UPPER BOUND ESTIMATION OF THE EXPECTATION OF THE NUMBER OF INFECTED NODES AS A RISK FOR CYBER INSURANCE RATE MAKING ON FINITE GRAPHS

• Main Author: Antonio, Yeftanus
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/46508
• Institution: Institut Teknologi Bandung

Predictions of increased cyber attacks in the next few years make cyber risk estimation become a popular topic today. Viruses, Trojans, and worms spread from one computer to another in a network. The process of spreading disease in Biological populations used to understand the process of spreading viruses on a computer network using an epidemic mathematical model approach. The risk related to the number of infected computers on different computer network topologies was obtained by using a simple stochastic epidemic model, namely, the Susceptible- Infectious-Susceptible (SIS) model, with modified contact parameters. The solution of the Kolmogorov differential equation for expectations of the number of infected computers for the SIS model was an upper bound. This study carried out estimations of upper bound on complete graphs, cycle graphs, wheel graphs, star graphs, and path graphs. The Gillespie algorithm, also known as the Stochastic Simulation Algorithm, was used to compare the upper bound and the samples mean of the number of infections. Cyber insurance rates with the expected value principle on a collective risk model for that graph topology was founded by using simulations approach. Based on the simulation results, increasing the average degree of graphs resulted in higher infections mean than a lower average degree of graphs.