Link-Signature-Based Discriminatory Channel Estimation (LS-DCE) for Physical Layer Security in Stationary and Mobile OFDM Transceivers

Abdulsahib Albehadili,Khair A. Al Shamaileh,Ahmad Y. Javaid,Vijay K. Devabhakuni
DOI: https://doi.org/10.1109/tvt.2020.2995540
IF: 6.8
2020-08-01
IEEE Transactions on Vehicular Technology
Abstract:In this work, we propose a physical layer security (PLS) scheme for orthogonal frequency-division multiplexing (OFDM) based transceivers, termed link-signature-based discriminatory channel estimation (LS-DCE). The proposed scheme exploits the space-time uniqueness of the channel responses (i.e. link-signatures) at different transceiver pairs to facilitate a reliable signal reception at legitimate receivers, while equivocating reception at non-legitimate adversaries. While most PLS studies focus on transmitted data to provide such an equivocation, LS-DCE focuses on channel estimations. Since the quality of estimates impacts the overall performance, LS-DCE facilitates distinctive detection error probabilities by enforcing a poor channel estimation at non-legitimate receivers as compared to legitimate ones. To validate the effectiveness of the proposed scheme, analytical, simulated, and experimental evaluations are performed. First, a theoretical analysis is established in terms of the probability of bit error (PBE) at the legitimate and non-legitimate receivers. Then, indoor/outdoor stochastic and deterministic channel models are adopted using correlated Nakagami-m simulation model and software-defined radio (SDR) testbed, respectively. With $-10,{rm dB}$ signal-to-noise ratio (SNR), it is found that a spatial correlation $leq 0.1$ between the legitimate and non-legitimate channels results in an average PBE of 0.48, 0.48, and 0.49 for BPSK, QPSK, and 16-QAM, respectively, at the non-legitimate receiver; and 0.45, 0.46, and 0.48 with $text{50},{rm dB}$ SNR.
telecommunications,engineering, electrical & electronic,transportation science & technology
What problem does this paper attempt to address?