3, Third Quarter 2013.

11. Yuxin Mao, “A Semantic-based Intrusion Detection Framework for Wireless Sensor Network”,

Networked Computing (INC), 6th International Conference, Gyeongju, South Korea 2010.

12. Rung-Ching Chen, Chia-Fen Hsieh and Yung-Fa Huang, “An Isolation Intrusion Detection

System for Hierarchical Wireless Sensor Network”, Journal of Networks, Vol. 5, Number 3

March 2010.

30

13. S. Agrawal and J. Agrawal, “Survey on Anomaly Detection using Data Mining Techniques,”

Procedia Computer Science, vol. 60, pp. 708–713, Jan. 2015. [Online]. Available: http:

//www.sciencedirect.com/science/article/pii/S1877050915023479

14. A. L. Buczak and E. Guven, “A Survey of Data Mining and Machine Learning Methods for

Cyber Security Intrusion Detection,” IEEE Communications Surveys Tutorials, vol. 18, no. 2, pp.

1153–1176, 2016.

15. L. Wang and R. Jones, “Big Data Analytics for Network Intrusion Detection: A Survey,”

International Journal of Networks and Communications, vol. 7, no. 1, pp. 24–31, 2017.

16. P. Mishra, V. Varadharajan, U. Tupakula, and E. S. Pilli, “A Detailed Investigation and Analysis

of using Machine Learning Techniques for Intrusion Detection,” IEEE Communications Surveys

Tutorials, pp. 1–1, Jun. 2018.

17. Z. M. Fadlullah, F. Tang, B. Mao, N. Kato, O. Akashi, T. Inoue, and K. Mizutani, “State-of-the-

Art Deep Learning: Evolving Machine Intelligence Toward Tomorrow‟s Intelligent Network

Traffic Control Systems,” IEEE Communications Surveys Tutorials, vol. 19, no. 4, pp. 2432–

2455, 2017.

18. E. Hodo, X. Bellekens, A. Hamilton, C. Tachtatzis, and R. Atkinson, “Shallow and Deep

Networks Intrusion Detection System: A Taxonomy and Survey,” arXiv:1701.02145 [cs], Jan.

2017, arXiv: 1701.02145. [Online]. Available: http://arxiv.org/abs/1701.02145

19. E. Benkhelifa, T. Welsh, and W. Hamouda, “A Critical Review of Practices and Challenges in

Intrusion Detection Systems for IoT: Towards Universal and Resilient Systems,” IEEE

Communications Surveys Tutorials, pp. 1–1, Jun. 2018.

20. B. B. Zarpelo, R. S. Miani, C. T. Kawakani, and S. C. de Alvarenga, “A survey of intrusion

detection in Internet of Things,” Journal of Network and Computer Applications, vol. 84, pp. 25–

37, Apr. 2017. [Online]. Available: http://www.sciencedirect.com/science/article/pii/

S1084804517300802

21. Paola Inverardi et al., “Distributed IDSs for enhancing Security in Mobile Wireless Sensor

Networks”, Proceedings of the 20th International Conference on Advanced Information

Networking and Applications (AINA‟06)1550-445X/06.

22. Rung-Ching Chen, Yung-Fa Huang, Chia-Fen Hsieh, “Ranger Intrusion Detection System for

Wireless Sensor Networks with Sybil Attack Based on Ontology”, New Aspects of Applied

Informatics, Biomedical Electronics and Informatics and Communications 2010.

23. Mohammad Saiful Islam Mamun, A.F.M. SultanulKabir, “Hierarchical Design Based Intrusion

Detection System For Wireless Ad Hoc Sensor Network” , International Journal of Network

Security & Its Applications (IJNSA), Vol.2, No.3 July 2010.

24. K.Q. Yan, S.C. Wang, C.W. Liu, “Hybrid Intrusion Detection System of Cluster-based Wireless

Sensor Networks”, 3rd IEEE International Conference on Computer Science and Information

Technology (ICCSIT), 2010.

31

25. Huang Lu et al., “A Novel Energy Efficient Routing Algorithm for Hierarchically Clustered

Wireless Sensor Networks”, International Conference on Frontier of Computer Science and

Technology 2009.

26. Xiaofang Li et al., “A Survey on Topology Control in Wireless Sensor Networks, 10th Intl. Conf.

on Control”, Automation, Robotics and Vision Hanoi, Vietnam, 17–20 December 2008.

