21 Mateen MSThesis

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/335652228
Towards Energy Efficient Routing in Blockchain Based Underwater WSNs via

Recovering the Void Holes
Thesis · September 2019

DOI: 10.13140/RG.2.2.12937.57444
CITATIONS READS
38 1,465
3 authors, including:
Abdul Mateen Sohail Iqbal

University of Management and Technology (Pakistan) National University of Sciences and Technology
25 PUBLICATIONS 245 CITATIONS 58 PUBLICATIONS 1,057 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Smart Grid View project
TEFL Certified View project
All content following this page was uploaded by Abdul Mateen on 11 April 2020.
The user has requested enhancement of the downloaded file.

Towards Energy Efficient Routing in Blockchain
Based Underwater WSNs via Recovering the Void
Holes
by
Hafiz Abdul Mateen
CIIT/FA17-RCS-009/ISB
MS Thesis
in
Computer Science
COMSATS University Islamabad, Islamabad - Pakistan
Spring, 2019
COMSATS University Islamabad

Holes
A Thesis presented to
COMSATS University Islamabad
in partial fulfillment
of the requirement for the degree of
MS (Computer Science)
by
Hafiz Abdul Mateen
Spring, 2019
i
Holes
A Post Graduate Thesis submitted to the Department of Computer Science as

partial fulfillment of the requirement for the award of Degree of MS (Computer
Science).
Name Registration Number

Hafiz Abdul Mateen CIIT/FA17-RCS-009/ISB
Supervisor:
Dr. Nadeem Javaid,

Associate Professor, Department of Computer Science,
COMSATS University Islamabad,
Islamabad, Pakistan.
Co-Supervisor:
Dr. Sohail Iqbal,

Assistant Professor, Department of Computing,
School of Electrical Engineering and Computer Science (SEECS),
National University of Sciences and Technology (NUST), Islamabad, Pakistan.
ii
Final Approval
This thesis titled
Towards Energy Efficient Routing in Blockchain Based Underwater

WSNs via Recovering the Void Holes
by
Hafiz Abdul Mateen
has been approved

for the COMSATS University Islamabad, Islamabad, Pakistan.
External Examiner:
Dr. Imtiaz Ahmad Taj
Professor, Dean Faculty of Engineering,
Capital University of Science and Technology (CUST),
Supervisor:
Dr. Nadeem Javaid,
COMSATS University Islamabad, Islamabad, Pakistan.
Co-Supervisor:
Dr. Sohail Iqbal,
School of Electrical Engineering and Computer Science (SEECS),
National University of Sciences and Technology (NUST),
HoD:
Dr. Majid Iqbal Khan,
COMSATS University Islamabad, Islamabad, Pakistan.
iii
Declaration
I Hafiz Abdul Mateen (Registration No. CIIT/FA17-RCS-009/ISB) hereby declare

that I have produced the work presented in this thesis during the scheduled period
of study. I also declare that I have not taken any material from any source except
referred to wherever due that amount of plagiarism is within acceptable range.
If a violation of HEC rules on research has occurred in this thesis, I shall be
liable to punishable action under the plagiarism rules of the HEC and COMSATS
University.
Date: 20 July, 2019

Hafiz Abdul Mateen
iv
Certificate
It is certified that Hafiz Abdul Mateen (Registration No. CIIT/FA17-RCS-009/ISB)

has carried out all the work related to this thesis under my supervision at the
Department of Computer Science, COMSATS University Islamabad, Islamabad,
Pakistan and the work fulfills the requirement for award of MS degree.
Date:
Supervisor:
Dr. Nadeem Javaid,

Associate Professor,
Department of Computer Science.
Co-Supervisor:
Dr. Sohail Iqbal,

School of Electrical Engineering and
Computer Science (SEECS).
HoD:
Dr. Majid Iqbal Khan,

Associate Professor,
Department of Computer Science.
v
DEDICATION
Dedicated
I am dedicating this thesis to my parents, who always support me
in my hard times and give me strength and encouragement to
achieve my goals. Thank you for supporting me and teaching me to
believe in me and my dreams. Without their support, I am unable
to complete this competitive phase of education. Without their
love and affection, none of this would matter.
vi
ACKNOWLEDGMENT
Thanks to ALLAH Almighty for giving me strength and ability to complete this
thesis. I deem it my pleasure to avail an opportunity to express the deepest appre-
ciation to my supervisor, Dr. Nadeem Javaid, Associate Professor at COMSATS
University Islamabad, for his enlightened supervision during my research period.
I owe a deep sense of gratitude to my family and friends (especially Arshed Sher,
Nimra Saleemi, Ayesha Anjum Butt and Tehreem Ashfaq) for their sincere guid-
ance, keen interest and encouragement throughout the degree.
Foremost, I must express my profound gratitude to my parents. It is because of

their moral support, guidance and encouragement that inspired me to accomplish
this degree. This achievement would not have been possible without them. In the
end, a great thanks to my ComSens family.
vii
ABSTRACT
Towards Energy Efficient Routing in Blockchain Based
Underwater WSNs via Recovering the Void Holes
Underwater Wireless Sensor Networks (UWSNs) are promising and emerging frame-
works having a wide range of applications. The underwater sensor deployment is
beneficial; however, some factors limit the performance of the network, i.e., less
reliability, high end to end delay and maximum energy dissipation. The provision-
ing of the aforementioned factors has become a challenging task for the research
community. In UWSNs, battery consumption is inevitable and has a direct im-
pact on the performance of the network. Most of the time energy dissipates due
to the creation of void holes and imbalanced network deployment. In this work,
two routing protocols are proposed to avoid the void hole and extra energy dis-
sipation problems due to which lifespan of the network will increase. To show
the efficacy of the proposed routing schemes, they are compared with the state of
the art protocols. Simulation results show that the proposed schemes outperform
their counterpart schemes.
By keeping in mind the emerging security issues in sensor networks, we have pro-
posed a blockchain based trust model for sensor networks to enrich the security of
the network. Additionally, this model provides security along with data immutabil-
ity. We have used a private blockchain because it has all the security features that
are necessary for a private sensor network. Moreover, private blockchain cannot
be accessed by using the Internet. In the proposed trust model, the Proof of
Authority (PoA) consensus algorithm is used due to its low computational power
requirement. In PoA consensus mechanism, a group of the validator is selected
for adding and maintaining blocks. Moreover, smart contracts are used to val-
idate and transfer cryptocurrency to service providers. In the end, transaction
and execution costs are also calculated for each function to testify the network
suitability.
viii
Journal Publications
1 Mateen, Abdul, Muhammad Awais, Nadeem Javaid, Farruh Ishmanov, Muham-

mad Khalil Afzal, and Saqib Kazmi. “Geographic and opportunistic recovery
with depth and power transmission adjustment for energy-efficiency and void
hole alleviation in UWSNs.” Sensors 19, no. 3 (2019): 709.
2 Ahmad, Ashfaq, Nadeem Javaid, Abdul Mateen, Muhammad Awais, and

Zahoor Khan. “Short-term load forecasting in smart grids: an intelligent
modular approach.” Energies 12, no. 1 (2019): 164.
ix
Conference Proceedings
1 Mateen, Abdul, Nadeem Javaid, Muhammad Awais, Nasir Khan, Urva Latif,
and Ihtisham Ullah. “Bio-Inspired Optimization Techniques for Home En-
ergy Management in Smart Grid.” In 2018 32nd International Conference on
Advanced Information Networking and Applications Workshops (WAINA),
pp. 250-257. IEEE, 2018.
2 Mateen, Abdul, Nadeem Javaid, Muhammad Bilal, Muhammad Arslan Fa-

rooq, Zahoor Ali Khan, and Fareena Riaz. “Towards Optimizing Energy
Efficiency and Alleviating Void Holes in UWSN.” In International Confer-
ence on Emerging Internetworking, Data & Web Technologies, pp. 516-527.
Springer, Cham, 2019.
3 Awais, Muhammad, Abdul Mateen, Zahoor Ali Khan, and Nadeem Javaid.
“A Survey on Multi-hop Routing Protocols for Efficient Resource Alloca-
tion in IoTs.” In 2018 Fifth HCT Information Technology Trends (ITT), pp.
239-244. IEEE, 2018.
4 Butt, Suhail Ashfaq, Abdul Mateen, Nadeem Javaid, and Zahoor Ali Khan.
“Towards the void hole alleviation for energy efficiency in WSN.” In 2018
Fifth HCT Information Technology Trends (ITT), pp. 318-324. IEEE, 2018.
5 Awais, Muhammad, Nadeem Javaid, Abdul Mateen, Nasir Khan, Ali Mo-
hiuddin, and Malik Hassan Abdul Rehman. “Meta Heuristic and Nature
Inspired Hybrid Approach for Home Energy Management Using Flower Pol-
lination Algorithm and Bacterial Foraging Optimization Technique.” In 2018
IEEE 32nd International Conference on Advanced Information Networking
and Applications (AINA), pp. 882-891. IEEE, 2018.
6 Awais, Muhammad, Zahoor Ali Khan, Nadeem Javaid, Abdul Mateen, Ay-
men Rasul, and Farooq Hassan. “Cluster-Based Routing Protocols with
Adaptive Transmission Range Adjustment in UWSNs.” In International
Conference on Emerging Internetworking, Data & Web Technologies, pp.
x
528-539. Springer, Cham, 2019.
7 Khan, Nasir, Nadeem Javaid, Muhammad Khan, Ahmed Subhani, Abdul

Mateen, and Arshad Iqbal. “Harmony Pigeon Inspired Optimization for
Appliance Scheduling in Smart Grid.” In 2018 IEEE 32nd International Con-
ference on Advanced Information Networking and Applications (AINA), pp.
1060-1069. IEEE, 2018.
8 Latif, Urva, Nadeem Javaid, Syed Shahab Zarin, Muqaddas Naz, Asma Ja-
mal, and Abdul Mateen. “Cost Optimization in Home Energy Management
System using Genetic Algorithm, Bat Algorithm and Hybrid Bat Genetic
Algorithm.” In 2018 IEEE 32nd International Conference on Advanced In-
formation Networking and Applications (AINA), pp. 667-677. IEEE, 2018.
9 Javaid, Nadeem, Ihtisham Ullah, Syed Shahab Zarin, Mohsin Kamal, Ba-
batunji Omoniwa, and Abdul Mateen. “Differential-Evolution-Earthworm
Hybrid Meta-heuristic Optimization Technique for Home Energy Manage-
ment System in Smart Grid.” In International Conference on Innovative Mo-
bile and Internet Services in Ubiquitous Computing, pp. 15-31. Springer,
Cham, 2018.
xi
TABLE OF CONTENTS
Dedication vi
Acknowledgements vii
Abstract viii
Journal Publications 69
Conference Proceedings 70
List of Figures xvi
List of Tables xvii
List of Algorithms xviii
List of Abbreviations xix
List of Symbols xxi
1 Introduction 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 Thesis Contributions . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Thesis Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Conclusion of the Chapter . . . . . . . . . . . . . . . . . . . . . . . 5
2 Related Work and Problem Statement 6
2.1 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1.1 Energy Efficiency . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1.2 Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.3 Topology Control . . . . . . . . . . . . . . . . . . . . . . . . 8

xii
2.1.4 Void Hole . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.5 Blockchain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.1 Sub-problem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2.2 Sub-problem 2 . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3 System Models 19
3.1 Proposed System Model 1 . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.1 End to End Adaptive Transmission Range and Depth Ad-

justment Based Model . . . . . . . . . . . . . . . . . . . . . 20
3.1.1.1 Successful Transmission . . . . . . . . . . . . . . . 20
3.1.1.2 Void Hole Occurrence . . . . . . . . . . . . . . . . 21
3.1.1.3 Void Hole Recovery . . . . . . . . . . . . . . . . . . 21
3.1.1.4 Establishing the Link with Neighbors . . . . . . . . 22
3.1.1.5 Forwarder Node Selection . . . . . . . . . . . . . . 24
3.2 Proposed System Model 2 . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.1 Ordinary Sensor Nodes . . . . . . . . . . . . . . . . . . . . . 26
3.2.2 Sink Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.3 Main Server . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.3.1 Hash Function and Merkle Tree . . . . . . . . . . . 27
4 Existing and Proposed Schemes 29
4.1 GEDAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2 LMPC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3 GEDPAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4 E2EVHR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
xiii
5 Simulation Results and Discussions 39
5.1 Simulations for Routing Protocols . . . . . . . . . . . . . . . . . . . 40
5.1.1 Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.1.2 Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.1.3 PAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.1.4 Depth Adjustment . . . . . . . . . . . . . . . . . . . . . . . 40
5.1.5 Energy Consumption . . . . . . . . . . . . . . . . . . . . . . 41
5.1.6 Network Parameters Setting . . . . . . . . . . . . . . . . . . 41
5.1.7 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . 42
5.2 Feasible Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2.1 Feasible Region between Energy Consumption and Through-

put using GEDPAR . . . . . . . . . . . . . . . . . . . . . . . 47
5.2.2 Feasible Region between Energy Consumption and PAR us-

ing GEDPAR . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.2.3 Feasible Region between Energy Consumption and Delay

using GEDPAR . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.2.4 Feasible Region between Energy Consumption and Through-

put using E2EVHR . . . . . . . . . . . . . . . . . . . . . . . 49
5.2.5 Feasible Region between Energy Consumption and PAR us-

ing E2EVHR . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.2.6 Feasible Region between Energy Consumption and Delay

using E2EVHR . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3 Simulations for Blockchain Trust Model . . . . . . . . . . . . . . . . 52
5.3.1 Simulation Tools . . . . . . . . . . . . . . . . . . . . . . . . 52
5.3.1.1 Remix IDE . . . . . . . . . . . . . . . . . . . . . . 53
5.3.1.2 Ganache . . . . . . . . . . . . . . . . . . . . . . . . 53
5.3.1.3 MetaMask . . . . . . . . . . . . . . . . . . . . . . . 53
5.3.1.4 MATLAB R2018a . . . . . . . . . . . . . . . . . . 54
5.3.2 Simulations Reasoning . . . . . . . . . . . . . . . . . . . . . 54
5.3.2.1 Execution Cost . . . . . . . . . . . . . . . . . . . . 54
xiv
5.3.2.2 Estimating Transaction Cost . . . . . . . . . . . . 54
5.3.2.3 Why Ether is Not Used Instead of Gas? . . . . . . 55
5.3.2.4 Comparison between Transaction and Execution

