Research Repository

Optimal Power Allocation in Block Fading Channels with Confidential Messages

Chorti, A and Papadaki, K and Poor, HV (2015) 'Optimal Power Allocation in Block Fading Channels with Confidential Messages.' IEEE Transactions on Wireless Communications, 14 (9). 4708 - 4719. ISSN 1536-1276

[img]
Preview
Text
optimal_power_allocation.pdf - Accepted Version

Download (909kB) | Preview

Abstract

© 2016 IEEE. The optimal power allocation for block fading (BF) networks with confidential messages is investigated under an $M$-block delay and power constraint. First, we study networks without channel state information (CSI) feedback to the transmitter and demonstrate that the optimal power allocation is the equidistribution of the power budget, denoted as the 'blind policy.' In blind scenarios secrecy can be achieved through receiver diversity; the probability of secrecy outage (PSO) is shown to decay exponentially with the diversity order of the legitimate user. Then, we investigate networks with CSI feedback. For comparison purposes, we restate the acausal secure waterfilling algorithm with full CSI before moving to the causal feedback scenario. In the latter, an approximate 'threshold policy' for the low SNR and an approximate 'high power policy' for the high SNR regimes are derived. Furthermore, a novel universal transmission policy is proposed across all SNRs, denoted as the 'blind horizon approximation' (BHA). Through numerical results, the BHA policy is shown to outperform both the threshold and high power policies when the legitimate user has an SNR advantage with respect to the eavesdropper, while it also compares well with the secure waterfilling policy.

Item Type: Article
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Faculty of Science and Health > Computer Science and Electronic Engineering, School of
Depositing User: Jim Jamieson
Date Deposited: 23 Aug 2015 15:52
Last Modified: 17 Aug 2017 17:34
URI: http://repository.essex.ac.uk/id/eprint/14631

Actions (login required)

View Item View Item