On the Impact of Adaptive Eavesdroppers in Multi-Antenna Cellular Networks

In this paper, the impact of adaptive smarter eavesdroppers on the secrecy performance of a multi-antenna cellular network is investigated. The eavesdroppers can act as either passive eavesdroppers or active jammers based on their distance to the active base stations (BSs). To analyze the secrecy performance, the BSs, cellular users, and eavesdroppers are modeled as independent Poisson point processes. The closed-form expressions of the connection outage probability and lower bound of the secrecy outage probability are derived using the stochastic geometry approach. Following that, the conditions under which the eavesdroppers can act as active jammers are obtained. Finally, the optimal power allocation between artificial noise and information signal and the secrecy code rate at each BS, as well as the guard zone of the eavesdroppers are obtained using the Stackelberg game approach, where the eavesdroppers are modeled as the leader and the BSs as the follower. The Stackelberg equilibrium is obtained through the proposed iterative algorithm. Numerical results verify the theoretical analysis and show that the secrecy performance can be degraded severely by the adaptive eavesdroppers.