Based on the trade data between countries and provinces of the year 1997, 2002, 2007, 2012, from two different spatial scales（provincial scale and national scale）research perspective, this paper investigates 9 provinces of China and 18 countries along the Silk Road Economic Belt, constructs trade ties networks by social network analysis and draws its visual topological graph and analyzes the topological characteristics quantitatively. The results demonstrate that network density of the provincial scale increases at first and then decreases, for national scale which keeps steady growth. Sichuan and Chongqing locate in the centre of the network, while Shannxi, Ningxia and Qinghai are on the periphery of the network；Xinjiang and Gansu gradually move toward the periphery, while Yunnan and Guangxi gradually move toward the centre；European countries mostly situate in the centre, while fives countries of central Asia are on the periphery, and the upward trend of China’s position is the most obvious. In the network of provincial scale there are 5 cohesive subgroups and there are 9 in the network of national scale. The trade network relation of provincial scale is basically balance, while the trade network relation of national scale shows core-periphery structure. Regional cooperation should be enhanced with innovation-driven development and optimized network structure, and the radiation function of the network core should completely plays its role and the regional economy integration process should be accelerated to realize co-development of the economic belt regions.