Physical Layer Security for Wireless Transmission

Abstract

Due to rapid growth in wireless technology, wireless networks have become indispensable in applications such as Internet of Things (IoT). The broadcast nature of wireless channels makes information exchange between legitimate nodes vulnerable to the eavesdrop ping. In addition to the conventional solution to secure information at application layer, physical layer security exploits channel ran domness in order to protect information from the eavesdropping. This work proposes methods to secure information from the active attacks of the adversary. First, we consider physical layer security for wireless communica tion systems under pilot contamination attack (PCA). The detec tion of PCA by an active Eve is essential for secure communica tion. We propose novel PCA detectors for legitimate nodes, which intend to establish secure communication in time division duplex (TDD) manner over a frequency selective wireless channel. We treat PCA detection as a binary hypotheses problem using the decision directed channel estimates. For PCA detection, we provide perfor mance analysis of the proposed PCA detectors. We also present simulation results and comparison with the performance analysis to demonstrate the accuracy of the analysis. We then present secret key generation at physical layer by exploit ing intrinsic channel randomness under PCA. For key generation under PCA, we use two-phase channel estimate using decoded pay load data as reference to mitigate the impact of PCA. The proposed secret key generation method achieves low key disagreement prob ability (KDP) under PCA. We also derive closed form expression DRSML QAU for KDP under PCA. We also present precoder design from the data aided channel estimate under PCA in data phase to enhance secrecy capacity of the reciprocal channels of the legitimate nodes. The pro posed method achieves high secrecy capacity in the presence of an active Eve. Finally, we propose a successive pilot contamination cancellation method to remove the impact of pilot contamination in the pilot phase. In addition to the PCA in pilot phase, Eve can also trans mit jamming signal in the data phase. We propose a successive pilot contamination and jamming removal approach in pilot and data phases, respectively. The proposed method enhances the se crecy capacity of the legitimate nodes by effectively estimating the reciprocal channel. The precoder design from the estimated chan nel steers information towards legitimate node and artificial noise (AN) in the null-space of the user in order to enhance the secrecy capacity of the legitimate channels