Abstract
In this paper, a novel anti-jamming mechanism is proposed to analyze and enhance the security of adversarial Internet of Battlefield Things (IoBT) systems. In particular, the problem is formulated as a dynamic psychological game between a soldier and an attacker. In this game, the soldier seeks to accomplish a time-critical mission by traversing a battlefield within a certain amount of time, while maintaining its connectivity with an IoBT network. The attacker, on the other hand, seeks to find the optimal opportunity to compromise the IoBT network and maximize the delay of the soldier's IoBT transmission link. The soldier and the attacker's psychological behavior are captured using tools from psychological game theory, with which the soldier's and attacker's intentions to harm one another are considered in their utilities. To solve this game, a novel learning algorithm based on Bayesian updating is proposed to find a $\epsilon$-like psychological self-confirming equilibrium of the game. Simulation results show that, based on the error-free beliefs on the attacker's psychological strategies and beliefs, the soldier's material payoff can be improved by up to 15.11% compared to a conventional dynamic game without psychological considerations.
Introduction
The Internet of Battlefield Things (IoBT) represents a paradigm shift in modern warfare, integrating soldiers, sensors, and autonomous vehicles to provide unprecedented situational awareness. However, the reliance on wireless communication makes these systems highly vulnerable to jamming attacks. Traditional security models often rely on "perfectly rational" game theory, such as the Nash Equilibrium, which assumes players optimize objective outcomes like data rates or power consumption.
In a high-stakes battlefield environment, human combatants are driven by psychological factors that traditional models overlook—specifically, the desire to frustrate the opponent. A soldier is not merely optimizing a data link; they are avoiding the psychological impact of a failed transmission they expected to succeed. Conversely, an attacker may prioritize jamming the most reliable link to maximize the soldier's frustration.
Methodology: Dynamic Psychological Games
The core of this research is the formulation of interactions as a dynamic psychological game. This framework introduces several key innovations:
- Psychological Utility Functions: Unlike standard games, utility here includes a "frustration" term, quantified as the positive gap between a player's expected payoff and their actual payoff.
- Second-Order Beliefs: To calculate frustration, players form beliefs about their opponent's beliefs (e.g., "What does the attacker think I am expecting?").
- Bayesian Updating Algorithm: To address the computational complexity of psychological equilibria, a learning algorithm is employed. Players use Bayesian updating to refine their first- and second-order beliefs over repeated interactions.
- $\epsilon$-like Psychological Self-Confirming Equilibrium (PSCE): This solution concept describes a state where players' beliefs converge to match actual strategies, even without a perfect global view of the opponent's mental state.
Key Insights and Results
The application of psychological game theory reveals strategic shifts that traditional models miss. In a standard Nash Equilibrium, an attacker might distribute jamming power for efficiency. However, in a Psychological Equilibrium (PE), the attacker targets the best available channel specifically because the soldier relies on it most, thereby maximizing frustration.
By anticipating this psychological intent, the soldier can adopt a more robust "evasive" strategy. Key findings from numerical simulations include:
- Performance Gain: Accounting for psychological factors improves the soldier's material payoff (reduced delay) by up to 15.11% over conventional models.
- Convergence: The Bayesian learning algorithm typically reaches an $\epsilon$-like PSCE within 210 iterations.
- Sensitivity to Delay: As the soldier's tolerable transmission delay ($\Delta$) increases, the material payoff at the psychological equilibrium increases by up to 38.11%, providing more flexibility to evade frustration-driven attacks.
Limitations and Future Directions
While this model provides a significant step forward, it focuses primarily on "frustration." Future research could incorporate other psychological factors such as fear, stress-induced cognitive bias, or overconfidence. Additionally, while the Bayesian approach is practical, the exponential growth of the state space with more IoBT devices suggests that advanced reinforcement learning techniques may be necessary for larger-scale network deployments.
Summary updated