Physical layer security is to enable confidential data transmission through wireless networks in the presence of illegitimate users, without basing on higher-layer encryption. The essence of physical layer security is to maximize the secrecy rate, that is the maxi- mum rate of information without intercepted by the eavesdroppers. Besides, the design of physical layer security considers the transmit power minimization. These two objectives conflict with each other. Consequently, the research on physical layer security designs often focuses on the two main classes of optimization problems: maximizing secrecy rate under the transmit power constraint and minimizing power consumption while guaranteeing the secrecy rate constraint. These problems are nonconvex, thus, hard to solve. In this thesis, we focus on developing optimization approaches to solve these two optimization problem classes. Our methods are based on DC (Difference of Convex functions) programming and DCA (DC Algorithm) which well-known as one of the most powerful approaches in optimization. In the first part, we consider three classes of secrecy rate maximization problems (chapters 2, 3, 4). In particular, chapter 2 studies the secure information transmission in a multiple-input single-output (MISO) relay system by using joint beamforming and artificial noise strategy under the deterministic uncertainty channel models of all links. Without using a relay, chapter 3 addresses the problem of transfer wireless information and power simultaneously in MISO secure system where scenarios of perfect channel state information and deterministic uncertainty channel models are concerned. Transmit beamforming without artificial noise and that with artificial noise are investigated. Under the assumption of statistical channel state information to eavesdroppers, chapter 4 studies the probability constrained secrecy rate maximization problem in multiuser MISO simultaneous wireless information and power transfer (SWIPT) system. The unified approach based on DC programming and DCA is proposed to solve three classes of optimization problems. The optimization problem in chapter 2 is recast as two general DC programs. The general DCA schemes are proposed to solve these two DC programs. In chapter 3, we consider four optimization problems in accordance with four scenarios. Exploiting the special structures of these original optimization problems, we transform it into four general DC programs for which the corresponding general DCA based algorithms are developed. In chapter 4, we first transform the considered problem into a tractable form. We then develop an alternating scheme to solve the transformed problem. Two general DC programs are proposed in each step of the alternating scheme. For solving these DC programs, we study a variant of general DCA, namely, DCA−ρ scheme. The convergence of alternating general DCA−ρ scheme is proven. The second part studies the transmit power optimization problem under the probability constraints of secrecy rate and harvested energy in a MISO SWIPT system (chapter 5). We reformulate the original problem as three general DC programs for which the corresponding general DCA-based algorithms are investigated. Numerical results demonstrate the efficiency of the proposed algorithms.
Physical layer security is to enable confidential data transmission through wireless networks in the presence of illegitimate users, without basing on higher-layer encryption. The essence of physical layer security is to maximize the secrecy rate, that is the maxi- mum rate of information without intercepted by the eavesdroppers. Besides, the design of physical layer security considers the transmit power minimization. These two objectives conflict with each other. Consequently, the research on physical layer security designs often focuses on the two main classes of optimization problems: maximizing secrecy rate under the transmit power constraint and minimizing power consumption while guaranteeing the secrecy rate constraint. These problems are nonconvex, thus, hard to solve. In this thesis, we focus on developing optimization approaches to solve these two optimization problem classes. Our methods are based on DC (Difference of Convex functions) programming and DCA (DC Algorithm) which well-known as one of the most powerful approaches in optimization. In the first part, we consider three classes of secrecy rate maximization problems (chapters 2, 3, 4). In particular, chapter 2 studies the secure information transmission in a multiple-input single-output (MISO) relay system by using joint beamforming and artificial noise strategy under the deterministic uncertainty channel models of all links. Without using a relay, chapter 3 addresses the problem of transfer wireless information and power simultaneously in MISO secure system where scenarios of perfect channel state information and deterministic uncertainty channel models are concerned. Transmit beamforming without artificial noise and that with artificial noise are investigated. Under the assumption of statistical channel state information to eavesdroppers, chapter 4 studies the probability constrained secrecy rate maximization problem in multiuser MISO simultaneous wireless information and power transfer (SWIPT) system. The unified approach based on DC programming and DCA is proposed to solve three classes of optimization problems. The optimization problem in chapter 2 is recast as two general DC programs. The general DCA schemes are proposed to solve these two DC programs. In chapter 3, we consider four optimization problems in accordance with four scenarios. Exploiting the special structures of these original optimization problems, we transform it into four general DC programs for which the corresponding general DCA based algorithms are developed. In chapter 4, we first transform the considered problem into a tractable form. We then develop an alternating scheme to solve the transformed problem. Two general DC programs are proposed in each step of the alternating scheme. For solving these DC programs, we study a variant of general DCA, namely, DCA−ρ scheme. The convergence of alternating general DCA−ρ scheme is proven. The second part studies the transmit power optimization problem under the probability constraints of secrecy rate and harvested energy in a MISO SWIPT system (chapter 5). We reformulate the original problem as three general DC programs for which the corresponding general DCA-based algorithms are investigated. Numerical results demonstrate the efficiency of the proposed algorithms.
