Jamming and Anti-Jamming Strategies of Mobile Vehicles

Anti-jamming games have become a popular research topic. However, there are not many publications devoted to such games in the case of vehicular ad hoc networks (VANETs). We considered a VANET anti-jamming game on the road using a realistic driving model. Further, we assumed the quadratic power function in both vehicle and jammer utility functions instead of the standard linear term. This makes the game model more realistic. Using mathematical methods, we expressed the Nash equilibrium through the system parameters in single-channel and multi-channel cases. Since the network parameters are usually unknown, we also compared the performance of several reinforcement learning algorithms that iteratively converge to the Nash equilibrium predicted analytically without having any information about the environment in the static and dynamic scenarios.

Automated summary

Research on anti-jamming games in VANETs using realistic driving models with quadratic power functions in utility functions. Mathematical methods used to find Nash equilibrium in single-channel and multi-channel scenarios. Comparison of reinforcement learning algorithms in static and dynamic scenarios without environment information.