Deception framework- Multiple indexing in self adaptation

Abstract

In this thesis we propose deception as a security mechanism. In deception an attacker's behavior is manipulated by sending him misleading information. One of the critical phases in deception is self-adaptation where the defender has to adapt to the changing pattern of the attacker and respond accordingly. However, determining the change pattern must be efficient and accurate. In this thesis we propose a novel algorithm that is based on a windowing scheme and a tolerance value. Our approach aims to find the optimum window size and tolerance value from an efficiency and accuracy perspective. In this thesis we derive the optimum window size and optimum tolerance value which is adaptable to the reference (query sequence) or attack. Simulation results show that using our approach, execution time with optimum window value is less when compared to a fixed window size. Similarly, identifying a change in pattern is more accurate with the proposed optimum tolerance than with a fixed tolerance value.