Artificial Noise Aided Physical Layer Security in Multi-Antenna Small-Cell Networks

TLDR

Physical layer security in multi‑antenna small‑cell networks is studied with base stations, users, and eavesdroppers modeled as independent Poisson point processes. The study adopts artificial‑noise‑aided transmission at each base station to enhance secrecy performance. Using stochastic geometry, the authors derive closed‑form connection and secrecy outage probabilities, analyze parameter effects, and extend the model to high‑cell‑load scenarios with zero‑forcing beamforming for multi‑user transmission. In low‑cell‑load regimes, adding base stations improves both connection and secrecy outage while extra antennas only reduce connection outage; for fixed‑rate transmission a threshold where artificial noise is unnecessary is identified, a semi‑closed‑form lower bound on average secrecy rate is derived, and the optimal user count maximizing secrecy area spectral efficiency is shown to be a fixed fraction of transmit antennas, all confirmed by simulations.

Abstract

In this paper, physical layer security in multi-antenna small-cell networks is investigated, where the multi-antenna base stations (BSs), cellular users, and eavesdroppers are all randomly distributed according to three independent Poisson point processes. To improve the secrecy performance, artificial noise (AN) aided transmission is adopted at each BS. Based on the stochastic geometry, we first derive the closed-form expressions of the connection and secrecy outage probabilities, and then comprehensively analyze the impact of different parameters through asymptotic analysis. It shows that in a low cell-load case, deploying more BSs will improve the connection and secrecy outage performance, and deploying more transmit antennas at each BS will only improve the connection outage performance. For a fixed-rate transmission, the condition under which AN becomes unnecessary is derived. We also derive a semi closed-form expression of the lower bound of the achievable average secrecy rate, which is numerically efficient to evaluate. Finally, we extend the study to a high cell-load case and adopt the zero-forcing beamforming scheme to support multi-user transmission. The connection and secrecy outage probabilities are also analyzed. Moreover, the optimal number of users maximizing the secrecy area spectral efficiency is discussed, and it is shown to be a fixed portion of the number of transmit antennas. Simulation results are presented to validate the theoretical analysis.