Abstract: Cooperative sensing can improve sensing accuracy, but may expose the system to the dangers of malicious users that may be present, and these malicious users may launch denial-of-service attacks against the network. In this paper, we propose a new detection algorithm known as the malicious user detection algorithm (MUDA). First, we conduct various sensing tests with secondary users using the sensing server, and then estimate whether or not this node is malicious based on the received signal strength of the node being tested. Then, we adopt an active transmissions strategy to improve detection accuracy and reduce the overheads. Finally, a comparison of our performance results with those of existing algorithms, demonstrates that MUDA not only achieves higher accuracy but also reduces 65% of the throughput loss caused by the sensing operation.