Game-based analysis of eavesdropping defense strategy in WMN with directional antenna

This paper is concerned with the framework for finding the optimal stochastic routing to defend intelligent eavesdropping attacks in the Wireless Mesh Network (WMN) with a cone-based, directional antenna. A game-theoretical model is used to find the best strategy to maximize the expected number of secure sessions (ES) under the most severe circumstance. In particular, the WMN has been modeled by the two-person constant-sum game with two players, namely, a network defender and a network eavesdropper. Numerical results show that the enhancement of antenna directionality by decreasing the beamwidth can help improve ESS only in the sufficiently large network case of downlink. In addition, the optimal defending strategy at the proposed game equilibrium has been found to give a better ESS improvement in the downlink case than in the uplink case.