Security Information Technology of Physical Layer Based on Channel-tap Power for Mobile OFDM Systems and Cognitive Radio Networks

Abstract

This dissertation investigates security information technologies for mobile orthogonal frequency division multiplexing (OFDM) systems and OFDM-based cognitive radio (CR) networks Traditionally security is viewed as an independent feature addressed above the physical layer (PHY) All widely used cryptographic protocols are established assuming the PHY layer is merely used to provide an error-free link However with the emergence of ad-hoc and decentralized networks higher-layer security techniques are complex and hard to be implemented To complement and enhance traditional security mechanism we study novel schemes for OFDM systems and CR networks based on the channel power-delay profile (PDP) considering channel time and frequency selectivities More specifically the PHY layer employing the properties of OFDM signals over time-variant and multipath fading channels is aimed to assist the overall authentication process To identify different transmitters (TXs) we employ the hypothesis test based on their PDPs to distinguish different transmission terminals in OFDM systems and primary user emulation attacks (PUEA) in OFDM-based CR networks The proposed PDP estimation is obtained based on the redundancy of cyclic prefix (CP) which is a common feature for almost all OFDM systems In OFDM systems it is advantageous to use the PDP as the wireless signature because it depends on the surrounding environment and is therefore considered to be very hard to mimic In OFDM-based CR networks channel-tap power is utilized as a radio-frequency fingerprint (RF) to directly detect users via PHY layer To improve the detection performance of PHY layer in fading channels the cooperative detection schemes using the fixed sample size test (FSST) and the sequential probability ratio test (SPRT) are devised for CR networks Although different users can be distinguished it is still impossible to exactly tell identity of a TX as primary user (PU) or PUEA using only PHY layer Hence to accurately know identities of PUs and PUEAs the cross-layer intelligent learning ability of a mobile secondary user (SU) is exploited to establish detection databases by seamlessly combining the quick detection of PHY layer with the accuracy of higher layer authentication The proposed methods helps PHY layer completely detect the identities of PUs and PUEAs Finally the performances are analyzed and simulations confirm the advantages of the proposed methods

Date of Award	2015 Jun 26
Original language	English
Supervisor	Wen-Long Chin (Supervisor)