Detecting the presence or absence of primary user is the key task of cognitive radio networks.
However, relying on single detector reduces the probability of detection and increases the probability of
missed detection. Combining two conventional spectrum sensing techniques by integrating their individual
features improves the probability of detection especially under noise uncertainty. This paper introduces a
modified two-stage detection technique that depends on the energy detection as a first stage due to its
ease and speed of detection, and the proposed Modified Combinational Maximum-Minimum Eigenvalue
based detection as a second stage under noise uncertainty and comperes it with the case of using
Maximum-Minimum Eigenvalue and Combinational Maximum-Minimum Eigenvalue as a second stage.