Doctoral Dissertations

Abstract

"In environments like international military coalitions on the battlefield or multi-party relief work in a disaster zone, multiple teams are deployed to serve different mission goals by the command-and-control center (CC). They may need to survey damages and send information to the CC for situational awareness and also transfer messages to each other for mission purposes. However, due to the damaged network infrastructure in the emergency, nodes need to relay messages using the store and forward paradigm, also called Delay-tolerant Networks (DTNs). In DTN, the limited bandwidth, energy, and contacts among the nodes, and their interdependency impose several challenges such as sensitive data leakage to malicious nodes, redundant data generation, limited and delayed important message delivery, non-interested messages in storage, etc. We aim to focus on solving these challenges. We propose message fragmentation for secure message transfer because existing public-private-key cryptographic approaches may not work due to the unavailability of Public Key Infrastructure (PKI). Besides, to ensure more message delivery, redundant fragments are generated. However, too much redundancy may consume the energy and bandwidth of the nodes while transferring similar messages. Hence, we propose to send diverse content and limit the redundancy. Again, the dynamic environment we consider is prone to many adverse and sudden events. We aim to respond to these events by sending the event-related message to the CC fast with the help of intermediate nodes. The nodes are interested in certain types of content defined by their mission and interest. Therefore, we target to learn nodes' interests using Reinforcement Learning so that the nodes can populate themselves with the messages according to their mission requirements and increase the collaboration among them. Our future work will include machine learning techniques for predicting important places where node encounters the most and to cache data for each other according to their interest, encounter frequency, and encounter locations"--Abstract, p. iv

Advisor(s)

Madria, Sanjay Kumar

Committee Member(s)

Luo, Tony T.
Cen, Nan
Das, Sajal K.
Sarangapani, Jagannathan

Department(s)

Computer Science

Degree Name

Ph. D. in Computer Science

Publisher

Missouri University of Science and Technology

Publication Date

Fall 2022

Pagination

xvii, 190 pages

Note about bibliography

Includes_bibliographical_references_(pages 183-189)

Rights

© 2022 Shudip Datta, All Rights Reserved

Document Type

Dissertation - Open Access

File Type

text

Language

English

Thesis Number

T 12189

Download

Share

 
COinS