Anda di halaman 1dari 9

iii

ABSTRACT

Wireless sensor networks provide a powerful means to collect information on a


wide variety of natural phenomena. WSNs typically consist of a cluster of densely
deployed nodes communicating with a sink node. During this communication data
from sensor nodes to sink may suffer from channel contention, interference, packet
loss, energy conservation. To overcome these in WSNs there is a need for a better
transport layer protocol. To deliver the data reliably in WSNs, Reliable Multi-Segment
Transport is used.

Directed diffusion is a communication paradigm intended for sensor network


applications. The directed diffusion protocol family has three different diffusion
algorithms. They are One-Phase Pull diffusion, Two-Phase Pull diffusion and Push
diffusion.

The project deals with reliable data transport in sensor network using RMST. It
is a reliable transport layer protocol for WSNs and it is meant to operate on top of the
gradient mechanism used in directed diffusion. RMST uses One-Phase Pull diffusion
algorithm to forward a large blob. RMST adds two important features to directed
diffusion. They are fragmentation/reassembly of segments, and reliable message
delivery. RMST includes caching mode and non-caching mode which provides hop-
by-hop recovery and end-to-end recovery.

RMST maintains a ‘hole map’, which has the list of lost fragments. Whenever
there is a fragment loss, it is notified with the help of the negative acknowledgement.
The advantage of the RMST is to reduce the number of negative acknowledgement by
selectively sending a negative acknowledgement to the entire set of lost fragments.
Due to single NACK the energy usage is reduced. Since One-Phase Pull diffusion
algorithm is used, the flooding of control messages in the sensor network is reduced.
Evaluation is performed for hop-by-hop (cached) recovery and end-to-end (non-
cached) recovery.
iv

ACKNOWLEDGEMENT

I owe a great deal to my “Beloved Parents” who were the confidential


backbone for this project.

I have a great pleasure in expressing a gratitude and hearty thanks to our


Principal Dr.S.Balakrishnan, B.E., M.S., Ph.D., for his unstained support and
encouragement towards the completion of this project work.

I express my gratefulness to our Head of the Department Dr.K.Muneeswaran,


M.E., Ph.D., for his timely help and encouragement for completion of this project.

I convey my sincere regards to Dr. Kannan Balasubramanian, M.Tech.,


Ph.D., who inspired me to embark this project.

I would extend my sincere thanks and express my deep sense of gratitude to my


guide Ms.P.Golda Jeyasheeli M.E., (Ph.D), Asst.Professor, Computer Science
Department for her constant support, guidance and continuous encouragement.

I sincerely thank all the teaching and non-teaching staff for their kind
cooperation during the period of this project.

I would also like to thank my friends for their encouragement and suggestions.

Finally, I thank “THE ALMIGHTY” for showering His blessings to release my


project in the way I visualized.
v

TABLE OF CONTENT

Chapter Topic Page No.


No.
Abstract iii
List of Tables viii
List of Figures ix
List of Abbreviations xi
1. INTRODUCTION 1
1.1 Motivation 1
1.2 Objective 2
1.3 Overview of the project 2
1.4 Report overview 3
2. LITERATURE SURVEY 4
2.1 Data Dissemination 4
2.1.1 Flooding 5
2.1.2 Gossiping 6
2.1.3 Rumor Routing 6
2.1.4 Sequential Assignment Routing 7
2.1.5 Directed Diffusion 7
2.2 Transport Protocols For Sensor Networks 8
2.2.1 Tcp/Ip 8
2.2.1.1 Loss Detection/Recovery 9
2.2.2 Pump Slowly, Fetch Quickly (PSFQ) 9
2.2.2.1 Loss Detection/Recovery 10
2.2.3 Reliable Multi-Segment Transport 11
(RMST)
2.2.3.1 Loss Detection/Recovery 11
Mechanisms
3. DIRECTED DIFFUSION 12
3.1 Interests And Gradients 14
3.2 Data Propagation 15
3.3 Reinforcement For Path Establishment 16
vi