The information theoretic description of secrecy capacities and secrecy capacity regions (Liang et al., 2009) is an important research area to support a better understanding of security on the physical layer. Based on the secrecy capacity expressions or achievable secrecy rates, the transmit strategies, including power allocation, beamforming and subcarrier allocation, are optimized in order to choose a certain operating point. In this chapter, we focus on the optimization of the physical layer transmit strategies for typical wireless communication scenarios. In single-user scenarios, the system design is more complicated with additional secrecy constraints, since the secrecy capacity expressions are in general not concave or convex in the transmit strategies. The secrecy rate terms usually consist of a difference of two parts. The first one corresponds to the amount of data that can be reliably transmitted to the intended user and it is therefore concave in the transmit strategies. The second one corresponds to the amount of data that is overheard by the eavesdropper and it is thus also concave in the transmit strategies. The resulting transmit optimization problems are non-convex optimization problems since the difference of two concave functions is not necessarily convex or concave. However, in the multi-carrier case, the problem can be reduced to a convex optimization problem that can be efficiently computed. We conjecture that also the multiple antenna (MIMO) scenario will be completely solved in the very next future. In multi-user scenarios, the secrecy rate regions of all four elements of network information theory, the broadcast, the multiple access, the relay, and the interference channel were recently studied. The attacker models of the MAC and the relay case are more difficult than the wellmotivated ones of the broadcast and the interference channel. Therefore, we focus on the broadcast and interference channel. The resource allocation for the parallel broadcast channel without secrecy is involved due to a hard combinatorial problem the matching of carriers to users. Interestingly, with secrecy constraints, the resulting programming problem is much simpler and the optimal power and resource allocation can be solved efficiently. A similar
In this paper, we consider a wireless network where a source tries to send messages to a destination while keeping them secret from multiple eavesdroppers. To enhance secrecy, the transmission is performed with a cooperation of friendly jammers which make artificial noise to confound eavesdroppers and beamforming technique to direct jamming signal to intended targets. The purpose is to determine beamforming coefficients in order to maximize secrecy rate under the total jammer power constraint. The existing method solved this problem in a special case in which the beamforming coefficients were designed to completely eliminate jamming signal at the destination and a heuristic approach was applied to obtain a suboptimal solution for this case. We address this problem via a new approach based on DC (Difference of Convex functions) programming and DCA (DC Algorithm). We first reformulate this problems as a general DC program, i.e. minimizing a DC function under some DC constraints, and develop a DCA based algorithm for solving it. The experimental results show that the secrecy rate obtained by the proposed algorithm is better than those achieved by the existing one.
Physical layer security is a promising technique for information security. In multi-antenna systems, secure transmission designs are of crucial importance for physical layer security. In this thesis, our research focuses on the secure transmission designs for jamming signal aided multi-antenna systems. We first investigate robust secure transmission designs for a cooperative jamming aided multiple-input-single-output (MISO) system. Two scenarios are considered: (a) eavesdroppers' channel state information (ECSI) is available and (b) ECSI is unavailable. In scenario (a), a quality-of-service (QoS)-based design is considered to minimize the worst case signal-to-interference-and-noise ratio at the eavesdroppers. A secrecy rate based design is also studied. In scenario (b), a QoS-based design is considered to maximize the power of jamming signals, and the secrecy rate based design is not applicable. We propose an algorithm for each design problem through semidefinite relaxation. Our analysis shows that, even though the beamforming scheme in our designs is fixed as single stream beamforming, this does not cause any loss of optimality. We further consider secure transmission designs for a simultaneous wireless information and power transfer enabled MISO heterogeneous cellular network. In the considered system, all base stations (BSs) send confidential messages to information receivers and energy signals to energy receivers (ERs). BSs collaboratively utilize the energy signals as jamming signals to cripple ERs' interceptions. We propose a sum logarithmic secrecy rate maximization design problem. To tackle the design problem, we propose a semidefinite relaxation and successive convex approximation based centralized secure transmission design algorithm. Moreover, an alternating direction method of multipliers based distributed secure transmission design is also proposed. Simulation results demonstrate the effectiveness of the proposed algorithms. Secure transmission is designed for a cooperative jamming aided MISO non-orthogonal multiple access system. We consider an outage constrained secure transmission design problem under the assumption that only statistical ECSI is available. In the design problem, the minimum probabilistic secrecy rate is maximized. Based on 0-norm and the path-following approach, we develop an algorithm to find a solution to the design problem. Simulation results demonstrate that the secrecy outage probability can be guaranteed through the proposed secure transmission design algorithm.
