Research On Channel Security And Power Optimization Based On Artificial Noise In The Physical Layer

Since Claude Shannon created the information theory, a lot of research of secure communication has been increased. Using the secret key to process transmission information once became the mainstream way of ensuring the security communication. But with the arrival of the era of big data. Malicious eavesdropper also can easily have a strong cluster computing ability. The ability of this traditional secret key encryption method began to fall short of our wish. So the researchers focused on the communication security technology of physical layer. Different from the Secret key encryption, the technology of secure communication which in physical layer, not only proceeds from the angle of information theory to deduce and calculate the positive secure capacity for ensuring the perfect security of the communication, but also can be considered from the angle of signal processing, to makes a preprocessing for the transmitted signal, such as superpose the artificial noise to increase the security capacity of channel. Up to this day, channel security problem in the physical layer and the method for optimal power allocation of channel which derived from this subject have already become a hot research topic in the field of communication security. Faced with an increasingly prominent communications security issues, how to find out the optimal power allocation scheme of the channel model correctly and effectively has become a very meaningful and challenging research work.This paper studies the problem of multi-class channel security and optimal power allocation in physical lay which based on the artificial noise(AN) method, propose the corresponding solutions for improving the security performance of communication channel in physical layer and the algorithm of channel optimal power allocation, through a comprehensive and thorough analysis and research, this paper has made the following research results:1.Consider the security communication problem for the broadcast channel which with a single eavesdropper(BCE), according to the definitions and concepts of achievable secure rate and equivocation rate, we propose and prove the existence of an outer bound for achievable secure rate region on channel model, further proved that this outer bound tight to the outer bound which obtained from using max-flow min-cut theorem of graph theory. Show the significance of the outer bound which proposed in this paper; Consider secure communication problems of MISO channel at the presence of an active attacker. Simulate attack situation from the attacker’s point of view and establishes models in wireless environment. Using the average mutual information as the mathematical description amount to prove the effect of active attack in this model, define the average mutual information loss rate which is the secure measurement index, and use it to measure the channel communication ability under active attack. Using the method of space connection which based on pre-coding in theory to solve the security communication problems for MISO wireless channel which under the active attack situations that proposed in this paper; Consider the security communication problem of SISO channel when the transmitter’s(Alice) antenna number is limited which cannot make extra redundancy. Propose a new 2-stage bilateral AN transmission method that liberate the possible constraint on Alice’s antenna number which due to the actual environment.2.Consider the optimal power allocation problem for the MIMO channel which generating AN to prevent eavesdropping. In order to achieve the maximum secrecy capacity, propose an efficient iterative algorithm which similar like one-dimensional search to find the optimal power allocation ratio between AN with information signal. After using random matrix theory to obtain a closed-form expression for the lower bound of the channel secrecy capacity, further prove the water-filling is the optimal power distribution solution under the scenario which proposed in this paper.3.Consider the problem that traditional AN method would be limited under the condition of channel interaction. Define interception degree ?, disrupt degree ? and channel risk degree ?, to characterize the index of channel secure measurement. Derive the relation expression between expectation of ? and the channel interaction coefficient, and use this expression to analyze the channel interaction degree effect on the communication security of MISOSE channel which based on AN method. Base on Ⅰ, Ⅱ ill-pose theory, select the most representative channel, like MIMOME and SISO to make the security analysis for ?, and the obtained conclusion has guiding significance for determining the existence of positive security capacity when make the channel design or modeling.The main innovations of this thesis are described as follows:1.This paper gives and proves an OB of BCE model in the research of achievable secure rate region for BCE model. Through the comparison with the cutting set constraint given by max-flow min-cut which is the threshold condition can determine the significance of channel OB. It is proved that the OB which given by this paper tight in the cutting set constraint. Through the comparison with the existing three types of representative OB of BC model. It is proved that the OB which proposed by this paper tight or at least equal to the other OB of same type model.2.We break the routine and from the attacker’s point of view to establish the model for active attack scenario in this paper. We use the average mutual information as the mathematical description amount firstly and use the ratio of the average mutual information to define the loss rate of average mutual information, it can show the effect of active attack which given by this paper in specific model. We propose the method of vector NS connection which based on pre-coding in theory to cope with the security communication problems for MISO channel which under the active attack situations that proposed in this paper. Simulation experiment and numerical calculation show that, the solution proposed in this paper can effectively defense the active attack which set by this paper.3.This paper proposes a new method which called 2-stage bilateral AN transmit in the research of the security communication problem for SISO channel when the Alice’s antenna number is limited. It not only seriously disrupts the information receiving on Eve, but also ensures the private information receiving on Bob by setting one transmit and one receive antenna on Bob, Shown by theoretical derivation, to compare with the method of multiple helper nodes, the method proposed in this paper have the advantage in terms of power consumption, which proves that the method proposed in this paper can effectively liberate Alice’s antenna number restriction which due to the external environment, and it is more universal in the practical situation.4.In the research of optimal power allocation problem for the MIMO channel which using AN method, we obtain a closed-form expression for the lower bound of the channel secrecy capacity based on the random matrix theory and prove WF is the optimal power distribution solution under the scenario which proposed in this paper. In order to find the optimal power allocation ratio between AN with information signal to achieve the maximum channel secrecy capacity, this paper proposes an efficient iterative algorithm which similar like one-dimensional search. Theory derivation and simulation results show that, the proposed algorithm can quickly converge and multiple antennas AN scenario can promote the secrecy capacity more than other AN case to a certain extent.5.We propose a method of uniform and integral processing multi antennas on Bob which called MIX-AN in the research of antenna numbers constraint for MIMO channel which based on AN and the AN method is limited under the condition of channel interaction. Through theoretical derivation, it is proved that this method can not only liberate the number restriction for Alice’s antenna which due to multi antenna on Bob, but also increase the redundancy dimensions of channel NS, so as to effectively guarantee the communication security for MIMO channel. Through defining three indexes which can characterize the channel secure measurement in this paper, we analyze the channel interaction degree effect on the communication security of MISOSE channel which based on AN method. Base on Ⅰ, Ⅱ ill-pose theory, we select the most representative channel to make the security analysis for channel risk degree. The theoretical deduction shows that it needs to ensure the channel risk index greater than 1 in the design of channel modeling, so as to obtain positive secrecy capacity.