Date of Graduation


Document Type


Degree Name

Master of Science in Computer Science (MS)

Degree Level



Computer Science & Computer Engineering


Brajendra Nath Panda

Committee Member

Ukash Nakarmi

Second Committee Member

Lu Zhang


Critical Infrastructure, Cyberattack, Damage assessment and recovery, Database security, Optimized damage assessment and recovery, Security


Critical infrastructures (CI) play a vital role in majority of the fields and sectors worldwide. It contributes a lot towards the economy of nations and towards the wellbeing of the society. They are highly coupled, interconnected and their interdependencies make them more complex systems. Thus, when a damage occurs in a CI system, its complex interdependencies make it get subjected to cascading effects which propagates faster from one infrastructure to another resulting in wide service degradations which in turn causes economic and societal effects. The propagation of cascading effects of disruptive events could be handled efficiently if the assessment and recovery are carried out as quickly as possible. To be an efficient system, it should reduce the impact by reducing the number of nodes undergoing service degradation. In general, the damage assessments include accessing and assessing log information which is very costly in terms of time spent and IO reads. A generic model thus should be very optimal in suggesting smaller number of assessments as possible and at the same time reduce the number of nodes undergoing unnecessary service degradations. This thesis investigates the CI systems in depth to optimize the damage assessment and recovery process so that it could help in resuming the operations of as many safe data items as quickly as possible. It also focuses on reducing the load imposed in terms of number of nodes towards damage assessment and recovery procedures through the proposed optimization model. The quick identification and categorization of the type of data items as damaged, undamaged, or skeptical within the impacted CI system is the key factor which makes this model highly efficient and helps this model to project better performance. The developed model and its algorithm have been implemented on a simulated data and environment whose results shows that the proposed model performs well in terms of time, speed, accuracy, complexity, efficiency, and performance.