Please use this identifier to cite or link to this item:
Title: A Secure Framework for Flying Ad-Hoc Networks
Authors: Singh, Kuldeep
Supervisor: Verma, Anil Kumar
Keywords: UAVs;FANETs;Trust;Security;FCTM;Threat model
Issue Date: 24-Aug-2018
Abstract: The use of Unmanned Aerial Vehicles (UAVs) has significantly increased for forming an ad hoc network due to their astonishing working capabilities in the tactical as well as civilian areas such as armed attacks, border surveillance, disaster management, rescue operation and transportation. These UAVs aided ad hoc networks popularly known as Flying Ad hoc Networks (FANETs). A FANET consist a group of homogeneous or heterogeneous flying agents called Unmanned Aerial Vehicles (UAVs) connected by wireless links. UAVs are capable to communicate with each other in a group, and interact with their neighboring nodes to acquire valuable information. In FANETs, UAVs performs not only as a client, but also have the responsibility to perform as a router and server in the network. The large degree of freedom and self-organising capabilities makes FANETs totally different from any other network solution. FANETs can be rapidly deployed anytime in real-time for communication because they do not require any external resources. These unique features make FANETs a suitable solution for various applications. In FANETs, intermediate UAVs are responsible for forwarding the data packets of other network’s UAVs and enable them to communicate out of their transmission range. In FANETs, UAVs are commonly constrained by limited energy and computation power. So, intermediate nodes can refuse to forward packets in order to preserve their resources, leads to disruption in network traffic. Nodes showing such behavior are known as selfish nodes. Selfishness is generally a passive behavior of the node, and effect the overall performance of the network by degrading it. FANETs are more prone to attacks due to the self-organized, anonymous, distributive and independent behavior of nodes.Trust management can be one of the security solutions for FANETs because to accomplish the task/mission with cooperation and coordination, nodes need to have trust on each other. Trust management schemes/techniques allowed a node to assess trustworthiness of other network nodes. A trust management technique helps in the detection of malicious and selfish node behavior and also enhances the overall network performance by segregating them. Measurement parameters and computation approach to evaluate trust for secure communication among nodes is crucial for the development of trust management technique. Trust evaluation in FANETs environment involves several intricate aspects like - node behavior assessment in terms of reliability and performance, correct recommendation etc. Motivated from the above, a fuzzy based trust model is proposed for FANETs. This model makes use of principles of fuzzy classification and optimization to manage trust relationships among nodes of FANETs. Trust is calculated based on the quality of service (QoS) (includes- reliability and performance parameters) and social parameter (recommendation). Social trust improves the trust evaluation process for those nodes which lacks in direct observation of their neighbor nodes due to some constraints such as hidden terminal problem, bad mounting attack, etc. There are various advantages of trust management- malicious node can be identified and isolated, evading of data forwarding to selfish and malicious nodes and it helps to select a higher trust level node, resulting in improvement of overall performance of FANETs. In the proposed model, network nodes communicate only with its immediate neighbor nodes (hop count=1), based on the fact that only immediate neighbor node can give recommendation correctly. Decay function is also used to replicate the real-world behavior of the UAVs. The node keeps only its neighbor node’s trust information which helps in reducing the processing time for trust evaluation, utilization of energy and memory.
Appears in Collections:Doctoral Theses@CSED

Files in This Item:
File Description SizeFormat 
PhD Thesis.pdf14.34 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.