27. BinghuiGuo et al., “The Interactive Core of A Random Wireless Sensor Network Model”,

International Conference on Computer Distributed Control and Intelligent Enviromental

Monitoring 2012.

28. Sarkar, S.K.; Basavaraju, T.G.; Puttamadappa, C. Routing Protocols. In Ad Hoc Mobile Wireless

Networks-Principles, Protocols, and Applications, 2nd ed.; CRC Press: Boca Raton, FL, USA,

2012; pp. 81–126.

29. Clausen, T.; Jacquet, P. Optimized Link State Routing Protocol (OLSR), RFC 3626

(Experimental). Available online: https://www.ietf.org/rfc/rfc3626.txt.pdf (accessed on 9 October

2018).

30. Perkins, C.E.; Bhagwat, P. Highly Dynamic Destination-Sequenced Distance-Vector Routing

(DSDV) for Mobile Computers. In Proceedings of the conference on Communications

architectures, protocols and applications (SIGCOMM94), London, UK, 31 August–2 September

1994; pp. 234–244.

31. Radhika Ranjan, R. Random Waypoint Mobility, Reference Point Group Mobility. In Handbook

of Mobile Ad Hoc Networks for Mobility Models; Springer: Boston, MA, USA, 2011; pp. 637–

670.

32. Fan, B.; Sadagopan, N.; Helmy, A. The Important Framework for Analyzing the Impact of

Mobility on Performance of Routing Protocols for Adhoc Networks. Ad Hoc Netw. 2003, 1,

383–403.

33. Hong, J.; Zhang, D. Impact Analysis of Node Motion on the performance of FANET routing

protocols. In Proceedings of the 14th International Conference on Wireless Communications,

Networking and Mobile Computing (WiCom2018), Chongqing, China, 18–20 September 2018.

34. Sakhaee, E.; Jamalippour, A.; Kato, N. Aeronautical Ad Hoc Networks. In Proceedings of the

IEEE Wireless Communications and Networking Conference (WCNC 2006), Las Vegas, NV,

USA, 3–6 April 2006; pp. 246–251.

35. Zheng, Y.; Wang, Y.; Li, Z.; Dong, L.; Jiang, Y.; Zhang, H. A Mobility and Load aware OLSR

routing protocol for UAV mobile ad-hoc networks. In Proceedings of the 2014 International

Conference on Information and Communications Technologies (ICT2014), Nanjing, China, 15–

17 May 2014.

36. Bamis, A.; Boukerche, A.; Chatzigiannakis, I.; Nikoletseas, S. A mobility aware protocol

synthesis for efficient routing in ad hoc mobile networks. Comput. Netw. 2008, 52, 130–154.

32

37. Yu, Y.; Ru, L.; Chi, W.; Liu, Y.; Yu, Q.; Fang, K. Ant colony optimization based polymorphism-

aware routing algorithm for ad hoc UAV network. Multimed. Tools Appl. 2016, 75, 14451–

14476.

38. Swidana, A.; Abdelghanya, H.; Saifana, R.; Zilic, Z. Mobility and Direction Aware Ad-hoc on

Demand Distance Vector Routing Protocol. Procedia Comput. Sci. 2016, 94, 49–56.

39. Khalaf, M.; Al-Dubai, Y.; Min, G. New efficient velocity-aware probabilistic route discovery

schemes for high mobility Ad hoc networks. J. Comput. Syst. Sci. 2015, 81, 97–109.

40. Perkins, C.E.; Royer, E.M. Ad-hoc on-demand distance vector routing. In Proceedings of the

Second IEEE Workshop on Mobile Computing Systems and Applications, New Orleans, LA,

USA, 25–26 February 1999.

41. Brahmbhatt, S.; Kulshrestha, A.; Singal, G. SSLSM: Signal Strength Based Link Stability

Estimation in MANETs. In Proceedings of the 2015 International Conference on Computational

Intelligence and Communication Networks, Jabalpur, India, 12–14 December 2015.

42. Alejandro Proan˜o and LoukasLazos, “Packet Hiding method for Selective Jamming Attacks”,

IEEE Transactions on Dependable and Secure Computing”, Volume 1, January/February 2012.

43. Abderrezak Rachedi and HendBaklouti, “MuDog: Smart Monitoring Mechanism for Wireless

Sensor Networks based on IEEE 802.15.4 MAC”, IEEE International Conference ICC 2011.

