Efficient and Transparent Wi-Fi Offloading for HTTP(S) POSTs

IEEE Trans. Mob. Comput.(2016)

引用 4|浏览16
暂无评分
摘要
With the emergence of online platforms for (social) sharing, collaboration and backing up, mobile users generate ever-increasing amounts of digital data, such as documents, photos and videos, which they upload while on the go. Cellular Internet connectivity (e.g., 3G/4G) enables mobile users to upload their data but drains the battery of their devices and overloads mobile service providers. Wi-Fi data offloading overcomes the aforementioned issues for delay-tolerant data. However, it comes at the cost of constrained mobility for users, as they are required to stay within a given area while the data is uploaded. The up-link of the broadband connection of the access point often constitutes a bottleneck and incurs waiting times of up to tens of minutes. In this paper, we advocate the exploitation of the storage capabilities of common devices located on the Wi-Fi access point’s LAN, typically residential gateways, NAS units or set-top boxes, to decrease the waiting time. We propose HOOP, a system for offloading upload tasks onto such devices. HOOP operates seamlessly on HTTP(S) POST, which makes it highly generic and widely applicable; it also requires limited changes on the gateways and on the web servers and none to existing protocols or browsers. HOOP is secure and, in a typical setting, reduces the waiting time by up to a factor of 46. We analyze the security of HOOP and evaluate its performance by correlating mobility traces of users with the position of the Wi-Fi access points of a leading community network (i.e., FON) that relies on major national ISPs. We show that, in practice, HOOP drastically decreases the delay between the time the photo is taken and the time it is uploaded, compared to regular Wi-Fi data offloading. We also demonstrate the practicality of HOOP by implementing it on a wireless router.
更多
查看译文
关键词
IEEE 802.11 Standards,Logic gates,Web services,Mobile handsets,Mobile communication,Browsers,HTML
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要