Cooperative Spectrum Sensing (CSS) is an effective approach to improve the detection performance of vacant frequency bands in Cognitive Radio (CR) networks. The CSS process imposes some security threats to the CR networks. One of these common threats is Primary User Emulation Attack (PUEA). In PUEA, some malicious users try to mimic primary signal characteristics and deceive CR users to prevent them from accessing the vacant frequency bands. The present study introduces a new CSS scheme, in the presence of a malicious PUEA, called Attack-aware CSS (ACSS). The idea is based on the estimation of attack parameters including probabilities of the fake PUEA signals’ presence in both occupied and unoccupied frequency bands. The obtained parameters are innovatively applied in Neyman-Pearson (N-P) or Log-Likelihood Ratio Test (LLRT) to improve collaborative sensing performance. Simulation results verify the performance improvement of the proposed method against PUEA compared with conventional method.