Secrecy Outage Analysis in Random Wireless Networks With Antenna Selection and User Ordering

Gaojie Chen,Justin P Coon

doi:10.1109/lwc.2017.2689024

Abstract

This letter investigates the secrecy outage probability in the downlink with ordered user equipment (UE) based on two ordering metrics. UEs and independently acting eavesdroppers (EDs) are positioned randomly according to a Poisson point process. We consider a transmit antenna selection (TAS) scheme at the base station (BS) to enhance secrecy performance and propose two metrics to order the UEs: one based on long-term average channel gain information from the BS to the UEs, and the other based on instantaneous channel gains. We derive closed form expressions for the secrecy outage probability subject to each of these ordering policies and verify our calculations through Monte Carlo simulations. Our results show that while TAS yields a performance improvement relative to single-antenna systems, the secrecy outage probability for TAS systems actually increases with the path loss exponent. Furthermore, we show that the importance of the specific user ordering policy that is adopted in these systems is reduced for high path loss environments or situations where large numbers of antennas are employed.

Highlights

Physical layer (PHY) security has gained a lot of interest since Wyner’s seminal paper [1]
The basic principle of PHY security is to exploit the inherent randomness of noise and wireless channels to ensure the confidentiality of messages against any eavesdropper (ED) regardless of its computing power [2]
Our results show that while transmit antenna selection (TAS) yields a performance improvement relative to single-antenna systems, the secrecy outage probability for TAS systems increases with the path loss exponent for both policies

Summary

INTRODUCTION

Physical layer (PHY) security has gained a lot of interest since Wyner’s seminal paper [1]. Studies have considered information theoretic security over wireless channels, covering such topics as cooperative relay and jammer networks, buffer-aided relay networks, multiple-input multiple-output communication (MIMO) with distributed beamforming, full-duplex networks, and cognitive radio networks [4]– [8]. All of these contributions focused on a small number of nodes and assumed the locations of EDs are known. We quantify the deterioration in secrecy performance with increasing ordinal UE index, i.e., cycling through the ordered list of UEs from best to worst

System Model

Secrecy Outage Probability

Policy I

Policy II

SIMULATIONS RESULTS