http://picasaweb.google.com/markus.endler/Favorites?feat=directlink Mobile Cloud Computing and Pervasive Social Networking: A glimpse of the (near) future Markus Endler LAC/PUC-Rio Agenda • Mobile Social Networking • Cloud computing and mobile world • Anatomy of future Mobile Cloud Apps • Examples of Mobile Cloud Apps • Current Barriers • Major Opportunities • Emerging Technologies • Pervasive Social Networks • Examples • Opinion & Outlook Mobile Social Network Growth M. Meeker forecast that next generation social networking mobile platforms will drive unprecedented change in communications and commerce. Indicators of the current growth of mobile social networking (source: Luke W. Data Monday: Mobile Social Networking, http://www.lukew.com/ff/entry.asp?932) • The second-most popular Web activity for mobile users to engage in on a daily basis is accessing a "social networking site or blog." In January 2008 1.8-million did it, in January 2009, 9.3-million (a monumental, 427% increase). • Four of the top ten domains accessed via mobile devices are social networking sites. • In September 2009: a one-year increase of 179% in subscribers accessing social networking sites from their mobile devices VS a 10% increase on the PC versions of the same sites. • Mixi’s (Japan’s leading social network) mobile monthly page views are three times desktop page views. Mobile Social Network Growth Indicators (cont.) • Irish mobile-phone users spend on average 45 minutes a day accessing social-networking sites via their mobiles • People that use Facebook on their mobile devices are almost 50% more active on Facebook than non-mobile users. • 25% of Internet users with mobile or wireless access, use Twitter or another micro-blogging service. • Almost 40 percent of Internet users with four or more Internet-connected devices use Twitter. The fewer the gadgets, the less a user is likely to connect via a micro-blog or status updating service. Other predictions: ABI Research's "Mobile Social Networking" (Sept. 2008) study: • More than 140 million mobile subscribers worldwide will use social networks on their phones by 2013, generating over $410 million in subscription revenues, according to P.S.: ABI said it was intentionally very conservative in its predictions. Novidades no Brasil • TIM e a Qualcomm anunciaram o lançamento da TIM App Store, primeira loja de aplicativos móveis multiplataforma – J2ME, Brew, Flash e Android, Windows Mobile, Palm, Symbian e LiMo) – milhares de desenvolvedores certificados pela Qualcomm poderão oferecer aplicativos para os clientes TIM. • Ericsson lança na Futurecom sua solução de Application Store, que facilita a comercialização e distribuição de aplicativos de maneira ágil e personalizada de acordo com o perfil do usuário • Motorola lança no Brasil o Droid/Milestone, primeiro smartphone da empresa rodando Android 2.0. Droid é 3G, tem teclado QWERTY em formato deslizante, tela sensível ao toque, câmera de 5 megapixels, Wi-Fi e GPS. • Droid vem com a ferramenta Motoblur de interação com redes sociais como Orkut, Facebook, Twitter e blogs. Ver YouTube: “Zumo: demo do MotoCliq e MotoBlur” Why cloud computing will become a disruptive force in the mobile world? 1. Mobile subscribers are a much larger user base 2. Smartphones are much richer devices than desktop machines, and they will become affordable for everyone in 2 years. 3. Wireless broadband is becoming a reality 4. Nowadays, mobile apps are distributed in a very operator-sectorized way. With Cloud computing you may be able to access more mobile applications - as long as you have access to the web. So far… • Google's App Engine hosts 45,000 conventional applications • Android Market has thousands of free mobile apps • Apple’s App Store has ≈ 95K iPhone applications, Estimates by end of 2010, ≈ 250K • Salesforce's mobile offering is limited to smart phones (Windows Mobile, Blackberry, and iPhone). • But as soon as technology makes app building for mobile devices as easy as for the web, we could have a “mobile revolution”. • New types of mobile apps are popping up all the time More Reasons... 5. Mobile devices have, and will always have much less power than desktop computers. Hence, the virtually unlimited computing & storage capacity of Clouds is even more a solution for mobiles. 6. There are more dominant mobile platforms than dominant desktop platforms. Mobile Cloud apps are easier cross-platform and allow to cover the whole mobile. 7. Clouds allow users to swap devices and still have access his data/information, as well as execute processing-intensive applications in the Cloud. 8. Wireless broadband networks evolve much faster than landlines, especially in third world countries. Broadband connectivity won’t be a problem. 9. Mobile development environments evolve faster than their desktop counterparts (see Android, iPhone OS, Symbian) Mobile Cloud Apps Google Apps Flickr, YouTube, Google Apps MySpace Anatomy of future Mobile Cloud Apps? Browsers to access CloudComp websites OR downloadable applications? Maybe both, but an additional sort: downloadable client-side apps • They will appear to be a typical mobile app – user might not even notice the difference, but… ... main difference will be in the back-end: • Mobile applications will store your data in the cloud as opposed to on the mobile device, and the applications will become more powerful as processing power is offloaded to the cloud. • mobile apps will communicate directly to the Cloud service itself, as opposed to needing the user to launch a web browser and navigate through the mobile web Mobile Cloud Apps 4 2 netbook 3 1 2 3 Smart phone 1. Downloadable small client-side mobile apps 2. MobClient-Cloud communication through an optimized, wireless- technology-aware protocol 3. Transparent transmission of user context data, transparent data syncronizations 4. Content search & adaptation, matchmaking, service selection, priv data storage, priv application processing, spontaneous notifications (push), etc. Examples of Mobile Cloud Apps • business-focused mobile productivity applications where: – collaboration, data sharing, multitasking, and scheduling are key factors – Example: Egnyte (http://www.egnyte.com) for document