Secrecy enhancement with artificial noise in decentralized wireless networks: A stochastic geometry perspective

Artificial noise has been considered as an effective and efficient way to improve the secrecy of wireless communications. However, most existing papers have been focused on the analysis of a single communication link between a transmitter and a legitimate receiver with some eavesdroppers. In this paper, we investigate the impact of artificial noise to the secrecy of a large scale decentralized wireless network where channels are modeled as Rayleigh fading and locations of legitimate nodes and eavesdroppers follow Poisson point processes. To study network throughput versus system parameters, we derive the secrecy transmission capacity and show how it is related to the number of transmitting antennas and the power allocation ratio between data and artificial noise. Analytical results provide insights on how system parameters can be designed to improve secrecy transmission capacity of a wireless network.