Research on the Fault-Tolerant Algorithm of Parallel Digital Terrain Analysis

In recent years,due to the increasing calculation demands for the massive spatial data analysis,the parallel computing based on high-performance computers has become an inevitable trend for Digital Terrain Analysis(DTA).At the same time,the reliability of the parallel system becomes a foremost key while the stability of the clusters with tens of thousands of processors is threatened constantly by a larger number of hardware and software failures.This paper takes parallel DTA technologies as research object and proposes a Neighboring-Algorithm Based Fault-Tolerant(N-ABFT) strategy so as to enhance the accuracy of failure detection in fault-tolerant software.By means of the check row/column,the N-ABFT algorithm can detect the transient and fail-stop failures after all the computing nodes finished the calculation.Finally,two algorithms based on different analytical windows are tested and the preliminary results are discussed.