Consider a multiple-input multiple-output (MIMO) beamforming system, where a multi-antenna base station transmits information and energy simultaneously to a multi-antenna information receiver (IR) and a number of multi-antenna energy receivers (ERs). The ERs are assumed to possess dual functionality such that they can also decode information. This gives rise to the physical layer security issue of providing confidential information to the IR while ensuring efficient energy harvesting for the ERs. Hence, we employ an artificial noise (AN)-aided secrecy approach at the transmitter, where the AN improves the security by interfering with the information decoding at the ERs while, at the same time, providing them with wireless energy to harvest. In this paper, we address the problem of jointly optimizing the beamforming matrix for the IR and the covariance matrix of the AN such that the secrecy rate is maximized subject to energy harvesting constraints and a total power constraint. The corresponding optimization problem is a difference of convex functions (DC) programming problem, which is generally non-convex. Nevertheless, we propose an alternating optimization (AO) strategy to tackle this problem for the cases of a square and a non-square IR channel. The performance of the proposed algorithms is demonstrated by simulation results.
Information security is one of the most critical issues in wireless networks as the signals transmitted through wireless medium are more vulnerable for interception. Although the existing conventional security techniques are proven to be safe, the broadcast nature of wireless communications introduces different challenges in terms of key exchange and distributions. As a result, information theoretic physical layer security has been proposed to complement the conventional security techniques for enhancing security in wireless transmissions. On the other hand, the rapid growth of data rates introduces different challenges on power limited mobile devices in terms of energy requirements. Recently, research work on wireless power transfer claimed that it has been considered as a potential technique to extend the battery lifetime of wireless networks. However, the algorithms developed based on the conventional optimization approaches often require iterative techniques, which poses challenges for real-time processing. To meet the demanding requirements of future ultra-low latency and reliable networks, neural network (NN) based approach can be employed to determine the resource allocations in wireless communications. \n \nThis thesis developed different transmission strategies for secure transmission in wireless communications. Firstly, transmitter designs are focused in a multiple-input single-output simultaneous wireless information and power transfer system with unknown eavesdroppers. To improve the performance of physical layer security and the harvested energy, artificial noise is incorporated into the network to mask the secret information between the legitimate terminals. Then, different secrecy energy efficiency designs are considered for a MISO underlay cognitive radio network, in the presence of an energy harvesting receiver. In particular, these designs are developed with different channel state information assumptions at the transmitter. Finally, two different power allocation designs are investigated for a cognitive radio network to maximize the secrecy rate of the secondary receiver: conventional convex optimization framework and NN based algorithm.
This paper addresses robust designs of secure simultaneous wireless information and power transfer for multiple-input single-output (MISO) multiuser systems. By using a bounded uncertainty model for the channel state information (CSI) from transmitters to legitimate receivers and under the assumption of only available statistic knowledge of the CSI to eavesdroppers, we consider the robust designs with both worst-case and chance-constrained optimizations. To jointly optimize transmit beamforming vectors and power splitting factors, we formulate three problems aiming at minimizing the system power, maximizing the sum secrecy rate, and maximizing the minimal secrecy rate, respectively, subject to both energy harvesting and secrecy outage constraints. These problems are intractable in their original forms due to the CSI models and their non-convexity. To reach computationally efficient solutions, we first rewrite the original problems by semi-definite relaxation, then transform the CSI uncertainty related constraints by S-procedure, and finally approximate the non-convex constraints by the first-order Taylor expansion. By the reformulations, we present iterative successive convex approximation algorithm for approaching the optimized solutions. Simulation results under various setups of system parameters are provided to examine the system performance and validate the effectiveness of the proposed schemes.