44. Noman Mohammed, HadiOtrok, Lingyu Wang, MouradDebbabi, and Prabir Bhattacharya,

“Mechanism Design-Based Secure Leader Election Model for Intrusion Detection in MANET”,

IEEE Transactions on Dependable and Secure computing, February 2011.

45. Xiao Zhenghong and Chen Zhigang, “A Secure Routing Protocol with Intrusion Detection for

Clustering Wireless Sensor Networks”, International Forum on Information Technology and

Applications, 2010.

46. Tao Shu, Marwan Krunz, and Sisi Liu, “Secure Data Collection in Wireless Sensor Networks

Using Randomized Dispersive Routes”, IEEE Transactions on Mobile Computing, July 2010.

47. Shanshan Chen, Geng Yang and Shengshou Chen, “A Security Routing Mechanism against Sybil

Attack for Wireless Sensor Networks”, International Conference on Communications and Mobile

Computing 2010.

48. Ali Modirkhazeni, NorafidaIthnin and Othman Ibrahim, “Secure Multipath Routing Protocols in

Wireless Sensor Networks: A Security Survey Analysis”, Second International Conference on

Network Applications, Protocols and Services 2010.

49. Guangcheng Huo, Xiaodong Wang, “DIDS: A Dynamic Model of Intrusion Detection System in

Wireless Sensor Networks”, IEEE, International Conference on Information and Automation

,Zhangjiajie, China June 20 –23, 2008.

50. Rodrigo Roman, Jianying Zhou and Javier Lopez, “Applying Intrusion Detection Systems to

Wireless Sensor Networks”, Proceedings of IEEE CCNC 2006.

33

51. Yun Wang, Xiaodong Wang, Bin Xie, Demin Wang, and Dharma P. Agrawal, “Intrusion

Detection in Homogeneous and Heterogeneous Wireless Sensor Networks”, IEEE transactions

on mobile computing, June 2008.

52. Piya Techateerawat, Andrew Jennings “Energy Efficiency of Intrusion Detection Systems in

Wireless Sensor Networks”, IEEE/WIC/ACM International Conference on Web Intelligence and

Intelligent Agent Technology 2006.

53. Kemal Akkaya and Mohamed Younis,” A survey on routing protocols for wireless sensor

networks”, Ad Hoc Networks, Elsevier 2003.

54. Mustafa IlhanAkbas, DamlaTurgut, “Lightweight routing with dynamic interests in wireless

sensor and actor networks”, Ad Hoc Networks, Elsevier 2013.

55. Nikolaos A. Pantazis et al., “Energy-Efficient Routing Protocols in Wireless Sensor Networks: A

Survey”, IEEE Communications Surveys & Tutorials, Vol. 15, No. 2, Second Quarter 2013.

56. Yuan, Y., Chen, H., & Jia, M. (2005, October). An optimized ad-hoc on-demand multipath

distance vector (AOMDV) routing protocol. In 2005 Asia-Pacific Conference on

Communications (pp. 569-573). IEEE.

57. Hong, J., & Zhang, D. (2019). TARCS: A Topology Change Aware-Based Routing Protocol

Choosing Scheme of FANETs. Electronics, 8(3), 274.

58. Mutlag, A.A.; Ghani, M.K.A.; Arunkumar, N.; Mohamed, M.A.; Mohd, O. Enabling

technologies for fog computing in healthcare IoT systems. Future Gener. Comput. Syst. 2019, 90,

62–78.

59. Kumari, A.; Tanwar, S.; Tyagi, S.; Kumar, N. Fog computing for Healthcare 4.0 environment:

Opportunities and challenges. Comput. Electr. Eng. 2018, 72, 1–13.

60. García-Valls, M.; Calva-Urrego, C.; García-Fornes, A. Accelerating smart eHealth services

execution at the fog computing infrastructure. Future Gener. Comput. Syst. 2018,

doi:10.1016/j.future.2018.07.001.

61. Farahani, B.; Firouzi, F.; Chang, V.; Badaroglu, M.; Constant, N.; Mankodiya, K. Towards fog-

driven IoT eHealth: Promises and challenges of IoT in medicine and healthcare. Future Gener.

Comput. Syst. 2018, 78, 659–676.

62. Ahmadi, H.; Arji, G.; Shahmoradi, L.; Safdari, R.; Nilashi, M.; Alizadeh, M. The application of

internet of things in healthcare: A systematic literature review and classification. Univer. Access

Inf. Soc. 2018, 1–33, doi:10.1007/s10209-018-0618-4.