sharing, backup and collaboration • Private consumers: maybe navigation, SocNet sharing and tagging applications • specialty applications, e.g. for remote home surveillance and control, • Examples: – Schlange Link: remote keyless entry system which lets you control your home from a distance: let someone into your house, manage your lights, your thermostat, your camera system, etc – iPhone apps for remote control of your PC or DVR – ScreamingToaster uses Cloud Comp for Blogging App on the BlackBerry Soonr - Your Mobile Cloud iPhone/iPod touch app Allows to • access any uploaded file from your mobile phone • Share with anyone, using just their mobile phone number. • Search all files in your cloud. • Secure mobile/computer files backup. • Print to any shared printer Current Barriers 1. lack of speedy mobile Internet access 2. lack of internet access everywhere (e.g. at countryside) and intermittent connectivity (remember: all data will be in the cloud) 3. Not yet efficient mobile protocols (high data throughput with variable connectivity quality) 4. lack of standard interfaces and communication protocols for interoperability with different clouds Major Opportunities Cloud can offer huge storage and processing capacity for: • Social Networking interactions, collaboration and information sharing • Profile matchmaking • Context processing and customized content delivery • Additional services (yellow pages, maps, mobile- tuned search engines, etc.) based on other information correlations Emerging technologies • HTML5 (local caching, native editing facilities, optimized protocols) • Smartcard Web server (OMA) = SIM card that connects directly with the carrier to push applications to mobile phones. URL: www.openmobilealliance.org/technical/release_program/SCWS_v1_0.aspx • TokTok (http://www.mytoktok.com/) is a technology that allows access to web services like search, Gmail and Google Calendar by voice. Standardization Efforts BONDI Specification (http://bondi.omtp.org) by the Open Source Mobile Terminal Group – standardization of a small set of key interfaces for web services from mobile devices OneAPI (http://oneapi.aepona.com/) aims to provide a commonly supported, lightweight, Web friendly API • goal: allow developers to uniformly access parts of network providers' capabilities, such as location services. • This overcome the current problems of Mobile app development which requires tight integration with operator systems. Each operator has its API to access subscribers and other network capabilities (charging, messaging, location, user context). Emerging SocNet Protocols NoseRub (http://noserub.com/quick-facts/#whatis) is a open-source protocol for decentralized social networking. • allows to implement distributed social networking applications that are able to manage, share and synchronize the user’s ´profiles, friends-list, etc. of several social networks. • also supports aggregation of public streams (profile aggregation (mashup) service • general protocol for implementing decentralized sharing/synchronization of services using different programming languages. NoseRub uses OpenID, RSS and FOAF. DSNP (http://www.complang.org/dsnp/) is a protocol for distributed social networking. • Instead of keeping your profile on a SocNet service, the user stores it wherever the he/she wishes, and then directly exchanges contacts/profiles with peers executing DSNP. Pervasive Soc Networks • Just social networks accessed by mobile users...? • I believe not: – Enhanced SocNet, where user context information plays a fundamental role • Location, day-of-the week, current activity (inferred from movements and application usage) – New matchmaking and context inference processing (in the Cloud) – Consists also of spontaneous (situation-oriented) and temporary social networks Examples of PSN formed spontaneously • Bob watches a nice girl sole-dancing on her favourite music, and is curious about which song this might be. He can “tune” his smartphone to the music being played on the girl’s smartphone. • Several people (which do not know each other) are at a bus stop and would like to find people to share a cab to their neighborhood. • Mary is having problems operating a ticket selling machine in the main hall of a railway station. She would like to know if any other person in the hall has managed to purchase the ticket minutes ago. If there is someone, she would like to ask for help. • Drivers stuck in a traffic jam want to share information about possible aternative routes. Some Research Challenges Resource Efficiency and Scalability • How to optimize and scale the transmission of context information to the Cloud without compromising application’s performance and draining network resources? Privacy • How to design PSN services with flexible privacy control options and which makes the user trust that her data will not be used in a bad way? Context representation and Inference Mechanism • Which context ontologies and inference mechanisms are useful (and implementable) for relevant and appealing PSN applications? Parallel Algorithms • Create Parallel algorithms (maybe using Map-Reduce?) for context inference, and context-aware search and matchmaking in the Cloud. There are many other interesting aspects... Opinion & Outlook ABI Research: – “Within the next 5 years, mobile cloud computing will become the predominant application development strategy.” Clouds will become: – a marketplace for apps – a powerfull computing and storage backbone – a context and matchmaking engine – infra-structure for Pervasive Soc Networks Some References • http://www.readwriteweb.com/archives/why_cloud_computing_is_th e_future_of_mobile.php • J. Bardram, & T. Hansen: The AWARE architecture: supporting context- mediated social awareness in mobile cooperation, Proceedings of the 2004 ACM conference on Computer supported cooperative work • Mikko Salminen, Kari Kallinen, Kliment Yanev, Niklas Ravaja , Timo Saari, Mobile Application for Increasing Contextual and Emotional Work Group Awareness, LNCS vol. 4662, 2007 • Ankur Gupta, Achir Kalra, Daniel Boston, Cristian Borcea, MobiSoC: a middleware for mobile social computing applications, ACM/Kluwer Mobile Networks and Applications (MANET), 2005.