Store-and-Forward Data Transfer using Optimized Intermediate Node

In this paper, we propose a new data transfer method for science big data. Current science requires much bigger data than ever before, but the time to receive the data still takes long due to the well-known last mile problem. We propose a new method which shows better performance for packet losses caused by the last mile problem, in transferring science big data. The proposed method uses intermediate storage server to enhance the transfer throughput, and we call it DTN. It is optimized for high performance transfer in storing and sending data, and separates the last mile from end-to-end path. We verified the improved performance by measuring the end-to-end throughput.