Abstract: Compressed Sensing (CS) theory provides a new solution for lowering acquisition cost and synchronous demodulation processing pressure of DS TT&C signal, furthermore sparsity is an important prerequisite for CS application, but the research on sparsity of the signal is seldom reported. In this paper, the sparsity of DS TT&C signal is in depth analyzed by building the basic dictionary, and a dual-stage dictionary learning method is proposed, moreover the basic learned dictionary and delay-Doppler dictionary are built based on the dual-stage dictionary learning method and detailed analysis of the signal model. Lastly, the performances of the basic dictionaries are verified by simulation experiments. The results show that DS TT&C signals receive a strong sparsity both in the built basic dictionaries, which provides a foundation for signal processing on the basis of CS.