In view of the emergence of mega data signals in wireless communication networks in recent years

this paper properly compresses these signals while performing signal sampling and realizes their reconstruction with small amount sampling value or observed value by using an appropriate reconstruction algorithm.On the basis of the original Orthogonal Matching Pursuit（OMP）-based compressed sensing reconstruction algorithm

the Alternating Step-size（AS）is introduced and an improved algorithm

i.e.GP-OMP algorithm proposed.Experimental results show that the spectrum information this algorithm obtained by using the previous sensing moment can not only effectively reduce the amount of computations by the reconstruction algorithm

but also the reconstruction time

with almostthe same results as those by using the original method.