Abstract: Due to the existence of cross-routing, there is a large gap between network topology detected from single point and the actual topology. The multi-point detection can make up the errors, but it's difficult to implement because of the high cost of erection and large amounts of data redundancy. The research is based on the massive data authorized by the CAIDA Skitter project and focuses on the analysis of ITM result from single vantage point in comparison with that from multiple vantage points. After calculating the difference between them on several eigenvalues, error distribution of data in each core detected from single point is analyzed based on the concept of coreness. Then, it is found that the intra-core distribution obeys the power-law and the inter-core distribution can be described by the fitted distribution formula. According to the procedure as above, the data results detected from single point can be revised to reduce errors and approach the integration data results measured from multi-points.