Research On Physical Layer Security In Cooperative Cognitive Radio Networks

Cognitive radio is emerging as a promising paradigm to increase the efficiency of spectrum utilization and relieve the spectrum scarcity problem,which has gained a significant development in the past decade.As a combination of cognitive radio and cooperative relay technologies,cooperative cognitive radio has attracted extensive attentions in global academic community.In cooperative cognitive radio networks(CCRNs),the primary users lease the spectrum for the secondary users,in order to improve the network performance with the help of cooperative relay.Similar to other traditional wireless networks,due to the broadcast nature of wireless channel,the wireless transmission in CCRNs is sensitive to wiretap attack.The secure transmission has been traditionally achieved through key-based encryption at network layer.However,with the progress of computing capability of the eavesdropper,the encryption method at upper layer may be compromised and becomes very challenging.Besides,due to the characteristics of self-organization,low-complexity and resource-constrained in CCRNs,the encryption method with high complexity faces the challenges of key distribution and management.Different from the secret key encryption,physical layer security technology,which can fully exploit the characteristics of radio spectrum including time-variant and randomness,is capable of further enhancing the security performance from the perspective of information theory and proves to be an important complementarity for the existing security system.Given the advantages of low computational complexity,low power cost and eliminating the need of key distribution and management,physical layer security becomes a particularly suitable solution for security concern in CCRNs.Existing studies on physical layer security only focus on a simple network,resulting in a poor performance in such a heterogeneous and complex environment.Moreover,some simple and impractical assumptions have been supposed in these works without fully considering the cooperative mechanism among primary and secondary users,which makes the existing approaches hard to generalize and realize in CCRNs.The thesis focuses on physical layer security in CCRNs and tackles the aforementioned issues as follows:1.Considering the cooperative feature of primary/secondary users,we propose a cooperative jamming based physical layer security scheme to achieve reliable and secure communications.Specifically,1)we employ semi-definite programming and zero-forcing beam-forming technology to drive the information/jamming beamformer as well as the decoding vector to deteriorate the signal to interference plus noise ratio at the eavesdropper;2)by applying geometric programming,we design the power allocation vector to maximize the secrecy capacity under global power constraints;3)we formulate the jammer selection problem as a coalition formation game.Based on the merge-and-split rules and the Max-Pareto order,we propose a coalition formation algorithm,and obtain a stable and convergent coalition structure of secondary users.2.Previous researches assume the instantaneous channel state information of all users is known a priori.However,this assumption is too idealistic and cannot be applied in cooperative cognitive radio networks.In this paper,we attempt to employ the optimal stopping theory to joint select proper jammer and relay nodes with only statistical channel state information can be known.Specifically,1)with the purpose of achieving secure communications,we formulate the selection process as an optimal stopping theory problem and design the reward function according to the secrecy capacity and time efficiency;2)we derive the expected reward function of each observation step and propose the optimal stopping rule.Based on the optimal stopping rule,we propose an optimal selection algorithm and prove the convergence property of this algorithm;3)we prove the existence of the optimal sensing order and meanwhile propose a dynamic programming based approach to obtain the order for optimizing algorithm performance.3.An incentive mechanism needs to be implemented since previous researches assume that the jammer uses its own power resource to jam the eavesdropper while getting no return.Inspired by this,we propose an energy harvesting based jammer incentive scheme to fulfill the individual rationality of primary/secondary users.Besides,unlike most of researches in this area which assume that the transmitters can get the perfect channel state information of the eavesdropper,in this paper,we pay attention to the case when only imperfect information can be known.Specifically,1)based on a norm-bounded channel uncertainty model,we employ the robust optimization to design the transmit covariance matrices of jammer and relay nodes for maximizing secrecy capacity in the worst case;2)we build a harvest-and-jam mode for the jammer,wherein the jammer uses energy harvesting technology to balance the revenue and expenditure;3)by adopting Vickrey auction,we propose an jammer incentive algorithm and design a reasonable pricing strategy.We derive the dominant strategy equilibrium of this algorithm and prove that truthfulness and individual rationality can be guaranteed.The proposed algorithm also guarantees an improved secure communication quality of the primary user,and meanwhile,stimulates the cooperative behavior of the helper.