To study the strategy for wideband spectrum sensing in cognitive radio system

the relation between the accuracy of channel state estimation and entropy was analyzed from the perspective of information entropy.And then a least entropy based randomized strategy（LERS） for wideband spectrum sensing with cooperation was proposed.By fully utilizing priori information and actually sensed results of channel states

the proposed strategy could select sensing channels to minimize total entropy

and the spectrum sensing interval of each channel was automatically regulated.Theoretic analyses and simulation results show that the least entropy based strategy effectively improves the accuracy of channel state estimation and the efficiency of spectrum sensing.