Deploying hybrid methodology in enhancing large file transfer

File transfer is important for interconnected network systems. File synchronization is a process that uses file transfer method to update files between distanced locations. For an organization with regional distanced companies, file synchronization plays an important role in maintaining data concurrency between the companies. Due to the demand of faster file transfer methods, many related research have been done to improve the performance of file transfer. One of the approaches is to identify the network traffic condition and perform file transfer when the network traffic is relatively low. Time series forecasting exhibits the possibility for predicting the network traffic condition in the future events accurately. However, there are a few issues that needed to be resolved in order to utilize forecasting in file transfer. The objective of this paper is to evaluate the total improvement of a proposed file transfer method built on a forecasting module. This paper consists of three parts. Part one is time series model determination. This consists of time series model parameter estimation method by using Box-Jenkins methodology. Part two is to deploy and evaluate the performance by implementing file chunking with forecasting method before initializing file transfer. This is to determine the performance of file chunking algorithm in assisting time series forecasting during file transfer. Part three is similar process in part two with slight changes in the file transfer order. In the end of the experiment, we will conclude the usability of the proposed method based on the experiment conducted for this paper.