63. Baker, S.B.; Xiang, W.; Atkinson, I. Internet of Things for Smart Healthcare: Technologies,

Challenges, and Opportunities. IEEE Access 2017, 5, 26521–26544.

64. Sanchez M, Manzoni P. Anejos: A java based simulator for ad-hoc networks. Future Generation

Computer Systems 2001; 17(5): 573–583

34

65. Aly M. El-Semary, Mohamed Mostafa A. Azim, “A Two-Tier Energy-Efficient Secure Routing

Protocol for Wireless Sensor Networks”, 7th International Conference on Information Assurance

and Security (IAS), 2011.

66. Alaauldin Ibrahim et al., “Integrated Comparison of Energy Efficient Routing Protocols in

Wireless Sensor Network: A survey”, IEEE Symposium on Business, Engineering and Industrial

Applications (ISBEIA), 2011.

67. J. Nechvatal, E. Barker, D. Dodson, M. Dworkin, J. Foti, and E. Roback, „„Status report on the

first round of the development of the advanced encryption standard,‟‟ J. Res. Nat. Inst. Standards

Technol., vol. 104, no. 5, pp. 435–459, 1999.

68. J. H. Kong, L.-M. Ang, and K. P. Seng, „„A comprehensive survey of modern symmetric

cryptographic solutions for resource constrained environments,‟‟ J. Netw. Comput. Appl., vol.

49, pp. 15–50, Mar. 2015.

69. K. Fall, K Varadhan, “ns Manual,” Available online: https://www.isi.edu/nsnam/ns/doc/

(accessed on 17 February 2019).

AUTHORS’ BIOGRAPHY

B. D. DEEBAK is presently working as Associate Professor in the department of

Computational Intelligence, School of Computer Science and Engineering at Vellore Institute

of Technology, Vellore, India. He previously associated with GMR Institute of Technology,

Rajam (AP) as Associate Professor in the Department of Computer Science and Engineering.

He also associated with Middle East Technical University (METU) Northern Cyprus Campus

during 2016-2017. He has more than 12 Years of Teaching Experience, Research in various

Engineering Institutions in India and Abroad. He received his B.Tech. (IT) from Anna

University in 2006, M.E. (Embedded System and Computing) from RTM Nagpur University,

Nagpur in 2009, and Ph.D. from SASTRA University, Thanjavur in 2016. His areas of

research include Multimedia Networks, Network Security and Machine Learning. He is an

active member in professional societies like IE (I), CSI and ISTE. He received a Research

Grant in 2012 from TCS Under Research Scholar Program collaborated with SASTRA

University and TCS Innovation Laboratory Bangalore, entitled “Secure Authentication

Schemes for Multimedia Client-Server Systems” for the amount of Rs.50 Lakhs. He also

collaborated with Middlesex University and King‟s College London during Post-Doc tenure

period in the project known as Newton Fund under British Council. His current research

interests include computer networks, wireless communication systems, wireless sensor

networks, Multimedia Networks, Routing and Security. He has published 12 papers in well

reputed publishers such as IEEE, Springer and Tubitak. He also serves as reviewer from

IEEE Communications Letters, IEEE Access, IEEE System and IEEE Sensor Journal.

Dr. Fadi Al-Turjman is a Professor at Antalya Bilim University, Turkey. He received his

Ph.D. degree in computing science from Queen‟s University, Canada, in 2011. He is a

leading authority in the areas of smart/cognitive, wireless and mobile networks‟ architectures,

protocols, deployments, and performance evaluation. His record spans more than 160

publications in journals, conferences, patents, books, and book chapters, in addition to

numerous keynotes and plenary talks at flagship venues. He has received several recognitions

and best papers‟ awards at top international conferences, and led a number of international

symposia and workshops in flagship ComSoc conferences. He is serving as the Lead Guest

Editor in several journals including the IET Wireless Sensor Systems (WSS), MDPI Sensors

and Wiley Wireless Communications and Mobile Computing (WCMC). He is also the

35

publication chair for the IEEE International Conf. on Local Computer Networks (LCN‟18).

He is the sole author for 3 recently published books about cognition and wireless sensor

networks‟ deployments in smart environments with Taylor and Francis, CRC New York (a

top tier publisher in the area).

AUTHORS’ PHOTO

B D DEEBAK

Fadi Al-Turjman