Innovation diffusion network was built by ER stochastic graph model.  Explored were the influence of the behaviors from innovation adoption individuals and the network structure on the innovation micro-adoption and macro-diffusion.  The experiment results show that successful innovation diffusion will be determined by internal and external influences.  The larger tendency of the potential adoption individual getting the innovation evaluation information from his/her homogenous neighborhoods, the larger the adoption rate is.  With the increasing of the number of potential adoption individual, the adoption rate fluctuates periodically. The network structure and individual behaviors have nothing to do with the diffusion speed. These results are benefit to  innovators' promoting their innovations.