Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
Opportunistic Persistent Data Storage- Interim Presentation
1. Opportunistic Persistent Data Storage
Author: Weerasooriya W.A.A.C.P.
Index Number: 11002192 Supervisor: Dr. Ranasinghe D.N.
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
2. Outline
• Motivation
• Research Problem
• From Proposal defense
• Design & Architecture
• Our Contribution
• Related Work
• To-Do
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
3. • Opportunistic Networks are type of Mobile Ad-
hoc Networks(MANET).
• Message delays are accepted.
• There are a lot of routing protocol mechanisms
that have been designed for ON
• Synthetic traces are widely used to evaluate
ON E.g. : ONE simulator
Motivation
There are a lot of papers which describes virtual
data storages over distributed systems. We can find
several storage protocol implementations for
MANET as well. But…..
According to the best of our knowledge there is not any
storage mechanism which uses
opportunistic properties
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
5. Outline
• Motivation
• Research Problem
• From Proposal defense
• Design & Architecture
• Our Contribution
• Related Work
• To-Do
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
6. Modeling a persistent data storage on
opportunistic networks making maximum
use of opportunistic properties
Research Goal
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
7. Research Questions
Q1: What is the best way to capture Social network
properties in ON?
Q2: What is the best persistent storage mechanism for
MANET that can be adjusted to use with opportunistic
properties of nodes?
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
8. Scope
• Synthetic traces- limitations( no. of subjects, simulations area/time)
Assumptions: It’s better to see the problem in the easiest way and deriving
a solution. Then we can go more deeper in to the problem, removing
assumptions one by one.
1. No peer failures
2. Number of nodes are fixed
3. One Writer- Many Readers
4. No reading while writing
5. Write globally- Read locally
6. Security etc.
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
9. Outline
• Motivation
• Research Problem
• From Proposal defense
• Design & Architecture
• Our Contribution
• Related Work
• To-Do
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
10. From Proposal Defense
•BitTorrent Protocol: In the very early phase of
our problem we only consider single variables.
After successful evaluation of this system we
try to focus on stream data.
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
11. Outline
• Motivation
• Research Problem
• From Proposal defense
• Design & Architecture
• Our Contribution
• Related Work
• To-Do
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
14. Outline
• Motivation
• Research Problem
• From Proposal defense
• Design & Architecture
• Our Contribution
• Related Work
• To-Do
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
15. Our Contribution
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
Virtual
Storage
Routing+ON
properties
MAC
PHYSICAL
Storage
Storage
Virtual
Storage
Routing+ON
properties
MAC
PHYSICAL Storage
SCORP
dLIFE
Data dissemination
Replica management
16. Outline
• Motivation
• Research Problem
• From Proposal defense
• Design & Architecture
• Our Contribution
• Related Work
• To-Do
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
17. Related Work- Our Path
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
Opportunistic
Persistent Data
Storage
Quorum based
Pub/sub
Social network properties
Routing for MANET
Mobility models
Content dissemination
18. Outline
• Motivation
• Research Problem
• From Proposal defense
• Design & Architecture
• Our Contribution
• Related Work
• To-Do
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
19. To-Do List: Implementation
• Implementation of the Routing protocol which uses Opportunistic
properties.
• Implementation of the Virtual Storage
• Our task should be carried out towards stream data also
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
20. To-Do list: Evaluation
• Evaluate the success rate: latest copy percentage
• Time to get the latest copy
• Energy consumption
• We have to evaluate our model using a routing protocol which does
not use opportunistic properties.
• At the same time have to evaluate using a mobility model which
does not exhibit opportunistic behavior: Random way point
September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
21. September 03, 2015 Weerasooriya W.A.A.C.P. 11002192
Thank you!
Your review will be golden!