Design of Multi-source Big Data Cross Source Scheduling System Based on Optimal K-means Clustering Algorithm

Traditional big data scheduling methods cannot effectively integrate massive data in the network, which easily leads to "rigid" operation of the system. In response to the above issues, this project intends to study a multi-source collaborative scheduling method for multi-source heterogeneous big data. K-average connections store data in binary format, and mine the index nodes of the data to achieve organizational indexing of the data. On this basis, the scheduling framework was used to obtain cross source parameters, and the location of the cross source scheduling center was determined through direct tracking, achieving the smooth application of the design of a multi-source big data cross source scheduling system based on the optimal K-means clustering algorithm. The scheduling time of the Design System 3 is relatively long, but its scheduling time does not change much as the amount of data increases. Through simulation and comparison, the experimental results showed that compared with traditional big data scheduling methods, the new scheduling method can effectively suppress the phenomenon of system rigid operation.