Costs . . . . . . . . . . . . . . . . . . . . . . . . . 59
6 Conclusion and Future Work 61
6.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
7 References 63
Journal Publications 69
Conference Proceedings 70
xv
LIST OF FIGURES
3.1 End to End Adaptive Transmission Range and Depth Adjustment

Based Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2 Void Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.3 Transmission Range Adjustment . . . . . . . . . . . . . . . . . . . . 23
3.4 Depth Adjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.5 Structure of Hello Message . . . . . . . . . . . . . . . . . . . . . . . 24
3.6 Blockchain Based System Model . . . . . . . . . . . . . . . . . . . . 27
3.7 Structure of Markle Tree . . . . . . . . . . . . . . . . . . . . . . . . 28
4.1 LMPC Layering Concept . . . . . . . . . . . . . . . . . . . . . . . . 34

4.2 Source to Destination Path Finding . . . . . . . . . . . . . . . . . . 35
5.1 Initial Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.2 Final Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.3 Depth Adjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.4 Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.5 Packets Received at Sink . . . . . . . . . . . . . . . . . . . . . . . . 45
5.6 Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.7 Average Energy Consumption . . . . . . . . . . . . . . . . . . . . . 46
5.8 Feasible Region for Energy Consumption and Throughput using
GEDPAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.9 Feasible Region for Energy Consumption and PAR using GEDPAR 49
5.10 Feasible Region for Energy Consumption and Delay using GEDPAR 50
5.11 Feasible Region for Energy Consumption and Throughput using
E2EVHR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.12 Feasible Region for Energy Consumption and PAR using E2EVHR 51
5.13 Feasible Region for Energy Consumption and Delay using E2EVHR 52
5.14 Network Deployment Cost . . . . . . . . . . . . . . . . . . . . . . . 55
5.15 State Checking Cost of Sink Nodes . . . . . . . . . . . . . . . . . . 55
5.16 Individual State Checking Cost of Sink Nodes . . . . . . . . . . . . 56
5.17 Transaction Cost of Sink Nodes . . . . . . . . . . . . . . . . . . . . 57
xvi
LIST OF TABLES
1.1 Basic Differences between UWSN and WSN . . . . . . . . . . . . . 3
2.1 Summarized Related Work . . . . . . . . . . . . . . . . . . . . . . . 11

2.1 Summarized Related Work (cont.) . . . . . . . . . . . . . . . . . . . 12
2.2 Summarized Related Work for Blockchain . . . . . . . . . . . . . . 13
5.1 Network Parameters Setting . . . . . . . . . . . . . . . . . . . . . . 41

5.2 Network Deployment Cost . . . . . . . . . . . . . . . . . . . . . . . 56
5.3 Current Active/De-active State of Nodes . . . . . . . . . . . . . . . 57
5.4 Current Active/De-active State of Nodes on Main Server . . . . . . 58
5.5 Comparison of Different Costs . . . . . . . . . . . . . . . . . . . . . 59
xvii
List of Algorithms
1 Periodic Beaconing . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2 Beacon Receive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3 Next Forwarder Node Selection . . . . . . . . . . . . . . . . . . . . 32
4 Void Hole Recovery for GEDAR . . . . . . . . . . . . . . . . . . . . 33
5 LMPC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
6 Void Hole Recovery for GEDPAR . . . . . . . . . . . . . . . . . . . 36
7 E2EVHR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
xviii
List of abbreviations
ADMM Alternating Direction-Method of Multipliers

CMA Confusion Mechanism Algorithm
CSI Channel State Information
DApps Distributed Applications
E2EVHR End to End Void Hole Recovery
FTF Futile Transaction Filters
GEDAR GEographic and opportunistic based Depth
Adjustment Routing
GEDPAR GEographic and opportunistic based Depth
and Power Adjustment Routing
GPS Global Positioning System
IoTs Internet of Things
IoVs Internet of Vehicles
IVs Intelligent Vehicles
IVTP Intelligent Vehicle Trust Point
LDB Locally Dynamic Blockchain
LMPC Layered Multi-path Power Control
LoRaWAN Long Range Wide Area Networks
MEC Mobile Edge Computing
MdO Multi-domain Orchestrator
msec millisecond
NC-DRDS NC Deterministic DS
NC-DS Network Coding based Distributed Storage
NC-RLDS NC Rate less DS
PPIP Pairwise Proportional Imitation Protocol
PAR Packet Acceptance Ratio
PDP Provable Data Possession
PDR Packet Delivery Ratio
PBFT Practical Byzantine Fault Tolerance
xix
PoA Proof of Authority
PoC Proof of Collaboration
PoW Proof of Work
RF Radio Frequency
SDN Software Dened Network
UWSN Underwater Wireless Sensor Network
WSN Wireless Sensor Network
xx
List of symbols
a(f ) Absorption coefficient

κ Beacon messages
Ecom Communication energy
nv Current void node
DistSDop Optimal distance from source to destination
σ Path loss constant
ν Set of next forwarder nodes
∆ Set of void nodes
N oiseship Shipping noise
m Slope for straight line
k Spreading factor
N oisether Thermal noise
N oisetotal Total noise
N oisetur Turbulence noise
N oisewind Wind noise
xD X-coordinate of receiving node
xS X-coordinate of sending node
yD Y-coordinate of receiving node
yS Y-coordinate of sending node
xxi
Chapter 1
Introduction
1
CHAPTER 1. 1.1. INTRODUCTION
1.1 Introduction
The planet Earth, on which we live our lives, consists of 70% water. Whereas,
the oceans hold more than 90% of total water. This much quantity shows the
importance of the water medium. To explore the underwater medium for getting
and sharing the important information, a network is deployed in a specific region.
Information transmission using Underwater Wireless Sensor Networks (UWSNs)
is one of the emerging technologies and is used for the betterment of ocean obser-
vation systems. Applications of UWSNs range from aquaculture to oil industry;
instrument monitoring to climate recording; pollution control to prediction on
natural disasters; search and survey purposes to submarine purposes.
The sensor node in a UWSN acquires the desired information and transmit towards
the next forwarder node which is closer to the sink1 [1]. This sink may be the
onshore data center or a simple sensor node over the water surface. In the data
forwarding procedure, the source2 node generates data packets and communicate
with its neighbors to find the potential node. Afterward, the potential neighbor
node finds the next potential node from its neighbors and transmit data packet
towards that potential node. To find the potential neighbor from the forwarder
node, some criteria and routing procedure are defined. This criterion may base
upon efficient energy utilization or alleviation of void holes.
Blockchain attracts great courtesy of researchers as they have faith in that this
technology will bring remarkable changes and opportunities to industries. Blockchain
is very powerful technology for resolving trusted communications in a decentralized
fashion. Basically, it deals digitally and exemplified by the Bitcoin. This tech-
nology was introduced back in 2008 and circulated by the cryptography mailing
group.
The main strength of blockchain is decentralization, which allows direct (point to

point) transactions. This approach can be used in distributed systems where trust
is not needed for nodes to do transactions. blockchain adopts many means such as;
time stamping, distributed consensus, data encryption and economic incentives.
It is used for solving the problems having inefficiency, high cost and insecure data
storage.In recent years, the research on blockchain technology has been motivated
to grow quickly with the rapid popularity and development of blockchain.
1
Sink: This word is alternatively used as sink node, sonobuoy, destined node and a destination
node.
2
Source: The words source node and initial node are alternatively used for the source.
2 Thesis by: Hafiz Abdul Mateen

The Radio Frequency (RF) waves cannot be used in an underwater medium. The
reason behind is that RF technology increases the energy consumption by increas-
ing the attenuation factor. At lower frequencies (30 to 300 HZ), water becomes
a conductor for RF waves. Moreover, the frequencies in this range require the
large size of antennas and these antennas require high transmitting power for data
transmission. These requirements cannot be fulfilled in UWSNs. Consequently,
RF waves cannot be used in UWSNs. Additionally, the technology of optical
waves requires very high precision on a single point for a transmitter and a re-
ceiver. However, sensor nodes move with the water current. In essence, we have
to use the acoustic waves in UWSNs. Nevertheless, the speed of acoustic waves is
almost five orders of magnitude less than the speed of RF waves [2].
The underwater medium is extremely unpredictable and challenging when com-

pared with Wireless Sensor Network (WSN). The major differences of UWSN with
WSN are: 1) high energy consumption 2) high propagation delay 3) low bandwidth
4) high dynamic topology-operation 5) less propagation speed 6) low efficiency 7)
low data transmission rate and 8) high environmental and noise interferences. The
comparison between UWSN and WSN is shown in Table 1.1.
Table 1.1: Basic Differences between UWSN and WSN
Base of difference UWSN WSN

Energy consumption Very high Low
Propagation delay High Low
Bandwidth Low High
Dynamic topology oper- High Low
ation
Efficiency Low High
Data transmission rate Low High
Environmental and High Low
noise interference
Communication medium Acoustic RF waves
waves
Speed of propagation 1200m/s to 3 ×108 m/s
1500m/s
The design of routing protocols has paramount importance in UWSN. These pro-
tocols indicate the routing path for data from the source node at the bottom
towards the sinks node at the surface of the ocean. Expressly, these protocols

face the different challenges which are associated with the underwater medium,
e.g., limited battery resources, interference, noise, reliable Packet Delivery Ratio
(PDR), high propagation delay, movements of sensors and void holes.
Efficient energy usage is one of the most important task during the design of a
routing protocol. As the batteries of sensor nodes in an underwater environment
are non-removable and have limited energy storage. This issue provides a strong
base for efficient battery utilization. Mostly, energy dissipates during the processes
of data packet transmission and reception. Efficient energy usage depends on var-
ious factors. For instance, the initial position and number of anchor nodes; sensor
nodes and the way in which nodes are deployed. The deployment of a network
must be one of the two types 1) sparse deployment and 2) dense deployment. The
sparse deployment leads toward the creation of a void hole and dense deployment
results in an excessive amount of sensors failure.
The energy and network stability have a direct relation. As more will be the
energy of sensors, longer will be the stability of the network and vice versa. Void
holes are areas within the transmission range of a network where a node cannot
find its next neighbor or forwarder. The void holes creation has following reasons
1) node becomes dead due to a lot of energy usage and 2) no forwarder node.
Topology control has the ability to overcome the undesired effects of UWSNs and
consequently to enhance the performance of routing protocols. The relationship
between topology control and UWSN is summarized as follows [7]:
• In UWSNs, wireless communication is provided by the acoustic channel to

enable networking services.
• UWSNs have many peculiar characteristics that enhance the challenges in

effective networking design.
• To overcome these challenges and increase the performance of the network,

the topology control method is the best solution.
Localization of sensor network in underwater is indispensable. The gathered data

is useless until it is not correlated with the specified position of the sensor node.
Localization in UWSNs is very important as it has many useful applications, e.g.,
target tracking, underwater environment monitoring, pollution control and geo-
graphic routing protocols. Nevertheless, UWSNs cannot use Global Positioning
System (GPS) due to high energy dissipation and high attenuation of RF signals
[9, 20].
CHAPTER 1. 1.2. THESIS ORGANIZATION
1.1.1 Thesis Contributions
The major contribution of this thesis can be listed as follows:
• GEographic and opportunistic routing with Depth and Power Adjustment

Routing (GEDPAR) and End to End Void Hole Recovery (E2EVHR) routing
techniques are proposed for underwater environment.
• GEDPAR and E2EVHR are compared with GEographic and opportunistic

routing with Depth Adjustment Routing (GEDAR) and Layered Multi-path
Power Control (LMPC) state of the art routing protocols to testify the per-
formance.
• Extensive simulations are performed to check the effectiveness of the pro-

posed schemes.
• Blockchain based trust model for sensor networks is also proposed in current
thesis to enrich the security of the network.
• Smart contract for trusted blockchain model is written and transaction and
execution costs are recorded.
1.2 Thesis Organization
The rest of this thesis is organized as follows: Chapter 2 includes related work
and problem statement. The system models for routing protocols and blockchain
based trust model are presented in Chapter 3. The existing and proposed schemes
along with algorithms’ are elaborated in Chapter 4. The detailed discussions on
extensive simulation results are written in Chapter 5. Finally in Chapter 6, the
thesis is concluded along with the future work.
1.3 Conclusion of the Chapter
In this chapter, overview of UWSNs, WSNs, blockchain and routing protocols are
provided. Energy efficiency, network lifetime and the concept of void holes are
also discussed in this chapter. In addition, the existing literature regarding the
challenges for routing protocols are also mentioned. Nevertheless, the state of the
art literature regarding the aforementioned things will be provided in Chapter 2.
Chapter 2
Related Work and Problem Statement
6
CHAPTER 2. 2.1. RELATED WORK
2.1 Related Work
In this chapter, we review and compare some recent works on the basis of covering a
specific area of UWSNs. The works which cover the energy efficiency and void holes
are compared in section 2.1.1. Additionally, the works that cover the concept of
localization or geographic routing are compared in section 2.1.2. Moreover, section
2.1.3 represents the comparison of topological control based schemes. The concept
of a void hole is presented in section 2.1.4. Finally, the works which implement
blockchain are presented in section 2.1.5. Nevertheless, Table 2.1 provides the
summary of these works.
2.1.1 Energy Efficiency
The research works [3, 4, 5, 6] proposed different schemes to enhance the energy-
efficiency. The works in [3] and [4] have used the multi-hop techniques. Whereas,
the work [3] is focusing on network reliability, mobility management, PDR and
energy efficiency. On the other hand, the work [4] is only focusing on energy
efficiency. Both works [3] and [4] achieve their objectives; however, end to end
delay is compromised. The authors in works [5] and [6] mainly focus on reliability
by covering one-hop from the forwarder node. The proposed scheme EBLE from
the work [5] aims to minimize the energy dissipation with packet size management.
The objective is successfully achieved on the cost of delay. The cooperative routing
is used in work [6] for data reliability and mobility management, while PDR and
efficient energy usage are the main aims. The objectives are achieved successfully;
however, the network performs poorly in sparse network deployment.
The works [10, 11, 12, 13] are also using energy efficiency techniques. The works
[10] and [12] provide the reliability. Both of works discuss the concept of multi-
hoping. The proposed scheme in the work [10] is beneficial for a large amount of
data packets; however, this proposed technique does not perform well in sparse
network deployment. The MLPR from [12] looks toward the efficient path for
routing by utilizing minimum energy. For the implementation of MLPR, more
memory is required for the extra operations at each node. The energy dissipation
schemes; SDVF and EBULC are proposed in works [11] and [13], respectively. Both
schemes consider mobility management for decreasing the energy consumption in
UWSNs. Results show that end to end delay in the works [11] and [13] is enhanced.

