The Internet of Things (IoT) is a new paradigm for connecting various heterogeneous networks. Cognitive radio (CR) adopts cooperative spectrum sensing (CSS) to realize the secondary utilization of idle spectrum by unauthorized IoT devices, allowing IoT objects can effectively use spectrum resources. However, the abnormal IoT devices in the cognitive Internet of Things will disrupt the CSS process. For this attack, we propose a spectrum sensing strategy based on weighted combining of the hidden Markov model. The method uses the hidden Markov model to detect the probability of malicious attacks at each node and reports to the Fusion Center (FC), which evaluates the submitted observations and assigns reasonable weight to improve the accuracy of the sensing results. Simulation results show that the algorithm proposed has a higher detection probability and a lower false alarm probability than other algorithms, which can effectively resist spectrum sensing data falsification (SSDF) attacks in cognitive Internet of Things and improve the performance of IoT devices.

物联网 (IoT) 是连接各种异构网络的新范式。认知无线电（CR）采用协同频谱感知（CSS）实现非授权物联网设备对空闲频谱的二次利用，让物联网对象可以有效利用频谱资源。然而，认知物联网中的异常物联网设备会扰乱 CSS 流程。针对这种攻击，我们提出了一种基于隐马尔可夫模型加权组合的频谱感知策略。该方法使用隐马尔可夫模型检测每个节点的恶意攻击概率，并报告给融合中心（FC），融合中心对提交的观察结果进行评估并分配合理的权重，以提高感知结果的准确性。