๐”– Bobbio Scriptorium
โœฆ   LIBER   โœฆ

[ACM Press the 1st ACM Workshop - San Francisco, California (2010.06.15-2010.06.15)] Proceedings of the 1st ACM Workshop on Mobile Cloud Computing & Services Social Networks and Beyond - MCS '10 - User-profile-driven collaborative bandwidth sharing on mobile phones

โœ Scribed by Jung, Eric; Wang, Yichuan; Prilepov, Iuri; Maker, Frank; Liu, Xin; Akella, Venkatesh


Book ID
125483609
Publisher
ACM Press
Year
2010
Weight
429 KB
Category
Article
ISBN
145030155X

No coin nor oath required. For personal study only.

โœฆ Synopsis


The advent of smart phones, along with the paradigm shift towards cloud-based services, presents new challenges to the cellular backbone infrastructure. Cisco predicts that mobile data traffic will double every year through 2014, with a CAGR of 108% from 2009 to 2014, reaching 3.6 exabytes per month. We propose to exploit the potential of smart phones in proximity cooperatively, using their resources to reduce the demand on the cellular infrastructure, through a decision framework called RACE (Resource Aware Collaborative Execution). RACE enables the use of other mobile devices in the promixity as mobile data relays. RACE is a Markov Decision Process (MDP) optimization framework that takes user profiles and user preferences to determine the degree of collaboration. Both centralized and decentralized policies are developed and validated through simulation using real mobile usage traces. We implemented a simple prototype on a network of HTC G1 phones running the Android 1.5 operating system to demonstrate the viability of the system.


๐Ÿ“œ SIMILAR VOLUMES


[ACM Press the 1st ACM Workshop - San Fr
โœ Chun, Byung-Gon; Maniatis, Petros ๐Ÿ“‚ Article ๐Ÿ“… 2010 ๐Ÿ› ACM Press โš– 74 KB

Mobile cloud computing applications run diverse workloads under diverse device platforms, networks, and clouds. Traditionally these applications are statically partitioned between weak devices and clouds, thus may be significantly inefficient in heterogeneous environments and workloads. We introduce