The networks of interdependencies in large-scale software systems are complex, understand and measurement become a key problem for developer. we propose that OO design structure can be quantitative analyzed and better understood via core-shell decomposition based on complex networks. The k-core analysis allows to characterize networks beyond the degree distribution and uncover structural hierarchies , centrality and evolution due to the specific architecture of the system. we analyze the core-shell structure of some popular open source software and discuss the differences and similarities, get some noticeable properties, the result show the k-core analysis provides an interesting characterization helping to understand the underlying structure and improve development.