Defender-Attacker Games with Asymmetric Player Utilities.

Area of Science:

• Systems Security
• Game Theory
• Risk Analysis

Background:

• Strategic attackers pose significant threats to system security, necessitating advanced defense mechanisms.
• Existing defense models often assume symmetric defender-attacker utilities, which may not reflect real-world scenarios.
• Divergent attacker and defender utilities can lead to suboptimal defense strategies, especially when attacker objectives are unknown.

Purpose of the Study:

• To investigate defense strategies against strategic attackers with potentially asymmetric utilities.
• To compare the effectiveness of a conservative defense strategy (unknown attacker utility) versus an optimal strategy (known attacker utility).
• To determine conditions under which a conservative defense strategy remains optimal.

Main Methods:

• Modeling a system defense scenario with asymmetric utilities: defender's utility as system survivability and attacker's utility as expected destroyed elements.
• Analyzing attack strategies under both symmetric and asymmetric utility assumptions.
• Comparing the conservative defense strategy against the optimal defense strategy.

Main Results:

• A conservative defense strategy can be optimal even when the attacker's utility is unknown to the defender.
• This optimality holds under asymmetric utilities when the contest intensity is less than one.
• The findings highlight the robustness of conservative strategies in uncertain security environments.

Conclusions:

• Conservative defense strategies offer a reliable approach against strategic attackers, particularly when their objectives are not fully known.
• System survivability can be maintained even with asymmetric utilities under specific conditions (contest intensity < 1).
• The research underscores the importance of considering utility asymmetry in designing effective security protocols.