Suprathreshold stochastic resonance can be used to explain some biologic phenomena. In this paper, the suprathreshold stochastic resonance of a non-linear multilevel threshold neuronal network system is studied. The expression of the mutual information is derived, and the effects of the noise intensity and the system parameter on mutual information are discussed. The numerical computation results show that in the process of information transmission the selection of the threshold is very important according to the different effects of additive noise and multiplicative noise on system. Moreover, we also find that the more the number of devices, the more apparent the phenomenon of suprathreshold stochastic resonance is.