Cooperative Pilot Jamming Strategies for Enhanced Physical-Layer Security

Abstract

Tackling the physical layer secrecy issue, existing studies on cooperative jamming generally assume the persistent jamming and perfect channel estimation at the receiver side. However, these two assumptions are incompatible since jamming signals can corrupt the pilot symbols for estimation. In addition, due to the energy budget limitation, persistent jamming is burdensome to the cooperative jammers. To address these issues, in this thesis, I first derive the achievable secrecy rate, applying channel estimation error effect in the equation, based on the cooperative jammer model that transmits artificial noise persistently in pilot phase and data phase. In this result, I reveal that physical-layer security can be enhanced by pilot jamming, which disrupts channel estimation at the eavesdropper. Moreover, in light of efficient use of jamming power, I propose the scheme of jamming power allocation between pilot jamming and data jamming, aiming at secrecy rate maximization. Since it is cumbersome to attain the cooperative jammer-to-eavesdropper channel information, we here deal with both the perfect channel state information (CSI) case and the imperfect cooperative jammer-to-eavesdropper CSI case to design beamforming at the coopertive jammer and to obtain the optimal jamming power allocation. Furthermore, I propose a cooperative jammer energy efficiency model based on the circuit power consumption model and verify the performance of the proposed jamming power allocation scheme, in terms of secrecy rate and energy efficiency, via numerical results. The result show that the pilot-only jamming scheme gains advantage from the shortened pilot phase duration, relative to data phase duration. Since pilot phase is shorter than data phase in general, it is beneficial to jam only pilot signals in terms of circuit power consumption. Motivated by this fact, in this thesis, I subsequently propose a cooperative jamming scheme that delivers artificial noise only in pilot phase and optimize the corresponding beamforming to maximize secrecy rate by means of difference-of-convex (DC) algorithm and rank relaxation method. Furthermore, to account of energy budget limitation at the cooperative jammer, the energy efficiency of the proposed pilot-only jamming scheme is evaluated. In addition, a performance comparison between the proposed pilot-only jamming scheme and the conventional data-only jamming scheme, which is revised to fit in our framework, is accomplished via computer simulations. Also, motivated by the fact that the pilot sequence can be publicly known from its finite and standardized characteristic, I additionally introduce a deterministic design of pilot jamming symbols to achieve the physical-layer security enhancement. On the other hand, for practical application and implementation, there have been great efforts to attain enhanced physical-layer security without CSI and location information of the eavesdropper. In this context, I finally turn to study a security enhancement in Poisson point process (PPP)-modeled stochastic wireless networks, dealing with a lack of the channel and location knowledge related to the eavesdropper. Specifically, I employ a system model with a protected zone, i.e., an eavesdropper-free area, and consider the worst scenario that the eavesdropper is located on the border of the protected zone. Assuming a pilot-assisted channel estimation at the receiver side, I derive the secrecy transmission rate, applying the channel estimation error effect in it and propose a power allocation between pilot and data transmission which maximizes the secrecy transmission rate. Numerical results show the existence of the optimal power allocation and show that the proposed scheme enhances the secrecy transmission rate.