The energy efficiency is focused in the works [14, 15, 16, 17]. In [14], some data
collection methods are discussed which used minimum energy for data transmis-
sion from source to the destination. In both [14] and [15], mobility management is
considered, while in the [14], reliability and packet size management is not consid-
ered. Nevertheless, the works [15, 16, 17] focus on the reliability of the network.
Additionally, [15] considers both types of forwarding strategies; single-hop and
multi-hop. While [14], [16] and [17] only focus on single-hop from the current
node. Moreover, the work in [15] considers the security issues of UWSNs. While
in [14], the authors discusses the problems of getting route information. In [16],
the complexity of the network is a major challenge. Additionally, the authors in
[17] work for energy efficiency by managing the size of data packet.
2.1.2 Localization
The authors in [1, 2, 9, 18, 22] discussed the geographic or localization-based rout-
ing. The work in [1] and [2] review the works in which the concept of localization
based routing is used. Both of these above, discuss reliability and none of them
work on mobility management or packet size management. Moreover in [1] and
[2], the concept of single-hop and multi-hop is devised. The challenges which are
discussed in these works are high interference, limited batteries of sensor nodes,
low bandwidth and malicious attacks. The work in [18] achieves the higher PDR
by finding the locations of alive nodes. Afterward, the data packets are sent to
these alive nodes, accordingly. The challenges discussed in [22] are localization,
feasible hardware, relevant simulation tools and low power gliders.
2.1.3 Topology Control
Authors in [4, 7, 8, 20] proposed topology control based solutions. TCEB and
GARM schemes are proposed for controlling the topology of UWSNs in [4] and
[8], respectively. In addition, the [7] classifies different topological protocols. From
[7], reliability and mobility is discussed. The work [7] focuses on single-hop and
multi-hop while the work [4] only focus on next forwarder node. The challenges
that discussed in [4, 7, 8, 20] are: high attenuation, mobility of sensor nodes, energy
efficiency, low bandwidth, connectivity loss, high bit rate error, high deployment
cost, complexities and optimal location of glider. Using dynamic topological strat-
egy, work in [4] achieves energy efficiency and the work in [8] enhances both PDR
and energy efficiency. In [20], mobility management is a major consideration using
EEL and the concept of multi-hoping. In addition, the work [20] achieves better
simulation results from compared ones.
2.1.4 Void Hole
The concept of a void hole is presented in [21, 23, 24, 25]. Void holes are the
regions within the network range from where further data delivery is not possible.
In other words, if a forwarder node does not have any further node for data packet
transmission then this node is called void node and the area where transmission is
not possible in called void holes. TORA is presented in [21] in order to avoid the
void holes. The proposed scheme uses the concept of multi-hoping to avoid void
holes and to improve energy efficiency. Nevertheless, reliability and complexity of
this scheme are not discussed.
2.1.5 Blockchain
In [31], the authors proposed a blockchain-oriented secure-service provisioning

mechanism for the lightweight IoT devices. Authors applied smart contracts to
check the validity of acquired services. High throughput and low latency using
consortium-blockchain with PoA is achieved. Analysis of packaging time, through-
put and latency by comparing PoA with PoW is also done in this work. Moreover,
the results show that the proposed scheme guards lightweight devices from un-
trusted edge service providers and insecure services.
In [32], it is expected by the researchers that conventional blockchain technology

cannot be effortlessly applied to mobile devices. It is due to the reason that Proof-
of-Work (PoW) prerequisites large computational ability and storage volume dur-
ing the mining process. For this, authors proposed Mobile Edge Computing (MEC)
enabled wireless blockchain framework in [33]. In which they used stochastic ge-
ometry theory and Alternating Direction-Method of Multipliers (ADMM) based
algorithm. The proposed algorithm is also compared with the existing centralized
solution. Simulation results demonstrate that the proposed algorithm is efficient.
In [34], Lin et al. proposed a blockchain-based solution for Long Range Wide Area
Networks (LoRaWAN). Authors integrate blockchain and LoRaWAN by consider-
ing the crowd sensing and sharing economy. They developed a LoRaWAN server
to solve the problems of trust on private network operators and lack of network

coverage. In this study, there is a mechanism proposed to verify the existence of

data at a specific time on a network. blockchain-based incentive mechanism was
proposed in [35] for data storage by the nodes of a WSN. The authors used Prov-
able Data Possession (PDP) technique instead of PoW to obtain better results.
They also applied preserving hash function to compare the existing data of nodes
with the new one. The only problem with PDP is that it can identify the damaged
data on nodes, but is unable to recover it.
The authors in [36] proposed a data transmission scheme based on the multi-
link concurrent communication tree model. It is to handle the failure nodes in
blockchain. Results show that proposed scheme works effectively for 15% of failed
nodes. However, if this number reaches the 30%, communication time and delay
will increase.
In [37], problem of user access control for network optimization in a data-intensive
application was identified. Proposed solution considers authenticity of Channel
State Information (CSI) using blockchain consensus and deep learning. Analysis
shows that the proposed scheme increases the spectral efficiency.
Branch based blockchain technology for Intelligent Vehicles (IVs) was proposed in
[38]. Branching is done at Locally Dynamic Blockchain (LDB). It is to handle the
large amount of data generated by IVs. While blockchain is used to keep track of
the data generated by IVs and to verify it. Additionally, the concept of Intelligent
Vehicle Trust Point (IVTP) is also introduced to build the trust. Problem with
branching is that duplicate state changes increase with increasing load.
The authors in [39] identified that there is a need to revise the current storage
mechanisms to save the storage space. They proposed the Network Coding based
Distributed Storage (NC-DS) framework and embed it into existing blockchain
to store the encoded packets of the block. They implemented NC Determinis-
tic DS (NC-DRDS) and NC Rate less DS (NC-RLDS) to save the storage space.
Analysis shows that the storage space can be largely saved using the above men-
tioned mechanism. However, due to different encoded packets on different nodes
consensus mechanism cannot be applied to defend from pollution attack.
Orchestration is the automated configuration, coordination and management of

computer systems and softwares. A Blockchain-based Distributed Applications
(DApps) framework for multi domain service orchestration was proposed in [40].
The authors used blockchain, smart contract and DApps to solve the automation
and distributed harmony issues in networking. Results show that it is essential
for the Multi-domain Orchestrators (MdO) blockchain network to be secure. Also,
Table 2.1: Summarized Related Work
CHAPTER 2.
Type of technique Technique/ Reliability Packet size Mobility Number Achievements Challenges Limitations
Protocol management management of hops
RE-PBR [3] X x X Multi-hop PDR and energy ef- Difficult to deploy dense In dense deployment, end
ficiency is enhanced network to end delay is increased
TCEB [4] x x x Multi-hop Less energy dissipa- Serious cause of energy End to end delay is en-
tion consumption is attenua- hanced
tion in UWSNs
EBLE [5] X X x Single-hop Lower energy dissi- Path loss/dead due to con- Energy consumption is de-
pation by balancing tinuous data packet trans- creased on the cost of de-
the traffic load mission lay
Cooperative X x X Single-hop Successful packet Wireless sensor nodes Network performance de-
routing [6] delivery and lower move with current grades in sparse conditions
Energy efficiency energy usage
CS [10] X x x Single- Beneficial for large Less number of resources Does not perform effec-
hop and amount of data and energy efficiency tively in Sparse network
multi-hop packets deployment
SDVF [11] X X X Single- Increase network Energy efficiency, network Increase in source to desti-
hop and lifetime and PDR complexity and routing se- nation delay
11
multi-hop curity
MLRP [12] X x x Multi-hop Find efficient path Loss of data during trans- More memory at each node
and minimize en- mission process is required for extra oper-
ergy dissipation ations
EBULC [13] x x X Multi-hop Energy efficiency Complexity of UWSNs Energy usage is minimizes
on the cost of end to end
delay
Energy ef- x x X Multi-hop Energy efficiency Overhead of routing infor- Delay is increased
ficient data enhanced success- mation and increase in op-
collecting fully erational time
method [14]
Review of X x X Single- - Security issues and energy Does not discuss the com-
existing tech- hop and consumption plexity of the reviewed
niques [15] multi-hop schemes
Retransmission X x x Single-hop Enhanced PDR Complexity of the network Proposed scheme is too
and redundant much complex to imple-
approach [16] ment
Integer-linear X X x Single-hop Lifetime of network Optimal solution for en- Source to destination de-
programming is increased ergy dissipation and data lay is increased
[17] packet size
Thesis by: Hafiz Abdul Mateen

2.1. RELATED WORK
Table 2.1: Summarized Related Work (cont.)
Type of technique Technique/ Reliability Packet size Mobility Number Achievements Challenges Limitations
CHAPTER 2.
Protocol management management of hops

Review on lo- X x x Single- - Malicious attacks Cannot explain how flood-
calization algo- hop and ing and path loss problems
rithms [1] multi-hop can be compromised
Review on X x x Single- - High interference, limited Do not discuss the PDR
localization hop and battery of nodes and low and void holes
based routing multi-hop bandwidth
Localization
algorithms [2]
Review of X x x Single- - Limited bandwidth, delay Considerable number of
different tech- hop and problems, localization and challenges are ignored
niques [9] multi-hop security issues
RBCN [18] x x x Multi-hop Increase in PDR Find the locations of alive End to end delay is com-
nodes promised
Overview of x x X Single- - Localization, hardwares, Issues related to localiza-
UWSN works hop and simulation tools and tion are not discussed, e,g.,
[22] multi-hop low-power glider malicious attack
EEL [20] x x X Multi-hop New algorithm and High cost and complexity Cannot find the optimal
Improvement in the issues point for localization and
results energy usage
12
TCEB [4] x x x Multi-hop Energy consump- Topology change is not End to end delay is in-
tion is reduced much efficient due to at- creased
due to dynamic tenuation
topology
Topology control based Classify topol- X x X - - Mobility of sensor nodes Does not provide efficient
ogy control al- makes difficulty in efficient algorithm
gorithm [7] usage of batteries, loss of
connectivity and high bit
rate error
GARM [8] x x x Single-hop PDR and energy ef- Optimal location of glider Proposed scheme works
ficiency enhanced and minimum channel at- better in predefined envi-
tenuation ronment
TORA [21] X x x Multi-hop End to end delay Low bandwidth, high la- Proposed scheme takes
Void node and alleviation of tency and error rate more time on computa-
void holes tions
GEDAR [24] X x X Multi-hop Void hole avoidance Computations and energy Energy consumption for
consumption depth adjustment is high
LMPC [25] X x x Multi-hop Void hole allevia- Dividing the network area Communication overhead
tions into layers due to multiple copies,
which results in communi-
cation delay

2.1. RELATED WORK
Table 2.2: Summarized Related Work for Blockchain
Type of technique Technique used Research area Consensus/ Achievement Limitation
CHAPTER 2.
Validation
algorithm
Blockchain oriented IoT PoA Proposed scheme Efficient coordination be-
secure service pro- guards lightweight tween devices is still miss-
visioning mechanism devices ing
[31]
LoRaWAN [34] Crowd sensing - Solved the trust and Computational complexity
security problems is increased
Multi-link concurrent WSN - Proposed scheme Delay and computation
communication tree works effectively till time is increased when
model [36] 15% of nodes failure node failure reached at
30%
Network optimiza- IoT IVTP IVTP is introduced Duplicate state changes in-
Blockchain tion in a data- for building the trust crease with increasing load
intensive application between devices
[38]
NC-DRDS and NC- Storage issue in IoT - Storage is largely Due to different encoded
RLDS [39] devices saved packets on different nodes,
13
consensus mechanism can-
not be applied.
DApps framework for Automation and dis- PoW Security is major con- Transaction confirmation
multi domain service tributed harmony is- cern time should be defined
orchestration [40] sues in networking
Blockchain-based Privacy protection - Increased the user’s Results obtained on a lim-
location privacy in IoT devices using participation ited data
protection incentive CMA
mechanism [41]
Blockchain is inte- IoV PoW Large data storage Throughput is enhanced
grated with IoVs [44] and security issues on the cost of delay
Blockchain based AI-powered net- PBFT The proposed frame- No reward is given to the
data sharing frame- works work performs well in data provider
work [46] terms of privacy, secu-
rity and scalability
Rolling blockchain IoT - Removed PoW and Security issues and pollu-
[49] provide the central tion attacks are ignored
storage
Decentralized trust IoV PoS and PoW Finding of correct Security of the vehicle
management system nonce for hash func- owner is compromised
[54] tion

