Malicious users who deliberately damage the networks or greedy users who enhance their communication benefits will violate interference constraint (VIC), thereby disrupting access priority and preventing normal communication. As sensing nodes become lighter and smaller, the energy storage of nodes that sense the VIC behavior is finite, which brings challenges to spectrum sensing under VIC behavior. To tackle the problem, in this letter, we build the corresponding mathematical model based on the spatial-temporal heterogeneous scenario and design two corresponding detection methods according to locations of SUs. Focusing on multi-sensing nodes cooperative detection scheme, to further improve the detection performance, we design two sensing nodes selection mechanisms based on energy consumption (EC) and energy consumption-signal to interference plus noise ratio (EC-SINR) in the data fusion process respectively. Finally, simulation results are provided to verify that the multi-sensing nodes cooperative detection scheme can improve the detection performance, the sensing nodes selection mechanism based on EC-SINR can reflect the value of the sensing nodes’ data better and its detection probability is about 0.1 ~ 0.2 higher than the EC.