科研之友
微信
新浪微博
Facebook
分享链接
摘要
Collaborative spectrum sensing (CSS) was envisioned to improve the reliability of spectrum sensing in centralized cognitive radio networks (CRNs). However, secondary users (SUs)' changeable environment and ease of compromise make CSS vulnerable to security threats, which further mislead the global decision making and degrade the overall performance. A popular attack in CSS is the called spectrum sensing data falsification (SSDF) attack. In the SSDF attack, malicious cognitive users (MUs) send false sensing results to the fusion center, which significantly degrades detection accuracy. In this paper, a comprehensive reputation-based security mechanism against dynamic SSDF attack for CRNs is proposed. In the mechanism, the reliability of SUs in collaborative sensing is measured with comprehensive reputation values in accordance with the SUs' current and historical sensing behaviors. Meanwhile a punishment strategy is presented to revise the reputation, in which a reward factor and a penalty factor are introduced to encourage SUs to engage in positive and honest sensing activities. The whole mechanism focuses on ensuring the correctness of the global decision continuously. Specifically, the proposed security scheme can effectively alleviate the effect of users' malicious behaviors on network decision making, which contributes greatly to enhancing the fairness and robustness of CRNs. Considering that the attack strategy adopted by MUs has been gradually transforming from simplicity, fixedness and singleness into complexity, dynamic and crypticity, we introduce two dynamic behavior patterns (true to false and then to true (TFT) and false to true and then to false (FTF)) to further validate the effectiveness of our proposed defense mechanism. Abundant simulation results verify the rationality and validity of our proposed mechanism.
- 出版日期2016-12
- 单位哈尔滨工程大学