2.1. RELATED WORK
transaction confirmation time should be well defined for better performance. At

the same, representing smart contract and interpreting it in the proposed frame-
work is still an open topic to be covered.
In [41], a blockchain-based location privacy protection incentive mechanism is pro-

posed. Confusion Mechanism Algorithm (CMA) is introduced to protect the user’s
information by encrypting the received information from the sensors. blockchain
is also there to further secure the user’s information and issue incentives based
on the frequency of participation. Results show that the proposed mechanism
increased the user’s participation largely from 20% in traditional to 80% in the
proposed mode. However, the results obtained may be one-sided as a limited data
(100 pieces of data) was collected.
Increased male participation can also be seen in the proposed system. Conven-
tional models of business are not fit for E-business on IoT. Therefore, an IoT
E-business model is proposed in [42]. This model uses blockchain and smart con-
tract techniques. There are two main commodities in the model. Decentralization
is achieved using blockchain and smart contract is used to control the ownership
of assets. Realization of exchange of ownership of assets and designing a uniform
data format is still missing.
In [43], the authors proposed blockchain-based big data sharing framework to build
the trust among edge devices. They use Proof of Collaboration (PoC) consensus
mechanism, Futile Transaction Filters (FTF) algorithm, Express Transactions (E-
TX) and hollow blocks to save computational resources, storage resources and to
enhance the network efficiency respectively. They achieved high throughput and
low network usage. Designing the blockchain layer in the proposed framework in
a green and efficient manner is still an issue.
In [44], blockchain is integrated with Internet of Vehicles (IoVs) to provide large

and secure data storage. The authors designed multi-blockchain architecture con-
sists of five blockchains according to the different data blocks to be stored. Results
show that this integration provides large and secure data storage. They achieved
high throughput with increasing data, but delay also increases. In [45], the authors
proposed a vehicular network architecture based on blockchain for smart city. The
authors used the blockchain with smart contract. However, the service providers
are not rewarded so they will not provide services effectively.
In [46], the authors proposed the first blockchain-based data sharing framework for
AI powered network operations. The authors used blockchain and smart contract

to overcome the issues of secure data sharing. For data validation the proposed
framework uses Practical Byzantine Fault Tolerance (PBFT) consensus algorithm.
The results show that the proposed framework performs well in terms of privacy,
security and scalability. However, no reward is given for data provider to motivate
them to share their data.
The authors in [47] by using blockchain technology a proof of concept architecture

is presented to implement an access management system for IoT. The permissions
and credentials of different IoT resources are stored globally. The results show that
when the WSNs are connected to multiple management hubs the proposed system
performs well. However, when the WSNs are connected with a single management
hub the architecture performs as centralized IoT system.
Sharma et. al proposed a distributed secure Software Defined Network (SDN)

architecture for IoT using the blockchain technique. The results show that the
proposed system performs well in terms of scalability, accuracy, defense effects
and efficiency. However, the data storage issue is ignored.
In [49], the authors proposed a method by considering the issues of research data
right management. The proposed method uses smart contract and blockchain to
manage research data right between publisher and user. The results show that
the cost of setting and execution of smart contract is approximately 1.87$ which
is very low as compared to Elsevier and Springer. However, the publishers are not
rewarded so they will not publish their research data for reuse. The authors in
[50] proposed a rolling blockchain concept for IoT devices. The IoT devices have
battery resources and computational power to carry out PoW. The results show
that the blockchain remains stable with increasing number of attacks. The lost
blocks depend on the density of the sensors and attack intensity. Rahmadika et.al
proposed a model for energy trading in a small environment by using blockchain.
Chord algorithm is used in the overlay network for fast distributed computation
of hash function mapping keys. Hash function mapping keys are distributed to
nodes responsible for them. The results show that 51% of the mining power is
required to add an invalid block in the blockchain. However, the attacks in the
overlay network are ignored in this work.
In [52], the authors model dynamics of mining pool selection as an evolutionary

game. During modeling the hash rate for puzzle solving and block propagation
delay are taken into account. Furthermore PoW for block mining and Pairwise

CHAPTER 2. 2.2. PROBLEM STATEMENT
Proportional Imitation Protocol (PPIP) to replicate the dynamics of the popula-

tion. The results show that the individual miners join the mining pool with the
minimum hash rate required.
The authors in [53] examine how blockchain technology facilitates the transition of
health-care from institution driven to patient driven interoperability. The authors
summarized the work by declaring the challenges which should be addressed before
the implementation of the proposed solution. However, the challenges are clinical
transaction volume, privacy and security, patient key management and reward in
form of cryptocurrency.
In [54], the author analyses some of the present blockchain networks to determine
that they satisfy Metcalf’s law or not. The digital currency the network has is
considers the value of the network. The number of unique addresses, engage with
the network per day is considered as users of the network. Furthermore, a new
model is proposed, the value of the network is proportional to the exponential of
the square root of the active users. The analysis shows that the growth in the value
of the network is proportional to the number of addresses participating actively in
the network.
The authors in [55] proposed novel hybrid network architecture by leveraging SDN
and blockchain technology. The proposed architecture is divided into two parts;
core network and edge network. By dividing the architecture into two parts, the
architecture has both centralized and distributed features and strengths. The
proposed architecture is compared to ethereum blockchain and the difference of
16.1s is observed in latency. However, edge nodes are not deployed efficiently and
have issues in enabling caching techniques at edge nodes.
2.2 Problem Statement
This section contains two problem statements. In section 2.2.1, the limitations for
existing state of the routing protocols for Underwater environment are discussed
and after that, solution for the problems are also proposed. In section 2.2.2, the
existing security limitations are discussed and trust model for the avoidance of
malicious nodes is proposed.

CHAPTER 2. 2.3. CONCLUSION OF THE CHAPTER
2.2.1 Sub-problem 1
In UWSN, each sensor has limited resources and requires effective utilization of
these resources. Efficient energy consumption has a major contribution to stabilize
the network for long term communication. In UWSNs, the packet is sent from the
source node to the sink node using different relay nodes. If a node cannot find
a forwarder node in its transmission range, it causes hindrance in the network
during communication.
In order to avoid the void holes in UWSNs, a routing protocol namely GEDAR
is presented in [24]. GEDAR addresses the issue by adjusting the depth of nodes;
however, the process of depth adjustment consumes lots of energy. In [25], LMPC
routing technique addresses the efficient data transmission by making the binary
tree from the root node. However, binary tree generation consumes high energy
and leads towards the transmission overhead. To overcome the aforementioned
problems, two routing protocol namely GEDPAR and E2EVHR are proposed for
avoiding the void holes and eliminating the extra energy consumption.
2.2.2 Sub-problem 2
In conventional routing protocols, central authority is required for facilitating the

authentication and identification of every device. The research work in [26] im-
plements the blockchain in networks for avoiding malicious attacks. Two types
of attacks are considered in this paper: greyhole attacks and blackhole attacks.
However, the performance of the network was gradually decreased and some un-
necessary computations was involved due to the PoW consensus algorithm. In
current work, a trust model is proposed for avoiding the malicious attacks and
provided the privacy and security to the sensor networks using the concepts of
blockchain.
In current chapter, first, the related work for sensor networks is elaborated. Sec-
ond, the state of the art work for blockchain is discussed. After that, a compre-
hensive and comparative analysis among current works is also performed in this
chapter. In the end, problem statements for our proposed work are written. In

Chapter 3, the proposed system models for routing protocols and blockchain based
trust model will be presented.

Chapter 3
System Models
19
CHAPTER 3. 3.1. PROPOSED SYSTEM MODEL 1
3.1 Proposed System Model 1
In this chapter, two system models are elaborated. System model in section 3.1.1
represents the end to end path recovery along with adaptive transmission range
and depth adjustment. The second system model for our another proposed routing
protocol is presented in Figure 4.2
3.1.1 End to End Adaptive Transmission Range and Depth

Adjustment Based Model
In this section, our first proposed system model is presented in Figure 3.1. This
system model consists of source nodes, relay nodes and sonobuoys. Source node
forwards data packets toward the destined sonobuoys during transmission. The
proposed protocol follows multi-hoping feature for packets transmission. Source
and relay nodes only use acoustic signals while radio waves are used for communi-
cation among sink node, submarine, satellite, base station and the main processing
unit.
In the proposed system model 1 (Figure 3.1), sensor nodes are randomly deployed
in an underwater medium. Nevertheless, sink nodes are deployed at the sea sur-
face. The same transmission range and energy are assigned to each sensor node.
Moreover, each sensor node has also the ability to adjust their depth from the
lower layer to the upper layer. During depth adjustment, nodes only move in a
vertical direction. The process of depth adjustment occurs in the case when a node
cannot find its next forwarder even by increasing the transmission range. There
are three different cases that are elaborated through the proposed system model
1.
• Successful transmission
• Occurrence of a void hole
• Void hole recovery
3.1.1.1 Successful Transmission
In the current scenario, successful transmission occurs when a packet which is

generated from the source is successfully received at the sink node. In this case, if
Satellite
Submarine
Main processing center

R1
Sink node
L4
New Depth
Base station
R1
R1
L3
Recovered node via enhancing
transmission range
Recovering node that enhanced
transmission range
L2
Node after depth adjustment
Node before depth adjustment
Relay node
L1
Source node
Radio waves
Acoustic waves
Figure 3.1: End to End Adaptive Transmission Range and Depth Adjustment
Based Model
a packet is transmitted from the initial node, it follows different paths to reach a
destination. During transmission process, packet moves from one depth to other
and one layer to other. However, the direction of the packets is the sink node.
3.1.1.2 Void Hole Occurrence
Void holes can be defined as nodes having no neighbors in their communication

range. Void holes are generated due to many reasons; e.g., 1) A node can be dead
due to again and again re-selection of the same node or 2) No next forwarder node
exists in the transmission range of current forwarder node. The occurrence of void
hole and blockage in transmission is summarized in Figure 3.2.
3.1.1.3 Void Hole Recovery
On the occurrence of void holes, we have to perform some recovery methods. In

this work, GEDPAR is proposed for void hole recovery purpose. GEDAR and
LMPC are also used for the same purpose. However, both of these methods have
some limitations, e.g., GEDAR always takes depth adjustment on the occurrence

Sink node
Void node
Source node
Figure 3.2: Void Node
of a void hole. On the other hand, LMPC uses the concept of the binary tree and
forward multiple copies of each packet.
In the proposed scheme, we can use the option of an increase in transmission range
or depth adjustment. Depth adjustment is used in the case when a neighbor node
could not found even in maximum transmission range. Figure 3.3 presents the
transmission range adjustment. The solid circle shows the original transmission
range while dotted circle shows the enhanced transmission range. Depth adjust-
ment procedure is shown in Figure 3.4. In this figure, R1 shows the minimum
transmission range and R2 indicates the maximum transmission range. In case,
if a node cannot find next forwarder in R2 then the current node must adjust its
depth vertically. The node in blue color shows the new depth.
3.1.1.4 Establishing the Link with Neighbors
Sensor nodes are deployed in an underwater environment and each sensor generates
a hello message (control message) to find its neighbors. This hello message is a
tuple of several things: source-id, destination-id, the status1 of a node (dead or
alive), type of node (source node or sink node), coordinates node and residual
energy of current node. This structure is presented in Figure 3.5.
1
Status: The words state and status are used alternatively to show the state of nodes.

Sink node
Source node
Figure 3.3: Transmission Range Adjustment
Sink node
Depth adjustment
R1 R2
Source node
Figure 3.4: Depth Adjustment
The estimated and the actual distances of each neighbor are calculated on the
basis of coordinates.

Hello packet
Destination Coordinates of current Residual

Source ID Status Type Energy
ID node
Figure 3.5: Structure of Hello Message
3.1.1.5 Forwarder Node Selection
According to our proposed system model, neighbor node is selected on the op-
timality basis. The criteria for optimality in this scenario for a node is to have
efficient energy and must lead towards the sink node. Optimal energy point for a
node is calculated using the following formula (adapted from [24]):
s
B
DistSDop = σ . (3.1)
(1 − 2(1−σ) )
Here, DistSDop represents the optimal distance from source to the destination
node. Where, σ is a path loss constant and calculation for B is as follows:
B = 2 × Ecom . (3.2)
Where, Ecom represents the communication energy. The slope for straight line m
from source node to the destination node is calculated as:
yS − yD
m= . (3.3)
xS − xD
Here, yS , yD , xS and xD are the coordinates of sending and receiving nodes. The
Eq. 3.4 describes the path loss during attenuation of signal. This equation is taken
from [24].
A(d, f ) = DistSDop k a(f )DistSDop , (3.4)

where, a(f ) is the absorption coefficient, k is a spreading factor and DistSDop in-
dicates the optimal distance from a source to the destination node. The absorption
coefficient is described by Throp’s formula.
0.11 × f 2 44 × f 2
10loga(f ) = ( ) + ( ) + 2.75 × 10−4 f 2 + 0.003. (3.5)
1 + f2 4100 + f 2
Transmitting probability for any node with distance d is calculated using the Eq.
3.6.
P (d, m) = (1 − P (DistSDop ))m . (3.6)
Turbulence noise N oisetur , shipping noise N oiseship , wind noise N oisewind and
thermal noise N oisether are calculated using the Eqs. 3.7, 3.8, 3.9 and 3.10, re-
spectively. These aforementioned equations are taken from [18].
N oiseT ur = 10((17−30×log10(f ))×0.1) . (3.7)
N oiseship = 10((40+20×(shp−0.5)+26×log10(f )−60×log10(f +0.03))×0.1 . (3.8)
0.5 )+(20×log10(f ))−40×log10(f +0.4))

N oisewind = 10((50+((7.5×wind)) . (3.9)
N oisether = −15 + (20 × log10(f )). (3.10)
Whereas, total noise N oisetotal is calculated by adding all noises as:
N oisetotal = N oiseT ur + N oiseship + N oisewind + N oisether . (3.11)
3.2 Proposed System Model 2
In this section, the blockchain base system model for malicious node avoidance in
WSNs is presented. The key components for this system model are classified as
follow:

• Main server
• Sink nodes
• Ordinary sensor nodes
3.2.1 Ordinary Sensor Nodes
These sensor nodes only monitor the environment and collect the real-time data
and upload this data to the associated sink node.
3.2.2 Sink Nodes
These nodes perform three major responsibilities; first, data collection from ordi-
nary sensor node; second, new node addition using PoA consensus mechanism and
last, smart contract execution which is published by the main server. Sink nodes
differentiate data on the base of id and location of the ordinary node. Each sink
node has its own database consists of hashes to keep the record of transactions.
Every sink node has the ability to communicate with ordinary sensor node, other
sink nodes and the main server. Sink node uses the private keys for accessing the
data from the main server.
3.2.3 Main Server
The main server is also known as endpoint or base station. The major tasks for
base station are to publish the smart contracts, issuance of activity and processing
of sensed data. The main server records each and every transaction along with
sink id and location in its immutable database. This database can only be accessed
by the main server itself or pre-authorized sink nodes.
It can be observed from the Figure 3.6 that the ordinary nodes are connected with
sink nodes. Every sink node will get the data from ordinary sensor nodes. Sink
nodes can send their data to other sinks as well as to the main server. Where, a
smart contract is implemented on sink nodes and issued by the main server. Sink
nodes can authenticate and blacklist any ordinary sensor node at any time on the
detection of malicious activity. Each sink has a communication record of its own
as well other nodes in its distributed ledger.

