Based on effectively using spectrum resources, Cognitive Radio (CR) is deemed to be a promising way that can alleviate the spectrum scarcity problem. Spectrum sensing, as a key component of CR technologies, has a significant effect on the performance of the CR system. Compared with the single user spectrum sensing, cooperative spectrum sensing improves the detection reliability by exploiting the diversity of CR users. After analyzing the existing cooperative spectrum sensing schemes, the energy saving and security issues in cooperative spectrum sensing were studied in this thesis.Cooperative spectrum sensing needs some information exchange in the procedure of cooperating, so that additional communication overhead is generated and more energy consumption is introduced to engaged CR users. Considering the communication overhead induced by the cooperative operation, we propose a novel cooperative spectrum sensing scheme with a light communication overhead based on each CR user's local confidence degree. The CR user does not send its sensing result if its local confidence degree is smaller than the confidence threshold. Therefore, the proposed cooperative spectrum sensing scheme can decrease the communication overhead and save energy in CR users.Because of the characteristics of CR, cooperative spectrum sensing is vulnerable to a variety of attacks, which will be harmful to the security of CR networks. Considering the security issues in the cooperative spectrum sensing, we analyze the impact of spatial correlation on cooperative spectrum sensing, and propose another cooperative spectrum sensing scheme based on the spatial correlation between CR users. This scheme prefers to choose the CR users which are far away from each other in space to perform cooperation, and avoid choosing the geographical closed CR users. In this way, the probability of multiple CR users being influenced by the primary user emulation (PUE) attacker at the same time is decreased, so that the impact of PUE attackers can be mitigated in cooperative spectrum sensing. |