I. Introduction
With the rapid development of computer networks, the shared resources in networks are gradually increasing, cloud computing resources are widely used, and the kinds of heterogeneous resources in them are also increasing, and the information data are growing exponentially, and the huge data have the characteristics of diversity, multi-source, heterogeneity and replication in their data structure due to the different sources and distribution. How to efficiently process and schedule these heterogeneous data has become a research hotspot in today's big data processing, in which the research on cross-source scheduling algorithms for big data is a topic of wide interest to a wide range of related scholars.