Database
Database Database
Database
Database Database
Ordinary Sensor Wireless Connection of Sink

Sink Node Main Server Database Database with Sink/Main Server
Nodes
Figure 3.6: Blockchain Based System Model
In this system model, the validity of data is checked at sink nodes. Noticeable
thing is that access on the main server is only granted to the sink nodes. The
main server checks the working status of sink nodes and ordinary nodes. It can
also remove any node if 1) it is dead or 2) involved in any suspicious activity.
3.2.3.1 Hash Function and Merkle Tree
For each transaction, a hash is generated which is called transaction hash. A hash
is just a function which takes the input value and generates the output value. This
output is a deterministic value against the input value. This can be mathematically
written as
f (a) = b, (3.12)
where, a is any input and b is associated output against the a; e.g., hash value for
“hi” in kessak-256 is “7624778dedc75f8b322b9fa1632a610d40b85e106c7d9bf0e74-
3a9ce291b9c6f”. Hash values are generally ‘irreversible’ which means that input
cannot be figured out by knowing the output except hit and trial method.
Main Hash
Hash ( Hash a + Hash b )
Hash a Hash b
Hash ( Hash a1 + Hash a2 ) Hash ( Hash b1 + Hash b2 )
Hash a1 Hash a2 Hash b1 Hash b2
Hash ( Hash D1 ) Hash ( Hash D2 ) Hash ( Hash D3 ) Hash ( Hash D4 )
Figure 3.7: Structure of Markle Tree
A hash tree or Merkle tree consists of the leaf nodes, which are a label with the
hash of data block and non-leaf nodes, which are labeled with the hash of its
own child nodes. This tree allows safe, secure and efficient content verification of
large data structure. A Merkle tree compiles all the transactions in a block by
composing a digital signature for the entire set of transactions. The structure for
this hash tree is presented in Figure 3.7.
In this chapter, we have discussed the proposed techniques along with system
models in detail. Blockchain based trust model for avoiding the malicious node
attacks in sensor networks is also presented. Additionally, the details and working
of this trust model are also elaborated in this chapter. In next chapter, the existing
and proposed schemes are discussed in details.

Chapter 4
Existing and Proposed Schemes
29
CHAPTER 4. 4.1. GEDAR
4.1 GEDAR
GEDAR is an opportunistic and depth adjustment-based routing protocol. In

GEDAR, each packet is sent to the forwarding set which consists of several neigh-
bors. Algorithm 1 shows the procedure of periodic beaconing in GEDAR. This
procedure requires S and D. Where, κ represents beacon messages. Lines (4-16)
elaborate on the overall procedure for distance and neighbor calculations. Lines
(8-11) add neighbors to the neighbor list. Line 6 shows that this procedure repeats
for each and every source node. Similarly, Algorithm 2 shows the steps upon re-
ceiving the beacon message. From Algorithm 2, lines (5-10) show the information
update of neighbors, while lines (7-8) show that the sequence of neighbor node is
updated.
Algorithm 1 Periodic Beaconing

1: node (S, D)
2: network deployment
3: κ: beacon message
4: if beacon is timed out then
5: κ.coordinates = distance (node)
6: if node N then
7: for s S do
8: if λs = 0 then
9: add in κ neighbor list (s.id, x-coordinates, y-coordinates)
10: λ=1
11: end if
12: end for
13: end if
14: broadcast λ
15: set new timeout
16: end if
Algorithm 3 elaborates the selection procedure of next potential node. The packet
will only be sent to the potential node and this criterion is defined in Algorithm
3. Where, ςj is a copy of {j and n1 show the highest priority of node. In this

CHAPTER 4. 4.2. LMPC
Algorithm 2 Beacon Receive

1: node (S, D)
2: κ: beacon message
3: if κ is from sink node then
4: update S(node) and κ
5: else update neighbor information (κ.id, κ.coordinates)
6: for s κ do
7: if sequence number(s,κ) > sequence number(s, S(node)) then
8: update S(node)
9: end if
10: end for
11: end if
algorithm, firstly, we define the set of candidate neighbor nodes. Each time ςj
takes a copy of {j for maintaining the list of nodes within its transmission radius.
Algorithm 4 involves the steps for the recovery of the void hole. First of all,
value for the current node is set to “1” for its identification and stops the beacon
messages (lines 1-3). The symbol ∅ shows that current node has no neighbor (line
4). In other words, it is a void node. ν is a set which contains the record of
next forwarder nodes. ∆ and nv are the void nodes set and current void node,
respectively. The distance for each forwarder near the current void is calculated
in line 10. In lines 11-14, this distance is compared with the transmission range.
If the distance is less than the transmission range, then the next forwarder node
is within the range of the current forwarder node and vice versa. In case, if no
forwarder node exists within transmission range then depth adjustment takes place
and the status for the void node is set to “0” from “1” (lines 17-18).
4.2 LMPC
In LMPC, for efficient transmission, multi-layer concept is implemented by divid-

ing the network environment into unequal layers; as, the working of LMPC totally
depends on the layers. We have already discussed that noise in deep water is less
than the shallow water. So, the size of a layer in deep water is greater than the
shallow water. The size of a layer has an inverse relation with noise attenuation;

CHAPTER 4. 4.2. LMPC
Algorithm 3 Next Forwarder Node Selection

1: node (S, D)
2: i=1
3: ςj = {j
4: while |{| > 0 do
5: ` = n1 {j
6: { j = { j - nj
7: for nu {j do
8: if D(n1, nu ) < 1/2ηc then
9: ` = ` U nu
10: { j = { i - nu
11: end if
12: end for
13: i=i + 1
14: end while
15: for `j do
16: for nk ς do
17: if D(nk , nt ) < rc then
18: ì = ì U nk
19: end if
20: end for
21: end for
greater the attenuation of noise, lower will be the layer size and vice versa. The
pictorial form of these layers is presented in Figure 4.1 and Algorithm 5 elaborates
the steps for LMPC routing scheme. From Algorithm 5, line 1 shows the input
and line 2 represents the parameters’ initialization. Total energy of each node is
calculated in lines 3-5. Total number of layers in which network is divided are de-
cided in line 6. Line 7 shows the deployment of nodes. Neighbor finding procedure
is done at lines 8-11. The start of the communication is represented by the line
12. On successful communication, acknowledgment message is sent to the source
node.

CHAPTER 4. 4.3. GEDPAR
Algorithm 4 Void Hole Recovery for GEDAR

1: if current node is void = 1 then
2: stop beacon messages

3: end if
4: ν = ∅: no neighbor node
5: ν: set of next forwarder nodes
6: ∆: set of void nodes
7: nv : is current void node
8: if |ν| > 0 then
9: for nu ν do
p
10: dist = (xv − xu )2 + (yv − yu )2
11: if dist ≤ rc then
12: (xv − xu )2 + (yv − yu )2 + (zv − zu )2
13: ν = ν U zv
14: end if
15: end for
16: end if
17: nv moves to new calculated depth
18: current node is void = 0
4.3 GEDPAR
GEDPAR is a proposed routing protocol and to show the efficacy of this protocol,
GEDAR and LMPC are taken as benchmark schemes. In GEDPAR, the concept
of depth adjustment is taken from the GEDAR. GEDPAR takes transmission
enhancement step on the appearance of void holes. Transmission enhancement
takes some extra energy; however, most of the void holes are removed in this
process. If a node cannot cover the void hole even by increasing the transmission
range then depth adjustment takes place for that node. Figure 3.1, Figure 3.2,
Figure 3.3 and Figure 3.4 show the pictorial summary of the proposed algorithms.
Working of Algorithm 1, Algorithm 2 and Algorithm 3 are the same for GEDAR
and GEDPAR. However, Algorithm 6 plays an important role for differentiating
GEDPAR from GEDAR.

CHAPTER 4. 4.4. E2EVHR
Sink node
Low depth region L3
Medium depth region L2
High depth region

L1
Source node
Source node Sink node Layer size
Relay node Source to sink packet delivery

Layer
Figure 4.1: LMPC Layering Concept
4.4 E2EVHR
This proposed protocol is the improved version of LMPC. However, in this routing
protocol first of all complete route finding procedure is executed. In E2EVHR,
binary tree is generated (as in LMPC) for ensuring the successful packet trans-
mission. Actually, after ensuring that there is no void hole in the routing path,
multiple copies of the packets are transmitted towards the sink node using bi-
nary tree. The layering concept for LMPC is presented in Figure 4.1. The major
difference between LMPC and E2EVHR is that E2EVHR avoids the void hole
by looking the path from source to the destination, while LMPC looks forward
one-hop from neighbor node.

Algorithm 5 LMPC
1: input: LMPC(node)
2: initialization
3: for each node n do
4: calculate total energy
5: end for
6: total number of layers L
7: deploy the nodes n
9: find neighbor: (neighbor.id, neighbor.coordinates, neighbor.layer)
10: end for
11: hello messages
12: packet sent from lower depths to the higher ones
13: acknowledgment message
14: end
Sink node
Low depth region L3
Medium depth region L2
High depth region

L1
Source node
Source node No potential node Source to sink path

Layer size
Relay node Sink node Source to sink packet delivery Layer
Figure 4.2: Source to Destination Path Finding

Algorithm 6 Void Hole Recovery for GEDPAR

1: LMPC(node)
2: if current node is void = 1 then
3: stop beacon messages
4: end if
5: ν = ∅: no neighbor node
6: ν: set of next forwarder nodes
7: ∆: set of void nodes
8: nv : is current void node
9: if |ν| > 0 then
10: enhance transmission radius
p
11: dist = (xv − xu )2 + (yv − yu )2
13: goto (23)
14: else
15: for nu ν do
p
16: dist = (xv − xu )2 + (yv − yu )2
18: (xv − xu )2 + (yv − yu )2 + (zv − zu )2
19: ν = ν U zv
20: end if
21: end for
22: nv moves to new calculated depth
23: end if
24: current node is void = 0
25: end if

Algorithm 7 E2EVHR
1: input: E2EVHR (node)
2: initialization
3: total number of layers L
4: deploy the nodes n
6: calculate total energy
7: end for
9: find neighbor: (neighbor.id, neighbor.coordinates, neighbor.layer)
10: end for
11: hello messages
12: for route finding do
13: find the path which ensures end to end delivery
14: save all the paths
15: end for
16: for potential neighbor do
17: select the potential neighbors
18: end for
19: packet sent towards the potential neighbor
20: acknowledgment message
21: end
Algorithm 7 presents the steps for E2EVHR protocol. The lines (1-2) are present-
ing the input and initialization, respectively. The line (3) involves the layering
while line (4) shows deployment of the nodes in the network. Total energy for
each node is calculated in lines (5-7). The neighbors for each node are calculated
in lines (8-10). In lines (12-15), the paths having no void node are evaluated.
The lines (16-17) help to find the potential neighbors from the current forwarder.
The packets are then transmitted toward the potential neighbor for continuing the
process of transmission. These potential neighbors help the packets to reach at
the destination, successfully.
Algorithms are processed according to the following order. Firstly, Algorithm 1 is

run which takes the input as sensor node S and sink node D. Secondly, Algorithm
2 is called in the response of Algorithm 1. After that, Algorithm 3 is called for
potential neighbor selection. In the end, Algorithm 6 is called (in case of void hole
occurrence) for void node recovery. While in the case of E2EVHR, void holes are
avoided by finding the path which ensures end to end delivery.
In this chapter, the existing and proposed schemes are discussed in details. Ad-
ditionally, pseudo-codes of routing protocols are also presented for better under-
standing the flow of work. Further, simulation results and discussions are ap-
pended in Chapter 5.

Chapter 5
Simulation Results and Discussions
39
CHAPTER 5. 5.1. SIMULATIONS FOR ROUTING PROTOCOLS
5.1 Simulations for Routing Protocols
Simulations are performed to validate the proposed routing techniques. The pro-
posed techniques are compared with GEDAR and LMPC. GEDPAR and E2EVHR
are greedy opportunistic routing protocols in which next forwarder node is selected
on the criteria of minimum distance from the current node. In the GEDPAR,
firstly, current node enhances transmission range when it finds no neighbor in its
transmission range. After that, if the current forwarder still not able to find any
node in its range then it executes depth adjustment. During depth adjustment,
the node moves from a deeper layer to the shallower one. The second proposed
routing algorithm (E2EVHR) takes some steps from LMPC.
Performance of GEDPAR and E2EVHR are compared on the basis of delay,

throughput, Packet Acceptance Ratio (PAR), depth adjustment and energy con-
sumption. These parameters can be defined as follows:
5.1.1 Delay
It can be defined as the time duration in which packet is successfully received at

a sink. It is measured in milliseconds (msec).
5.1.2 Throughput
Throughput is calculated as the total number of packets received at the sink node
to the total number of packets sent from source node in percentage.
5.1.3 PAR
PAR is a ratio between packets received at the sink to the packets sent from the
source node.
5.1.4 Depth Adjustment
The number of times nodes move from deep water to the shallow water to find the
next neighbor node in their range.

