Research On Physical Layer Security Of Cognitive Relay Networks Based On SINR

The growing wireless service demand promote the development of the high-speed,reliable and secure wireless communication technologies,but the problems such as the scarcity,the low utilization of spectrum resources and the low reliability,the easy eavesdrop of communication links have become the bottleneck of development.Combined with the ideas of cognitive radio and cooperative communication,the cognitive relay networks not only can improve the diversity gain,but also can improve the utilization of spectrum resources,which have become the research hotspot.The expansion of communication scales brings more serious security threats,and the improvement on computer power is gradually enlarging the defect of current secure mechanisms based on the encryption theory,an absolute secure technology based on information theory——physical layer security has aroused the interest of researchers.In this thesis,the physical layer security in cognitive relay networks have been studied.Different from the usual safe indexes,such as security capacity,security outage probability and security error-rate,this paper use the SNR/signal to interference plus noise ratio(SINR)for analysis.Under the premise that the information can be transmitted safely and the quality of the primary user can not be affected,the Secondary-user relay clusters use beamforming to weight the received signal as well as transmitting a certain form of artificial noise.Our works are described as follows:1.Aiming at the cognitive relay network adopting the AF(amplify-and-forward)relay strategy,this paper considered two cases in the actual scenes: the low-load and high-load of relay clusters,the corresponding optimization targets are put forward respectively.Low-load means that the relay cluster can provide enough total transmit power,in which target the maximum received SINR;While in the case of high-load,the goal is to minimize the transmit power of the relay cluster under the condition of meeting the minimum demand of the received SINR.The problem of power distribution can be solved by transforming into an equivalent convex problem by means of the semi-definite relaxation(SDR)technique,and the performance can be further optimized by the binary search method under the high-load case.The deduced scheme is verified by simulation.Whether the relay cluster at the low-load or the high-load,the performance has been significantly improved.In addition,the complexity analysis showed that the scheme can be solved in polynomial time,and the correctness of SDR is proved by the analysis of the KKT condition.2.For the full-duplex cognitive relay networks based on the "green communication" background,this paper propose the secure transmission method to improve the energy-efficiency.Based on the consideration of the self-interference cancellation rate and the forward fairness,the goal is to obtain more total system energy-efficiency.Two kinds of relay forwarding strategies need to be considered: AF and DF(decode-and-forward),respectively,the relay clusters use collaborative beamforming and artificial noise to forward.The mentioned above problem is solved by the hill-climbing algorithm combined with SDR.Finally,the simulation results and analysis showed that the total energy-efficiency is effectively improved.Compared with the DF,the performance of the AF relay strategy is obviously higher.