With the evolution of communication systems toward the sixth‐generation technology (6G), intelligent cognitive communication has gained considerable attention. As an important part of intelligent cognitive communication, cognitive radio (CR) offers promising prospects for efficient spectrum utilization. However, with the introduction of cognitive capabilities, CR networks (CRNs) face not only common security threats in wireless systems, but also unique security threats, including primary user emulation (PUE) attacks, endangering communication reliability and confidentiality. In order to enhance the defense ability of CRNs against PUE attacks, this paper proposes an integrated sensing and communication (ISAC)‐assisted approach. Leveraging ISAC technology, our scheme enhances location detection precision. We introduce a high‐resolution perception signal parameter estimation method and a position‐based identity authentication scheme. Furthermore, deep reinforcement learning is used to dynamically optimize the authentication threshold to ensure the stability of authentication in dynamic scenarios. Simulation results show that the proposed scheme is effective in resisting PUE attacks and improves the security and reliability of CRNs.

Aiming to address the security and timeliness challenges in reconfigurable intelligent surface (RIS)‐assisted integrated sensing and communication (ISAC) system with finite blocklength (FBL), this paper jointly investigates the communication security, sensing security, and information freshness performance of the system in the presence of communicating eavesdropper and sensing eavesdropper. Specifically, based on statistical channel state information (CSI), approximate closed‐form expressions for secrecy throughput, average age of information (AoI), and channel parameter estimation errors are derived and analyzed to characterize the performance of communication security, information freshness, and sensing security. The asymptotic analyses between secrecy throughput and blocklength, number of antennas, and number of RIS reflecting elements are established. Furthermore, an optimization problem for maximizing sum secrecy throughput is established under the timeliness, sensing security, transmit power, and RIS unit modulus constraints. To handle the intractable stochastic nonconvex problem, a joint alternating optimization method based on noncooperative game and stochastic successive convex approximation (NCG‐SSCA) is proposed by jointly designing RIS phase shift, transmit beamforming vector, sensing signal covariance, and blocklength. Simulation results validate our theoretical derivations and conclusions in the performance analysis. It is also shown that compared with SSCA and stochastic gradient descent (SGD) methods, the NCG‐SSCA method proposed in this paper achieves an increase in sum secrecy throughput by 10.4% and 16.3% with faster convergence speed.

This paper addressed the trade‐off between timeliness and reliability in joint communication and over‐the‐air computation offloading (JCACO) system under short‐packet communications (SPCs). The inevitable decoding errors introduced by SPC lead to errors in the data aggregation process of over‐the‐air computation (AirComp). Due to limited resources, pursuing high reliability may prevent the JCACO system from meeting delay requirements, resulting in a trade‐off between reliability and timeliness. To address this issue, this paper investigates the timeliness and reliability of the JCACO system. Specifically, the moment generating function method is used to derive the delay outage probability (DOP) of the JCACO system, and the outage probability of AirComp is calculated based on the errors that occur during its data aggregation process. The paper established an asymptotic relationship between blocklength, DOP, and AirComp outage probability (AOP). To balance timeliness and reliability, an AOP minimization problem is formulated under constraints of delay, queue stability, and limited resources based on computation offloading strategies and beamformer design. To overcome the issues of slow convergence and susceptibility to local optima in traditional algorithms, this paper proposed a stochastic successive mean‐field game (SS‐MFG) algorithm. This algorithm utilizes stochastic continuous convex approximation methods to leverage Nash equilibria among different users, achieving faster convergence to the global optimal solution. Numerical results indicate that SS‐MFG reduces AOP by up to 60%, offering up to a 20% improvement in optimization performance compared to other algorithms while also converging faster.