5.1.5 Energy Consumption
Total energy consumption of network during transmission, reception, transmission

range enhancement and depth adjustment. This total energy consumption of
network is measured in Joule (J).
5.1.6 Network Parameters Setting
The network is deployed over the area of 1500m × 1500m × 1500m. The num-
ber of nodes and sinks are 100 and 45, respectively. Initially, nodes are deployed
randomly. The initial transmission range of each node is 245m and nodes can
transmit up to 270m using some extra energy. This happens only when current
forwarder cannot find the next node in its transmission area. The initial energy of
each node is 100J. The velocity of acoustic waves and bandwidth for the network
is considered 1500m/s and 3000kHz, respectively. Transmission energy, reception
energy and idle time energy are considered as 2W, 0.1W and 10 × 10−3 W, re-
ceptively. Size of hello packet is 100bytes while the sizes of all other packets are
150bytes. Summary of these parameters is presented in Table 5.1.
Table 5.1: Network Parameters Setting
Parameter Value
Network dimensions 1500m × 1500m × 1500m
Number of sink nodes 45
Other nodes 100
Minimum transmission range 245m
Maximum transmission range 270m
Initial energy of nodes 100J
Velocity of acoustic waves 1500m/s
Bandwidth 3000kHz
Packet transmission energy 2W
Packet reception energy 0.1W
Idle time energy 10 × 10−3 W

5.1.7 Simulation Results
Initial deployment of our model is presented in Figure 5.1. This is a 3D deployment

which covers x-axis, y-axis, and z-axis, respectively. The x-axis represents the
width of the deployed wireless network. Y-axis indicates the breadth and z-axis
cover the depth of the network. From the Figure 5.1, red “*” and blue “+” are
representing the sink nodes and source nodes, respectively.
1500
1000
Depth
500
0
1500
1000 1500
1000
500 500
Breadth 0 0
Width
Figure 5.1: Initial Deployment
In Figure 5.2, final deployment is presented. This final deployment is done after
adjusting the transmission range and depth of all nodes. These transmission ranges
and depths are only adjusted in case of void hole occurrence. When void hole
occurs, first transmission range is enhanced for finding the next forwarder node
using some extra energy. If none of the forwarder neighbors is found then depth
adjustment occurs. Depth adjustment is done after enhancing transmission range
because it consumes more energy.
Figure 5.3 depicts the depth adjustment of nodes. We can see from the Figure 5.3
that most of the depth adjustment is done during the start of network deployment.
Once the network is deployed and initial depth adjustments are done then there
exist only a few occasions on which depth adjustment is required. A large amount
of energy is dissipated during the process of depth adjustment. So, we make sure
that the depth adjustment only occurs when it is necessary. Otherwise, try to avoid
the nodes by enhancing the transmission range. It is clear from the Figure 5.3
1500
1000
Depth
500
0
1500
1000 1500
1000
500
500
Breadth 0 0
Width
Figure 5.2: Final Deployment
that in GEDPAR routing protocol, the nodes require fewer depth adjustments as
compare to GEDAR. This step further involved in lessening the energy dissipation.
70
GEDAR
Number of depth adjustments
60 GEDPAR
50
40
30
20
10
0
0 50 100 150
Time (msec)
Figure 5.3: Depth Adjustment
The throughput of proposed routing protocols is compared with GEDAR and

LMPC. Figure 5.4 shows this comparison and assures the efficiency of proposed
schemes. According to simulation results, LMPC performs better than GEDAR

while GEDPAR outperforms both GEDAR and LMPC. The efficiency of the pro-
posed scheme is better than LMPC and GEDAR by the percentage of 13% and
37%, respectively. As we can see from the Figure 5.4 that the second proposed
technique also outperforms the counterparts. Firstly, E2EVHR performs poor
during the end to end node finding process; however, when the proposed scheme
finds the end to end routes than its throughput is improved drastically.
Figure 5.4: Throughput
Figure 5.5 shows the performance of GEDAR, GEDPAR, E2EVHR and LMPC
with respect to PAR. PAR is already defined in section 5.1.3 as it is a ratio between
packets received at sonobuoys to the packets sent from source nodes. We measure
PAR from the range of 0 to 1. Here, 0 indicates the minimum PAR and 1 represents
maximum PAR. In our scenario, during PAR calculation it is noticed that PAR
of GEDAR, LMPC, E2EVHR and GEDPAR is about 0.6, 0.8, 0.93 and 0.95,
respectively. These statistics clearly show that GEDPAR and E2EVHR perform
best among the comparing ones.
Delay is calculated in terms of time duration which is required by a packet to

reach from source to destination node. Basically, calculation of delay starts from
the time when the first packet is transmitted from the source node. The calculation
of delay continues until the last data packet is received at a sink. In addition, delay
contains total duration from the generation of the first packet at a source node to
the reception of the last packet at the sink node. E2EVHR takes more time at start
due to the necessary calculations for source to sink path finding procedure. When
1.2
0.8
PAR
0.6
GEDAR
GEDPAR
0.4
LMPC
E2EVHR
0.2
0 50 100 150
Time (msec)
Figure 5.5: Packets Received at Sink
it finds the complete path by ensuring that there is no void node in the route
then it transfers the packets directly from selecting paths. On the other hand,
GEDPAR involves more calculations and the reason is that it takes more time to
the comparing schemes. Delay of the comparing schemes is presented in Figure
5.6. Moreover, Figure 5.6 also shows the trade-off and this trade-off authenticates
the simulation results. This means that for getting minimum energy consumption
and throughput, we have to compromise on computational time.
In Figure 5.7, the comparison among the different routing protocols with respect
to energy consumption is presented. From the Figure 5.7, GEDPAR consumes less
energy as compared to the GEDAR and LMPC. GEDAR consumes more energy
then counterparts because it focuses on depth adjustment during the void hole
avoidance. Depth adjustment takes 1500mJ energy for one meter while transmis-
sion range enhancement takes less energy than depth adjustment (adapted from
[24]). Figure 5.7 is also representing the average energy consumption of all com-
paring schemes. Simulations are performed multiple times to check the average
results. LMPC uses multiple transmissions for one packet which resulted in en-
ergy dissipation. The proposed routing protocol (GEDPAR) consumes less energy
because it covers the void holes by increasing the transmission range. GEDPAR
only changes the depth when no forwarded node is found even by increasing trans-
mission range.

CHAPTER 5. 5.2. FEASIBLE REGIONS
Figure 5.6: Delay
104
3.5
Average energy consumption (J)
2.5
1.5
0.5
0
GEDAR LMPC GEDPAR E2EVHR
Figure 5.7: Average Energy Consumption
5.2 Feasible Regions
Feasible region is the area where all possible solutions of a particular problem ex-
ists. In this section, mathematical formulation using linear programming is used
to find the feasibility of the proposed protocol. To achieve the optimal solutions,
we define some constraints (listed below). By defining these constraints, the co-
ordinates of feasible regions are calculated. In this work, we have calculated the
coordinates for three feasible regions according to the proposed schemes (GEDPAR
and E2EVHR).
In Figure 5.8- 5.10, we annotate the feasible regions as P1, P2, P3 and P4. Figure
5.8 presents the feasible region of energy consumption versus throughput. Fig-
ure 5.9 illustrates the feasible region (energy consumption versus PAR) for the
GEDPAR routing protocol. While in Figure 5.8, the feasible region of energy
consumption versus delay is presented.
In Figure 5.11- 5.13, we annotate the feasible regions for E2EVHR as P1, P2,
P3 and P4. Figure 5.11 presents the feasible region of energy consumption ver-
sus throughput, whereas the Figure 5.12 illustrates the feasible region of energy
consumption versus PAR, for the E2EVHR routing protocol. In Figure 5.11, the
feasible region of energy consumption versus delay is presented.
5.2.1 Feasible Region between Energy Consumption and

Throughput using GEDPAR
The feasible region between energy consumption and throughput of the network
is calculated by taking the following parameters into account.
• Maximum energy consumption and maximum throughput
• Maximum energy consumption and minimum throughput
• Minimum energy consumption and maximum throughput
• Minimum energy consumption and minimum throughput
Four points are taken in order to draw the feasible region between energy consump-
tion and throughput. Figure 5.8 shows the feasible region for energy consumption
and throughput.

105
15
Number of packets received at sink

P4
10
P3
5
P1
P2
0
0 5000 10000 15000
Energy consumption
Figure 5.8: Feasible Region for Energy Consumption and Throughput using
GEDPAR

PAR using GEDPAR
Feasible region for energy consumption and PAR is shown in Figure 5.9. The four
points for the feasible region of energy consumption and PAR are taken as:
• Maximum energy consumption and maximum PAR
• Maximum energy consumption and minimum PAR
• Minimum energy consumption and maximum PAR
• Minimum energy consumption and minimum PAR

Delay using GEDPAR
Figure 5.10 depicts the coordinates of feasible region for energy consumption and
delay. We take four points between energy consumption and delay to draw the
feasible region.

15000
P4
10000 P3
PAR
5000
P1
P2
0
0 5000 10000 15000
Energy consumption
Figure 5.9: Feasible Region for Energy Consumption and PAR using GED-
PAR
• Maximum energy consumption and maximum delay
• Maximum energy consumption and minimum delay
• Minimum energy consumption and maximum delay
• Minimum energy consumption and minimum delay

Throughput using E2EVHR
The feasible region between energy consumption and throughput of the network
is calculated by taking the following parameters into account.
• Maximum energy consumption and maximum throughput
• Maximum energy consumption and minimum throughput
• Minimum energy consumption and maximum throughput
• Minimum energy consumption and minimum throughput

15000
P4
10000
P3
Delay
5000
P1
P2
0
0 5000 10000 15000
Energy consumption
Figure 5.10: Feasible Region for Energy Consumption and Delay using GED-
PAR
Four points are taken in order to draw the feasible region between energy consump-
tion and throughput. Figure 5.11 shows the feasible region for energy consumption
and throughput.
25
Number of packets received at sink
20 P4
15
10
P1
5
P3
P2
0
-5
0 5 10 15 20 25
Energy consumption
Figure 5.11: Feasible Region for Energy Consumption and Throughput using
E2EVHR


PAR using E2EVHR
Feasible region for energy consumption and PAR is shown in Figure 5.12. The
four points for the feasible region of energy consumption and PAR are taken as:
• Maximum energy consumption and maximum PAR
• Maximum energy consumption and minimum PAR
• Minimum energy consumption and maximum PAR
• Minimum energy consumption and minimum PAR
25
20
P4
15
PAR
10
P1
5
P3
P2
0
-5
0 5 10 15 20 25
Energy consumption
Figure 5.12: Feasible Region for Energy Consumption and PAR using
E2EVHR

Delay using E2EVHR
Figure 5.13 depicts the feasible region for energy consumption and delay. We take
four points between energy consumption and delay to draw the feasible region.

CHAPTER 5. 5.3. SIMULATIONS FOR BLOCKCHAIN TRUST MODEL
• Maximum energy consumption and maximum delay
• Maximum energy consumption and minimum delay
• Minimum energy consumption and maximum delay
• Minimum energy consumption and minimum delay
30
25
P4
20
15
Delay
10
P1
5
P2 P3
0
-5
0 5 10 15 20 25
Energy consumption
Figure 5.13: Feasible Region for Energy Consumption and Delay using
E2EVHR
5.3 Simulations for Blockchain Trust Model
In this section, we will discuss the simulation results which are obtained using
different tools and the reasons for these results. In section 5.3.1, simulation tools
are explained. Moreover, results and their reasoning will be discussed in section
5.3.2.
5.3.1 Simulation Tools
In order to take the simulation results, we have used multiple tools. Four tools are
reviewed for developing and testing the smart contract. First, we write a smart
contract on Remix IDE. Second, Ganache is used to show the clear deployment
visualization of the smart contract. Third, to connect the Etherium node with
browser, MetaMask which is the extension of Chrome browser, is used. Fourth,
execution cost and transaction cost of different transactions are obtained from
Remix IDE and later plotted with MATLAB.
5.3.1.1 Remix IDE
It is an open source tool which allows you to debug, test and compile smart
contract from browser. IDE stands for the integrated development environment
and it helps programmers for designing the software and other different tasks
relating to software development.
5.3.1.2 Ganache
Ganache provides a clear visualization of the smart contract deploying transac-

tions. Ganache grants you access to 10 accounts and each having 100 Ethers
for test purpose. When a transaction or a smart contract is deployed on the
blockchain, Ganache immediately confirms this transaction or smart contract. In
the result of any transaction or smart contract deployment, the transaction log
is increased and ether is deducted. Each transaction has details of fund trans-
fer, contract creation and contract call, etc. along with the sender’s address and
transaction hash.
5.3.1.3 MetaMask
MetaMask is a Google Chrome extension which connects the Etherium node and
browser. It allows us to send and receive ethers from Ganache. For this, the
connection of Ganache wallet with MetaMask is possible by using a private key.
MetaMask is also connected with Remix IDE. When Remix IDE and MetaMask
are connected, ether will be deducted on each transaction.

5.3.1.4 MATLAB R2018a
MATLAB R2018a is a comfortable tool for researchers to work within. It is a

multi-dimensional mathematical computing environment used to solve linear pro-
gramming problems within seconds. However, we have used this tool for getting
plots by passing the values which were obtained by the Remix IDE.
5.3.2 Simulations Reasoning
In this section, the simulation tools, results and reasoning of these results will be
discussed. Nevertheless, we provide an overview of some important terms which
require for understanding the term gas price and about its calculation.
5.3.2.1 Execution Cost
Execution cost is a cost which requires the gas as a fuel on the execution of the
functions (code lines). It also requires storage allocation for different variables as
a result of the execution of operations. On the other hand, the transaction cost
requires for sending the data on the blockchain.
5.3.2.2 Estimating Transaction Cost
The total ether cost of a transaction is based on the following two factors:
• Gas used: it can be defined as total gas consumed by the transaction.
• Gas price: it is a price (in ether) of one unit of gas specified in the transaction.
This transaction cost can be calculated by the following formula:
Costtotal = Ugas × Pgas . (5.1)
Where, Ugas and Pgas represent the gas used in transaction and price specified for
that transaction, respectively.