And Pruning
3.3.1 Path Establishment Using Positive 16
Reinforcement
3.3.2 Local Repair of Failed Paths 17
3.3.3 Path Pruning and Negative 18
Reinforcement
3.4 Directed Diffusion Protocol Family 20
3.4.1 Two-Phase Pull Diffusion 20
3.4.2 Push Diffusion 25
3.4.3 One-Phase Pull Diffusion 28
4. RELIABLE MULTISEGMENT TRANSPORT 30
PROTOCOL
4.1 RMST Architecture 30
4.1.1 Overview of RMST 31
4.1.2 Basic Services of RMST 32
4.1.3 Non-Caching Mode and Caching 32
Mode
4.1.4 RMST Support for Loss Detection and 33
Repair
4.1.5 The Back-Channel 34
4.1.6 Node Failure 34
4.1.7 Support for Caching 34

5. IMPLEMENTATION 36
5.1 Overview 36
5.2 One-Phase Pull Diffusion 36
5.3 Reliable Multisegment Transport Protocol 37
5.4 Evaluation Of RMST 38
6. CODING AND SIMULATION RESULTS 39
6.1 Classes And Functions 39
6.1.1 RMST Filter 39
6.1.2 RMST Source 41
vii

6.1.3 RMST Sink 41


6.1.4 One phase pull Diffusion 42
6.2 Emstar Scripts 45
6.3 Simulation Results 45
6.3.1 Interest Dissemination 45
6.3.2 Gradient Establishment 46
6.3.3 Sensor nodes position 47
6.3.4 Data Transmission 48
6.3.5 Data Received 49
6.3.6 Transmission without Holes 50
6.3.7 Transmission with Holes 51
6.3.8 Request for Lost Fragments 52
6.3.1 Evaluation Of RMST 53

7. CONCLUSION AND FUTURE WORK 54


7.1 Conclusion 54
7.2 Future Work 54
APPENDIX 55
REFERENCES 56

LIST OF TABLES

S.No. Table Caption Table No. Page No.


1. RMST Filter 6.1.1 39
2. RMST source 6.1.2 41
viii

3. RMST sink 6.1.3 42


4. One phase pull diffusion 6.1.4 42

LIST OF FIGURES

S.No. Figure Caption Figure No. Page No.


1. Interest propagation 3.1 13
2. Initial gradients set up 3.2 13
ix

3. Data delivery along reinforced path 3.3 14


4. Gradients after the sink reinforces the
3.4 17
empirically lowest delay path.
5. Gradients after local repair caused by
3.5 18
connection quality degradation.
6. Gradients after several rounds of
3.6 19
reinforcement, multiple paths.
7. Sink sends interest message in to network 3.7 22
8. Source receiving interest message 3.8 23
9. Source sends exploratory data 3.9 23
10. Sink receiving exploratory data 3.10 24
11. Sink sending reinforcement message 3.11 24
12. Source sending data to sink 3.12 25
13. Source sends exploratory data 3.13 26
14. Sink receiving exploratory data 3.14 26
15. Sink sending reinforcement message 3.15 27
16. Source sending data to sink 3.16 27
17. Source receiving interest message sent by
3.17 28
sink
18. Source sending data to sink
3.18 29

19. RMST in conjunction with diffusion node 4.1 30


20. Interest messages disseminate through the
6.1 45
network
21. Gradient establishments between the sensor
6.2 46
nodes
22. Positions of the node in the sensor field 6.3 47
23. Source sending the data fragments 6.4 48
24. Sink received the data sent by source 6.5 49
25. Holes not found when perfect model used 6.6 50
26. Holes found when noise introduced. 6.7 51
27. Single NACK sent for the lost fragments 6.8 52
28. Evaluation of RMST in caching mode 6.9 53
29. Evaluation of RMST in non-caching mode 6.10 53
x

LIST OF ABBREVATIONS

WSNs Wireless sensor networks

RMST Reliable Multi-Segment Transport

NACK Negative Acknowledgement

Qos Quality of service

TCP/IP Transport Control Protocol/Internet


Protocol
xi

PSFQ Pump Slowly Fetch Quickly

JPEG Joint Photographic Experts Group

MTU Maximum Transmission Unit

MAC Medium Access Control

Anda mungkin juga menyukai