In　stochastic　networks,　nodes　usually　function　dependently　and　interact　with　other　nodes　through　connectivity　links　or　dependency　links.　In　this　paper,　the　model　for　stochastic　networks　considering　sub-networks　with　connectivity　and　dependency　links　of　Erdos-Renyi　(ER)　topology　is　proposed,　which　is　defined　as　networks　with　arbitrary　pair　of　nodes　randomly　connected/depended　by　a　constant　probability.　The　reliability　evaluation　framework　for　the　proposed　networks　is　developed,　where　both　of　the　extended　multi-valued　decision　diagram　(MDD)　method　and　Monte　Carlo　simulation　(MCS)　are　involved.　The　MDD　method　is　proposed　to　assess　the　reliability　of　deterministic　stochastic　networks　with　ER　connectivity　and　dependency,　where　arbitrary　time　to　failure　distributions　of　nodes　are　allowed.　Based　on　the　reliability　evaluation　for　a　stochastic　network　with　a　deterministic　structure,　the　MCS　is　employed　to　achieve　the　reliability　analysis　of　corresponding　stochastic　networks.　Numerical　examples　are　presented　to　demonstrate　the　proposed　stochastic　network　model　and　reliability　evaluation　framework,　where　the　probability　distributions　for　the　reliability　of　stochastic　networks　are　provided.