Industries are key to economy, fundamentally for a nation's basis. Advancing industrial structures is a vital approach to modernized industrial system, industrial core competitiveness and entering middle-upper end of global value chains. This paper uses industrial commonality index to measure China's 31 provinces' 2012 to 2021 similarity matrix of industrial structures, and applies social network analysis (SNA) and secondary assignment procedure (QAP) to study the spatial correlation characteristics and functioning mechanism. Industrial commonality index can effectively show the asymmetry between inter-provincial industrial structural similarity and regional relation. Spatial correlation network of inter-provincial industrial structures can be divided into 4 domains, the first domain includes Beijing, Tianjin, Jiangsu, Guangdong, Zhejiang, Shanghai and Chongqing, at the top of industrial structural network which plays a leading role in optimizing industrial structures. The second domain includes Shanxi, Liaoning, Fujian and Shandong, playing a bridging and mediating role amid industrial migration, interactive with the first domain and outflowing to the third domain. The third domain includes Inner Mongolia, Hebei, Jiangxi, Sichuan, Henan, Hubei, Hunan, Anhui, Guangxi, Shaanxi and Jilin, both receiving the second domain's outflowing and outflowing to the forth domain. The forth domain includes the rest, which needs to receive industries from more developed areas in industrial structural adjustment. QAP suggests that labor inputs, human capital, capital types, and end consumption and geographic neighboring may be partially interpretated as spatially correlated, and path to inter-provincial industrial migration and receiving may be optimized.