Fast data transfer technique for large data in grid environment

• Main Author: Chong, Mien May
• Format: Thesis
• Language: English
• Published: 2015
• Online Access: http://psasir.upm.edu.my/id/eprint/57090/1/FSKTM%202015%203RR.pdf; http://psasir.upm.edu.my/id/eprint/57090/

Nowadays, multimedia files such as videos and images are being improved in terms of their quality, where the dimension of the video or image is improving from two dimensional (2D) to become three or four dimensional (3D or 4D). These quality improvements have also increased the size of the multimedia data. Thus, the large size of video data has used a lot of time to transmit through the network. To solve the problem, in our research, we proposed a new video splitting technique, which is the "Exponential-and-Uniform-based (ExpoNUni) Splitting Technique" to improve the transmission time and maintain the quality of the data. Our "ExpoNUni" Splitting Technique has performed a better result compared to the existing techniques such as Fibonacci-based Splitting with V1min Technique and Uniform-based Splitting Technique. For the Initial Delay, overall, our splitting technique by average has 26.23 % lesser initial delay time for comparing with the existing Fibonacci-based Splitting with V1min Technique. It shows an 11.62% average of initial delay time lesser comparing to Uniform-based Splitting Technique.For download time, the overall result that compared with the existing Fibonacci-based Splitting with V1min Technique, our splitting technique by average 1.43% lesser time for downloading a complete video file to the user‘s PC from the grid storage, while the Uniform-based Splitting Technique, it shows 1.67% decrement of average downloading time. To prove the Exponential-and-Uniform-based splitter component is able to reduce the transmission time in a Grid-based Video Storage Framework, we also compare the uploading time and downloading time between the framework without the splitter component and the framework with our splitter component. The results show that the framework with our splitter component proofs it takes lesser data transmission time,where overall with an average of 9.86% lesser time for the uploading time and overall with an average of 12.24% lesser time for the downloading time. Besides, the MATLAB tool proves that our splitting technique did not affect the quality of the retrieved data.