The existing spectrum sensing schemes based on free probability theory (FPT) have a high sensing performance especially in cases of low signal to noise ratio (SNR) and a small number of samples. However they all based on data fusion rule when performing cooperative spectrum sensing. The secondary users (SUs), in this case, have to transmit large amounts of sensing data to a fusion center (FC). It may cause the FC overloaded, which is considered as an extreme challenge to the practical application. In this paper, a new cooperative FPT-based spectrum sensing scheme is proposed, which utilizes the OR decision fusion rule to overcome the aforementioned shortcomings of such an algorithm. Simulation results over MIMO rayleigh fading channel demonstrate that, with decision fusion, the performance of our proposed scheme is much higher than that of the eigenvalue-based detector, even though the latter uses data fusion rule. Therefore, the proposed scheme is more effective and practical on the overall performance.