Optimal Power Allocation for Artificial Noise under Imperfect CSI against Spatially Random Eavesdroppers

TL;DR

This paper derives optimal power allocation strategies for artificial noise in multiantenna systems with imperfect CSI, balancing secrecy outage probability and secrecy rate against spatially random eavesdroppers.

Contribution

It provides the first analytical solutions for power allocation between information signal and artificial noise under imperfect CSI in secure communications.

Findings

Optimal power allocation strategies depend on channel estimation error.

More power to the information signal reduces SOP with rate constraint.

More power to artificial noise increases secrecy rate under SOP constraint.

Abstract

In this correspondence, we study the secure multiantenna transmission with artificial noise (AN) under imperfect channel state information in the presence of spatially randomly distributed eavesdroppers. We derive the optimal solutions of the power allocation between the information signal and the AN for minimizing the secrecy outage probability (SOP) under a target secrecy rate and for maximizing the secrecy rate under a SOP constraint, respectively. Moreover, we provide an interesting insight that channel estimation error affects the optimal power allocation strategy in opposite ways for the above two objectives. When the estimation error increases, more power should be allocated to the information signal if we aim to decrease the rate-constrained SOP, whereas more power should be allocated to the AN if we aim to increase the SOP-constrained secrecy rate.