QSAR/QSPR studies use physicochemical parameters and topological indices (e.g. Randic, ABC, GA) to predict the bioactivity of chemicals. A topological index is created by converting a chemical structure into a numerical value. Topological indices connect with physicochemical parameters such as boiling point, stability, and strain energy of chemical compounds. Graph theory has found extensive use in this field of study. Hayat and Imran (2014) conducted a recent study on the topological features of networks [13]. In this research, we expand the concept to interconnection networks and derive analytical closed results of the general Randic' index RaðGÞ for different values of ''a'' in butterfly and Benes networks. We also calculate first Zagreb, ABC, and GA indices for these important types of networks. We create two new mesh derived networks by applying graph operations to m-n mesh networks. We next analyze topological indices for these networks.