Cloud Computing

Introduction

Most of today's smartphone applications are geared towards an individual user and only use the resources of a single phone. There is an opportunity to harness the collective sensing, storage, and computational capabilities of multiple networked phones to create a distributed infrastructure that can support a wealth of new applications. (Lindemann 370) These computational resources and data are largely underutilized in today's mobile applications. Using these resources, applications could conveniently use the combined data and computational abilities of an entire network of smartphones to generate useful results for clients both outside and within the mobile network. This interface and the underlying hardware would create a mobile-cloud upon which compute jobs could be performed. We define mobile-cloud computing to be an extension of cloud computing in which the foundational hardware consists at least partially of mobile devices.

Some mobile applications already extract and aggregate information from multiple phones. Tweetie Atebits for the iPhone uses locations from other phones running the application to allow users to see recent Twitter posts by nearby users. Video and photo publishing applications such as YouTube and Flickr allow users to upload multimedia data to share online. The Ocarina application Smule for the iPhone allows users to listen to songs played by other users of the application, displaying the location of each user on a globe. Such smartphone applications are "push"-based and centralized, meaning that users push their information to a remote server where it is processed and shared.

It is possible to use a networked collection of smartphones in a more opportunistic way. Each smartphone has some amount of storage, some amount of compute power, some sensing abilities, some multimedia data, and some amount of energy. (Hull960) Each of these capabilities is currently only available to and utilized by the smartphone's owner. What if these capabilities were somehow offered to other users and applications? What if we 1 could harness a collection of smartphones to support large-scale distributed applications, using smartphones as the basis for a cloud computing infrastructure? Each smartphone would be equipped to perform individual, local computations on its local data in support of a larger, system-wide objective, and the outcomes of each smartphone's local actions would be aggregated to meet the needs of the overall application. Applications could use these resources abstractly, oblivious to the underlying implementation on a smartphone network.

Smartphone technology

Advances in mobile hardware and software have allowed users to perform tasks that were once only possible on personal computers and specialized devices like digital cameras and GPS personal navigation systems. (Sanjay 21)Using smartphones like the Apple iPhone, Android phones, and the BlackBerry, mobile users can now make full use of the Internet, capture and manage photos and videos, play music and movies, and play complex games. They have nearly ubiquitous access to the Internet via 3G services, WiFi, and peer-to-peer networking and can switch between networks automatically.

Sensors enable many interesting applications on smartphones. They provide information about the location, movement, and orientation of the phone and the environment's temperature and lighting. For example, ...