5.3.2.3 Why Ether is Not Used Instead of Gas?
Actually, it is made to decouple the cost of any operation from the market price of
ether. As we know that the cryptocurrency prices are volatile and ether also has
no exception. The gas limit for each action is constant and this the reason why
we use gas instead of ether.
106
6
4
Cost (gas)
0
Transaction Cost Execution Cost
Figure 5.14: Network Deployment Cost
104
3
2.5
2 Transaction Cost
Execution Cost
Cost (gas)
1.5
0.5
0
Current states of Sink nodes Current states of Sink nodes on main server
Figure 5.15: State Checking Cost of Sink Nodes

Table 5.2: Network Deployment Cost
Parameter Value
status 0x1 Transaction mined and execution succeed
transaction hash 0x98933, ..., 75f15
contract address 0x08970, ..., 659fb
from 0xca35b, ..., a733c
to Clustering.(constructor)
gas 300000000 gas
transaction cost 5674465 gas
execution cost 4272505 gas
hash 0x98933, ..., 75f15
input 0x608...40029
decoded input {}
decoded output -
logs []
value 0 wei
104 Transaction Cost Execution Cost

3.5
2.5
Cost (gas)
1.5
0.5
0
Sink Node 1 Sink Node 2 Sink Node 3 Sink Node 4 Sink Node 5
Figure 5.16: Individual State Checking Cost of Sink Nodes

104 Transaction Cost Execution Cost

5
Cost (gas)
3
0
Sink Node 1 Sink Node 2 Sink Node 3 Sink Node 4 Sink Node 5
Figure 5.17: Transaction Cost of Sink Nodes
Table 5.3: Current Active/De-active State of Nodes
Parameter Value
transaction hash 0xfbaef, ..., 840f96
from 0xca35b7, ..., fa733c
to Clustering.StateOfSNs() 0xbbf28, ..., 732db
transaction cost 27886 gas (Cost only applies when called by a con-
tract)
execution cost 6614 gas (Cost only applies when called by a con-
tract)
hash 0xfbaef, ..., 840f96
input 0x38e, ..., fccad
decoded input {}
decoded output {
“0”: “string: Current state of Sink Node 1 is: 1”,
“5”: “string: State of selected Ordinary Node is:
1”
}
logs []

Table 5.4: Current Active/De-active State of Nodes on Main Server
Parameter Value
transaction hash 0x8f442, ..., 7434c0
from 0xca35b, ..., fa733c
to Clustering.StateOfSNonMainServer() 0xbbf28, ...,
732db
transaction cost 27911 gas (Cost only applies when called by a con-
tract)
execution cost 6639 gas (Cost only applies when called by a con-
tract)
hash 0x8f442, ..., 7434c0
input 0x44e, ..., ce4a7
decoded input {}
decoded output {
“5”: “string: State of selected Ordinary Node is:
1”
}
logs []
Simulations for the calculation of execution and transaction costs are performed in
Remix IDE. During simulations, it is observed that the transaction cost is always
high with the comparison of execution cost. We perform a cost analysis for a
smart contract using the aforementioned costs in term of gas consumption. Figure
5.14 shows the transaction and execution costs required for network deployment.
We can see from the Figure 5.14 that the operation for network deployment is the
most expensive. It can also be seen from the Table 5.2 that the transaction cost
for network deployment is 5674465 gas which is higher than the execution cost
(4272505 gas).
Figure 5.15 depicts the state checking cost of sink nodes. State checking cost
means that the cost which is used for ensuring the active or de-active status of
sink nodes. To check whether the sink nodes are active or not, their status has
been checked by other sink nodes and then by the main server. This active or
de-active states must be the same in both places. If the status is same then the

node is working perfectly, otherwise, the node is malicious and need to be removed
from the network. This state checking cost is presented in Figure 5.15. Similarly,
Figure 5.16 shows the transaction and execution costs for each sink node‘s state.
Moreover, status for each node is presented in Figure 5.16 and this status is cross-
verified by the main server as shown in Table 5.4. Nevertheless, notable thing is
that the active status of nodes is represented by “1” and vice versa by “0”.
Transaction and execution costs per transaction are shown in Figure 5.17. We can
easily observe the difference between both of aforementioned costs. Transaction
cost is always high and execution cost is low. We have already discussed the
reasons for the higher transaction cost than the execution cost.
Table 5.5: Comparison of Different Costs
Transaction Cost Execution Cost

Function
(gas) (gas)
Network Deployment Cost 5674465 4272505
Per Transaction Cost
Sink Node 1 46325 23837
Sink Node 2 46019 23595
Sink Node 3 46237 23749
Sink Node 4 46193 23705
Sink Node 5 46457 23969
Current State (Active/De-active) of Sink Nodes 27886 6614
Current State (Active/De-active) of Sink Nodes According to Main Server 27911 6639
Individual State Checking Cost
Sink Node 1 30427 9155
Sink Node 2 30493 9221
Sink Node 3 31964 10692
Sink Node 4 32118 10846
Sink Node 5 30515 9243
5.3.2.4 Comparison between Transaction and Execution Costs
We execute different functions and calculate the transaction and execution costs
in terms of gas. After that, these costs are plotted using MATLAB. Moreover,
the aforementioned costs for different functions are provided in Table 5.2, Table
5.3 and Table 5.4, respectively. Nevertheless, we also provided comprehensive
comparison in Table 5.5.

The detailed discussions of simulation results are part of this chapter. The adap-
tivity of this proposed work is tested by comparing with the state of the art
routing protocols. Moreover, the transaction and execution costs for our proposed
blockchain based trust model are also calculate using different tools. The compar-
ison of these costs are also held in this chapter. Nevertheless, in Chapter 6, the
conclusion of this thesis along with future directions is bestowed.

Chapter 6
Conclusion and Future Work
61
CHAPTER 6. 6.1. CONCLUSION
6.1 Conclusion
In this thesis, imbalance and unnecessary energy dissipation are avoided by cov-
ering the void hole in an efficient way. We proposed two routing protocols namely
GEDPAR and E2EVHR for void hole recovery. In order to show the efficiency of
the proposed protocols, comparative analysis is performed with the existing state
of the art protocols: GEDAR and LMPC. Simulations result show that GED-
PAR outperforms GEDAR and LMPC in terms of throughput by the percentage
of 13% and 37%. PAR of GEDAR, LMPC, E2EVHR and GEDPAR is 0.6, 0.8,
0.93 and 0.95, respectively. While E2EVHR also outperforms the counterparts in
terms of throughput, PAR and energy efficiency. However, the proposed protocol
(GEDPAR) is minimizing the energy consumption at the cost of affordable delay.
Security is one of the emerging and big challenge in the field of sensor networks.
So in this thesis, security issue is tackled using the concepts of blockchain. Now,
blockchain based trust model for providing the security in sensor networks is pro-
posed. This model provides the security from malicious nodes using the concepts
of blockchain. It also ensures no loss of data, due to the blockchain’s properties of
distributed ledger and immutability.
Future Work
In the future, we will explore and implement some other routing protocols to
avoid void holes. To implement these techniques on a test bed for getting more
precision in results will be the new direction of our research. An idea about
the implementation of “IoTs” is also under consideration. Moreover, we can also
implement blockchain in any state of the art routing protocol and compare the
performance with original one.

Chapter 7
References
63
References
[1] Wang, Hua, Yingyou Wen, Yao Lu, Dazhe Zhao and Chunlin Ji. “Secure Lo-
calization Algorithms in Wireless Sensor Networks: A Review.” In Advances in
Computer Communication and Computational Sciences, pp. 543-553. Springer,
Singapore, (2019).
[2] Latif, Kamran, Nadeem Javaid, Ashfaq Ahmad, Zahoor Ali Khan, Nabil Al-
rajeh, and Majid Iqbal Khan. “On energy hole and coverage hole avoidance in
underwater wireless sensor networks.” IEEE Sens. J 16, no. 11 (2016).
[3] Khasawneh, Ahmad, Muhammad Shafie Bin Abd Latiff, Omprakash Kai-
wartya and Hassan Chizari. “A reliable energy-efficient pressure-based routing
protocol for the underwater wireless sensor network.” Wireless Networks 24,
no. 6 (2018).
[4] Hong, Zhen, Xiaoman Pan, Ping Chen, Xianchuang Su, Ning Wang and Wenqi
Lu. “A Topology Control with Energy Balance in Underwater Wireless Sensor
Networks for IoT-Based Application.” Sensors 18, no. 7 (2018).
[5] Wang, Hao, Shilian Wang, Eryang Zhang and Luxi Lu. “An Energy Balanced
and Lifetime Extended Routing Protocol for Underwater Sensor Networks.”
Sensors 18, no. 5 (2018).
[6] Khan, Anwar, Ihsan Ali, Atiq Ur Rahman, Muhammad Imran and Hasan
Mahmood. “Co-EEORS: Cooperative Energy Efficient Optimal Relay Selection
Protocol for Underwater Wireless Sensor Networks.” IEEE Access (2018).
[7] Coutinho, Rodolfo WL, Azzedine Boukerche, Luiz FM Vieira and Antonio
AF Loureiro. “Underwater Wireless Sensor Networks: A New Challenge for
Topology Control–Based Systems.” ACM Computing Surveys (CSUR) 51, no.
1 (2018).
[8] Jin, Zhigang, Ning Wang, Yishan Su and Qiuling Yang. “A Glider-Assisted
Link Disruption Restoration Mechanism in Underwater Acoustic Sensor Net-
works.” Sensors 18, no. 2 (2018).
[9] Nayyar, Anand, Vikram Puri and Dac-Nhuong Le. “Comprehensive Analysis
of Routing Protocols Surrounding Underwater Sensor Networks (UWSNs).”
In Data Management, Analytics and Innovation, pp. 435-450. Springer, Singa-
pore, (2019).
64
CHAPTER 7. REFERENCES
[10] Wu, Fei-Yun, Kunde Yang and Rui Duan. “Compressed Sensing of Underwa-
ter Acoustic Signals via Structured Approximation l0 Norm.” IEEE Transac-
tions on Vehicular Technology (2018).
[11] Khosravi, Mohammad Reza, Hamid Basri and Habib Rostami. “Efficient rout-
ing for dense UWSNs with high-speed mobile nodes using spherical divisions.”
The Journal of Supercomputing 74, no. 2 (2018).
[12] Gomathi and Martin Leo Manickam. “Energy Efficient Shortest Path Routing
Protocol for Underwater Acoustic Wireless Sensor Network.” Wireless Personal
Communications 98, no. 1 (2018).
[13] Hou, Rui, Liuting He, Shan Hu and Jiangtao Luo. “Energy-Balanced Unequal
Layering Clustering in Underwater Acoustic Sensor Networks.” IEEE Access
6 (2018).
[14] Iwata, Masanari, Suhua Tang and Sadao Obana. “Energy-Efficient Data Col-
lection Method for Sensor Networks by Integrating Asymmetric Communica-
tion and Wake-Up Radio.” Sensors 18, no. 4 (2018).
[15] Muhammed, Dalhatu, Mohammad Hossein Anisi, Mahdi Zareei, Cesar

Vargas-Rosales and Anwar Khan. “Game Theory-Based Cooperation for Un-
derwater Acoustic Sensor Networks: Taxonomy, Review, Research Challenges
and Directions.” Sensors 18, no. 2 (2018).
[16] Jan, Mian Ahmad, Zhiyuan Tan, Xiangjian He, and Wei Ni. “Moving Towards
Highly Reliable and Effective Sensor Networks.” Ad Hoc & Sensor Wireless
Networks 40, no. 3-4 (2018).
[17] Yildiz, Huseyin Ugur, Vehbi Cagri Gungor, and Bulent Tavli. “Packet size op-
timization for lifetime maximization in underwater acoustic sensor networks.”
IEEE Transactions on Industrial Informatics 15, no. 2 (2019).
[18] Khalid, Muhammad, Yue Cao, Naveed Ahmad, Waqar Khalid, and Piyush
Dhawankar. “Radius-based multipath courier node routing protocol for acous-
tic communications.” IET Wireless Sensor Systems 8, no. 4 (2018).
[19] Sher, Arshad, Aasma Khan, Nadeem Javaid, Syed Ahmed, Mohammed Aal-
salem, and Wazir Khan. “Void Hole Avoidance for Reliable Data Delivery in
IoT Enabled Underwater Wireless Sensor Networks.” Sensors 18, no. 10 (2018).
[20] Yuan, Yali, Chencheng Liang, Megumi Kaneko, Xu Chen and Dieter Hogrefe.
“Topology Control for Energy-Efficient Localization in Mobile Underwater
Sensor Networks using Stackelberg Game.” arXiv preprint arXiv:1805.12361
(2018).
[21] Rahman, Ziaur, Fazirulhisyam Hashim, Mohd Fadlee A. Rasid and Mohamed
Othman. “Totally opportunistic routing algorithm (TORA) for underwater
wireless sensor network.” PloS one 13, no. 6 (2018).

[22] Heidemann, John, Milica Stojanovic and Michele Zorzi. “Underwater sensor
networks: applications, advances and challenges.” Phil. Trans. R. Soc. A 370,
no. 1958 (2018).
[23] Javaid, Nadeem, Abdul Majid, Arshad Sher, Wazir Khan, and Mohammed
Aalsalem. “Avoiding Void Holes and Collisions with Reliable and Interference-
Aware Routing in Underwater WSNs.” Sensors 18, no. 9 (2018).
[24] Coutinho, Rodolfo WL, Azzedine Boukerche, Luiz FM Vieira, and Antonio
AF Loureiro. “Geographic and opportunistic routing for underwater sensor
networks.” IEEE Transactions on Computers 65, no. 2 (2016).
[25] Xu, Junfeng, Keqiu Li, Geyong Min, Kai Lin, and Wenyu Qu. “Energy-
efficient tree-based multipath power control for underwater sensor networks.”
IEEE Transactions on Parallel and Distributed Systems 23, no. 11 (2012).
[26] Ramezan, Gholamreza, and Cyril Leung. “A Blockchain-Based Contractual

Routing Protocol for the Internet of Things Using Smart Contracts.” Wireless
Communications and Mobile Computing (2018).
[27] Ahmed, Farwa, Zahid Wadud, Nadeem Javaid, Nabil Alrajeh, Mohamad
Souheil Alabed, and Umar Qasim. “Mobile Sinks Assisted Geographic and
Opportunistic Routing Based Interference Avoidance for Underwater Wireless
Sensor Network.” Sensors 18, no. 4 (2018).
[28] MacGillivray and Gorman, “Connecting the IoT:The Road to Success,” In-
ternational Data Corporation (IDC) Report, (2018).
[29] Butt, Suhail Ashfaq, Abdul Mateen, Nadeem Javaid, and Zahoor Ali Khan.
“Towards the void hole alleviation for energy efficiency in WSN.” In 2018 Fifth
HCT Information Technology Trends (ITT), pp. 318-324. IEEE, (2018).
[30] Mateen, Abdul, Muhammad Awais, Nadeem Javaid, Farruh Ishmanov,

Muhammad Khalil Afzal, and Saqib Kazmi. “Geographic and Opportunis-
tic Recovery with Depth and Power Transmission Adjustment for Energy-
Efficiency and Void Hole Alleviation in UWSNs.” Sensors 19, no. 3 (2019).
[31] Liu, Mengting, F. Richard Yu, Yinglei Teng, Victor CM Leung, and Mei Song.
“Computation ooading and content caching in wireless blockchain networks
with mobile edge computing.” IEEE Transactions on Vehicular Technology 67,
no. 11 (2018).
[32] Xu, Yang, GuojunWang, Jidian Yang, Ju Ren, Yaoxue Zhang, and Cheng
Zhang. Towards Secure Network Computing Services for Lightweight Clients
Using Blockchain.“ Wireless Communications and Mobile Computing 2018
(2018).
[33] Lin, Jun, Zhiqi Shen, Chunyan Miao, and Siyuan Liu. Using “blockchain to
build trusted lorawan sharing server.” International Journal of Crowd Science
1, no. 3 (2017).

