(A) study on physical layer security for multiple-input multiple-output wireless networks

Abstract

Security is an important issue in wireless networks as they are vulnerable to eavesdropping by third parties due to the broadcast nature of wireless medium. Traditionally, wireless networks are secured by conventional cryptographic technologies which only provide computational security based on the assumption that an attacker has limited computational power and lacks efficient algorithms to decipher secret messages. Unlike conventional cryptographic technologies, physical layer security is able to provide unconditional or perfect security, which is measured in an information theoretic framework. In this dissertation, we consider the physical layer security for multiple-input multiple-output (MIMO) systems and explore two research directions, the secure communication problem and the secret key agreement problem in MIMO systems. First, we study an artificial noise (AN) scheme for securing wireless communications over multiple-input multiple-output multi-eavesdropper (MIMOME) channels. It is assumed that the full channel state information (CSI) of the legitimate receiver is known to the legitimate transmitter while only the statistics of CSI of the eavesdropper are available at the legitimate transmitter. Based on these assumptions, we derived an insightful lower bound on the achievable ergodic secrecy rate which explicitly shows the impact of system parameters, such as the number of antennas at all terminals and the signal-to-noise ratio at the legitimate receiver, on the growth rate of the achievable ergodic secrecy rate. Furthermore, we derived an optimal power allocation between information signal and AN based on the tight lower bound. Let the number of antennas at the legitimate receiver and eavesdropper be $N_B$ and $N_E$, respectively. Then, our analytic results show that we need to put $\sqrt{N_B}/(\sqrt{N_B} + \sqrt{N_E})$ of the total transmit power to the information signal to maximize the ergodic secrecy rate. We carried out extensive numerical evaluations under various system configurations, and the results confirm our analytic results. Second, we present a secret key agreement (SKA) protocol for a multi-user time-division duplex system where a base-station (BS) with a large antenna array (LAA) shares secret keys with users in the presence of non-colluding eavesdroppers. In the system, when the BS transmits random sequences to legitimate users for sharing common randomness, eavesdroppers can attempt a pilot contamination attack (PCA) in which each eavesdropper transmits its target user's training sequence in hopes of acquiring a possible information leak by steering a beam towards the eavesdropper. We show that there is a crucial complementary relation between the received signal strengths at the eavesdropper and its target user. This relation tells us that the eavesdropper inevitably leaves a trace that enables us to devise a way of measuring the amount of information leakage to the eavesdropper even if the PCA parameters are unknown. To this end, we derive an estimator for the channel gain from the BS to the eavesdropper and propose a rate-adaptation scheme for adjusting the secret key length under PCA. Extensive analysis and evaluations are carried out under various setups, which show that the proposed scheme adequately takes advantage of the LAA to establish the secret keys under PCA. Lastly, we study how to incorporate two-way training, i.e. conducting training processes in both the up-link and down-link transmissions, into the SKA protocol for the purpose of strengthening the proposed SKA protocol against PCA and/or improving secret key generation efficiency. It will be shown that the two-way training provides the SKA protocol with a considerable enhancement of estimation for an eavesdropper's channel. The secrecy performances of the proposed SKA protocol is quantitatively analyzed in terms of secrecy outage probability and average secret key rate, which shows that the two-way training improves the security performance especially in fast fading environments.