INTRODUCTION e trend of computing technology is that everything is merged into the Internet [1-3] and “Big Data” are integrated to comprise complete information for collective intelligence. Sensors and human beings become two major contributors for Big Data. With the advance of low cost data capturing devices and storages, most of data are collected via sensors [4]. Human-generated data on the web has become even more popular in the recent years. e top ve biggest data in 2014 come from archives, government websites, social media, the media, and business apps [5]; all these data record human behaviors and social activities. Data is generated in a way exceeding the human limits to use them [6]; 90% of the data in

CONTENTS Introduction 139 e Protocols of Internet Data Transfer 141

e Mechanism of TCP 141 File Transfer Protocol (FTP) 142 e Challenges of Big Data Transfer on the Internet 142

Ecient Big Data Transfer 143 Data Segmentation for Multipathing and Parallel TCP Streams 144

Multipathing 144 Parallel TCP Streams 144

Multihop Path Splitting 145 Hardware-Supported Dynamic Bandwidth Control: Aspera FASP 146

e Mechanism of Aspera FASP 147 e Performance of Aspera FASP 147

Optimizing Bandwidth: BitTorrent 148 All-to-One Gather of Big Data 149 One-to-All Broadcast of Big Data 150 Conclusion 153 References 153

the world was created in the past few years alone [1] and the amount of data is doubling every year in every science domain [7]. e big-sized data exceeds the ability of general information technologies to capture, store, manage, and process within a tolerable elapsed time [8]. erefore, Big Data is dened as “high-volume, high-velocity, and/or high-variety information assets that require new forms of processing to enable enhanced decision making, insight discovery and process optimisation” [9].