[34] Ren, Yongjun, Yepeng Liu, Sai Ji, Arun Kumar Sangaiah, and Jin Wang.
“Incentive Mechanism of Data Storage Based on Blockchain for Wireless Sensor
Networks.” Mobile Information Systems 2018 (2018).
[35] Li, Jiao. “Data Transmission Scheme Considering Node Failure for
Blockchain.” Wireless Personal Communications103, no. 1 (2018).
[36] Lin, Di, and Yu Tang. “Blockchain Consensus Based User Access Strategies
in D2D Networks for Data-Intensive Applications.” IEEE Access 6 (2018).
[37] Singh, Madhusudan, and Shiho Kim. “Branch based blockchain technology
in intelligent vehicle.” Computer Networks 145 (2018).
[38] Dai, Mingjun, Shengli Zhang, Hui Wang, and Shi Jin. “A low storage room
requirement framework for distributed ledger in blockchain.” IEEE Access 6
(2018).
[39] Rosa, Raphael, and Christian Esteve Rothenberg. “Blockchain-Based Decen-

tralized Applications for Multiple Administrative Domain Networking.” IEEE
Communications Standards Magazine 2, no. 3 (2018).
[40] Jia, Bing, Tao Zhou, Wuyungerile Li, Zhenchang Liu, and Jiantao Zhang. “A
Blockchain-Based Location Privacy Protection Incentive Mechanism in Crowd
Sensing Networks.” Sensors 18, no. 11 (2018).
[41] Zhang, Yu, and Jiangtao Wen. “The IoT electric business model: Using
blockchain technology for the internet of things.” Peer-to-Peer Networking and
Applications 10, no. 4 (2017).
[42] Xu, Chenhan, Kun Wang, Peng Li, Song Guo, Jiangtao Luo, Baoliu Ye, and
Minyi Guo. “Making big data open in edges: A resource cient blockchain-based
approach.” IEEE Transactions on Parallel and Distributed Systems (2018).
[43] Jiang, Tigang, Hua Fang, and Honggang Wang. “Blockchain-based Internet
of vehicles: distributed network architecture and performance analysis.” IEEE
Internet of Things Journal (2018).
[44] Sharma, Pradip Kumar, Seo Yeon Moon, and Jong Hyuk Park. “Block-VN: A
Distributed Blockchain Based Vehicular Network Architecture in Smart City.”
JIPS 13, no. 1 (2017).
[45] Zhang, Guozhen, Tong Li, Yong Li, Pan Hui, and Depeng Jin. “Blockchain-
Based Data Sharing System for AI-Powered Network Operations.” Journal of
Communications and Information Networks 3, no. 3 (2018).
[46] Novo, Oscar. “Scalable Access Management in IoT using Blockchain: a Per-
formance Evaluation.” IEEE Internet of Things Journal (2018).
[47] Sharma, Pradip Kumar, Saurabh Singh, Young-Sik Jeong, and Jong Hyuk
Park. “Distblocknet: A distributed blockchains-based secure sdn architecture
for iot networks.” IEEE Communications Magazine 55, no. 9 (2017).

[48] Pãnescu, Adrian-Tudor, and Vasile Manta. “Smart Contracts for Research
Data Rights Management over the Ethereum Blockchain Network.” Science &
Technology Libraries 37, no. 3 (2018).
[49] Kushch, Sergii, and Francisco Prieto-Castrillo. “A Rolling Blockchain for a

Dynamic WSNs in a Smart City.” arXiv preprint arXiv:1806.11399 (2018).
[50] Rahmadika, Sandi, Diena Rauda Ramdania, and Maisevli Harika. “Security
Analysis on the Decentralized Energy Trading System Using Blockchain Tech-
nology.” Jurnal Online Informatika 3, no. 1 (2018).
[51] Liu, Xiaojun, Wenbo Wang, Dusit Niyato, Narisa Zhao, and Ping Wang.
“Evolutionary game for mining pool selection in blockchain networks.” IEEE
Wireless Communications Letters7, no. 5 (2018).
[52] Gordon, William, and Christian Catalini. “Blockchain technology for health-
care: facilitating the transition to patient-driven interoperability.” Computa-
tional and structural biotechnology journal 16 (2018).
[53] Alabi, Ken. “Digital blockchain networks appear to be following Metcalfe’s

Law.” Electronic Commerce Research and Applications 24 (2017).
[54] Sharma, Pradip Kumar, and Jong Hyuk Park. “Blockchain based hybrid net-
work architecture for the smart city.” Future Generation Computer Systems
86 (2018).
[55] Yang, Zhe, Kan Yang, Lei Lei, Kan Zheng, and Victor CM Leung.
“Blockchain-based decentralized trust management in vehicular networks.”
IEEE Internet of Things Journal (2018).

CHAPTER 7. JOURNAL PUBLICATIONS
Journal Publications
1 Mateen, Abdul, Muhammad Awais, Nadeem Javaid, Farruh Ishmanov, Muham-

mad Khalil Afzal, and Saqib Kazmi. “Geographic and opportunistic recovery
with depth and power transmission adjustment for energy-efficiency and void
hole alleviation in UWSNs.” Sensors 19, no. 3 (2019): 709.
2 Ahmad, Ashfaq, Nadeem Javaid, Abdul Mateen, Muhammad Awais, and

Zahoor Khan. “Short-term load forecasting in smart grids: an intelligent
modular approach.” Energies 12, no. 1 (2019): 164.

CHAPTER 7. CONFERENCE PROCEEDINGS
Conference Proceedings
1 Mateen, Abdul, Nadeem Javaid, Muhammad Awais, Nasir Khan, Urva Latif,
and Ihtisham Ullah. “Bio-Inspired Optimization Techniques for Home En-
ergy Management in Smart Grid.” In 2018 32nd International Conference on
Advanced Information Networking and Applications Workshops (WAINA),
pp. 250-257. IEEE, 2018.
2 Mateen, Abdul, Nadeem Javaid, Muhammad Bilal, Muhammad Arslan Fa-

rooq, Zahoor Ali Khan, and Fareena Riaz. “Towards Optimizing Energy
Efficiency and Alleviating Void Holes in UWSN.” In International Confer-
ence on Emerging Internetworking, Data & Web Technologies, pp. 516-527.
Springer, Cham, 2019.
3 Awais, Muhammad, Abdul Mateen, Zahoor Ali Khan, and Nadeem Javaid.
“A Survey on Multi-hop Routing Protocols for Efficient Resource Alloca-
tion in IoTs.” In 2018 Fifth HCT Information Technology Trends (ITT), pp.
239-244. IEEE, 2018.
4 Butt, Suhail Ashfaq, Abdul Mateen, Nadeem Javaid, and Zahoor Ali Khan.
“Towards the void hole alleviation for energy efficiency in WSN.” In 2018
Fifth HCT Information Technology Trends (ITT), pp. 318-324. IEEE, 2018.
5 Awais, Muhammad, Nadeem Javaid, Abdul Mateen, Nasir Khan, Ali Mo-
hiuddin, and Malik Hassan Abdul Rehman. “Meta Heuristic and Nature
Inspired Hybrid Approach for Home Energy Management Using Flower Pol-
lination Algorithm and Bacterial Foraging Optimization Technique.” In 2018
IEEE 32nd International Conference on Advanced Information Networking
and Applications (AINA), pp. 882-891. IEEE, 2018.
6 Awais, Muhammad, Zahoor Ali Khan, Nadeem Javaid, Abdul Mateen, Ay-
men Rasul, and Farooq Hassan. “Cluster-Based Routing Protocols with
Adaptive Transmission Range Adjustment in UWSNs.” In International
Conference on Emerging Internetworking, Data & Web Technologies, pp.

CHAPTER 7. CONFERENCE PROCEEDINGS
528-539. Springer, Cham, 2019.
7 Khan, Nasir, Nadeem Javaid, Muhammad Khan, Ahmed Subhani, Abdul

Mateen, and Arshad Iqbal. “Harmony Pigeon Inspired Optimization for
Appliance Scheduling in Smart Grid.” In 2018 IEEE 32nd International Con-
ference on Advanced Information Networking and Applications (AINA), pp.
1060-1069. IEEE, 2018.
8 Latif, Urva, Nadeem Javaid, Syed Shahab Zarin, Muqaddas Naz, Asma Ja-
mal, and Abdul Mateen. “Cost Optimization in Home Energy Management
System using Genetic Algorithm, Bat Algorithm and Hybrid Bat Genetic
Algorithm.” In 2018 IEEE 32nd International Conference on Advanced In-
formation Networking and Applications (AINA), pp. 667-677. IEEE, 2018.
9 Javaid, Nadeem, Ihtisham Ullah, Syed Shahab Zarin, Mohsin Kamal, Ba-
batunji Omoniwa, and Abdul Mateen. “Differential-Evolution-Earthworm
Hybrid Meta-heuristic Optimization Technique for Home Energy Manage-
ment System in Smart Grid.” In International Conference on Innovative Mo-
bile and Internet Services in Ubiquitous Computing, pp. 15-31. Springer,
Cham, 2018.
View publication stats

21 Mateen MSThesis

Uploaded by

Copyright:

Available Formats

21 Mateen MSThesis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

21 Mateen MSThesis

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Towards Energy Efﬁcient Routing in Blockchain Based Underwater WSNs via

Thesis · September 2019

Abdul Mateen Sohail Iqbal

SEE PROFILE SEE PROFILE

Smart Grid View project

TEFL Certified View project

The user has requested enhancement of the downloaded file.

COMSATS University Islamabad, Islamabad - Pakistan

Towards Energy Efficient Routing in Blockchain

COMSATS University Islamabad

A Post Graduate Thesis submitted to the Department of Computer Science as

Name Registration Number

Dr. Nadeem Javaid,

Dr. Sohail Iqbal,

This thesis titled

Towards Energy Efficient Routing in Blockchain Based Underwater

has been approved

I Hafiz Abdul Mateen (Registration No. CIIT/FA17-RCS-009/ISB) hereby declare

Date: 20 July, 2019

It is certified that Hafiz Abdul Mateen (Registration No. CIIT/FA17-RCS-009/ISB)

Dr. Nadeem Javaid,

Dr. Sohail Iqbal,

Dr. Majid Iqbal Khan,

Foremost, I must express my profound gratitude to my parents. It is because of

1 Mateen, Abdul, Muhammad Awais, Nadeem Javaid, Farruh Ishmanov, Muham-

2 Ahmad, Ashfaq, Nadeem Javaid, Abdul Mateen, Muhammad Awais, and

2 Mateen, Abdul, Nadeem Javaid, Muhammad Bilal, Muhammad Arslan Fa-

7 Khan, Nasir, Nadeem Javaid, Muhammad Khan, Ahmed Subhani, Abdul

List of Figures xvi

List of Tables xvii

List of Algorithms xviii

List of Abbreviations xix

List of Symbols xxi

1.1.1 Thesis Contributions . . . . . . . . . . . . . . . . . . . . . . 5

1.2 Thesis Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.3 Conclusion of the Chapter . . . . . . . . . . . . . . . . . . . . . . . 5

2 Related Work and Problem Statement 6

2.1 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.1.1 Energy Efficiency . . . . . . . . . . . . . . . . . . . . . . . . 7

2.1.3 Topology Control . . . . . . . . . . . . . . . . . . . . . . . . 8

2.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

2.3 Conclusion of the Chapter . . . . . . . . . . . . . . . . . . . . . . . 17

3.1 Proposed System Model 1 . . . . . . . . . . . . . . . . . . . . . . . 20

3.1.1 End to End Adaptive Transmission Range and Depth Ad-

3.1.1.1 Successful Transmission . . . . . . . . . . . . . . . 20

3.1.1.2 Void Hole Occurrence . . . . . . . . . . . . . . . . 21

3.1.1.3 Void Hole Recovery . . . . . . . . . . . . . . . . . . 21

3.1.1.4 Establishing the Link with Neighbors . . . . . . . . 22

3.1.1.5 Forwarder Node Selection . . . . . . . . . . . . . . 24

3.2 Proposed System Model 2 . . . . . . . . . . . . . . . . . . . . . . . 25

3.2.1 Ordinary Sensor Nodes . . . . . . . . . . . . . . . . . . . . . 26

3.2.2 Sink Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.2.3 Main Server . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.2.3.1 Hash Function and Merkle Tree . . . . . . . . . . . 27

3.3 Conclusion of the Chapter . . . . . . . . . . . . . . . . . . . . . . . 28

4 Existing and Proposed Schemes 29

4.5 Conclusion of the Chapter . . . . . . . . . . . . . . . . . . . . . . . 38

5.1 Simulations for Routing Protocols . . . . . . . . . . . . . . . . . . . 40

5.1.4 Depth Adjustment . . . . . . . . . . . . . . . . . . . . . . . 40