The phenomenal growth of mobile and embedded devices coupled with their ever-increasing computational and communications capacity presents an exciting new opportunity for real-time, distributed intelligent data analysis in ubiquitous environments. In these contexts centralized approaches have limitations due to communication constraints, power consumption (e.g. in sensor networks), and privacy concerns. Distributed online algorithms are highly needed to address the above concerns. The focus of this talk is on distributed stream mining algorithms that are highly scalable, computationally efficient and resource-aware. These features enable the continued operation of data stream mining algorithms in highly dynamic mobile and ubiquitous environments.
João Gama, Associate Professor, University of Porto and Researcher at LIAAD/INESC TEC U. of Porto working at the Machine Learning group. His main research topic is in machine learning for evolving data and models, including distributed learning. He is member of the Editorial Board of Machine Learning Journal, Data Mining and Knowledge Discovery, Intelligent Data Analysis and New Generation Computing, He served as Program Chair of ECML 2005, DS09, ADMA09, ECMLPKDD 2015 and a series of Workshops data stream mining with ACM SIGKDD and SIGAPP. He is author of a recent book on Knowledge Discovery from Data Streams.
New vehicle applications have recently emerged, ranging from navigation safety to location aware content distribution and intelligent transport. These applications require efficient communications support. For example in a world of autonomous vehicles, V2V communications will be essential for stable cruise control and platooning. Vehicles will provide services to fellow drivers (e.g., congestion or spectrum crowd sourcing) as well as to external customers. An example of the latter is urban sensing. Vehicles have become rich sensor platforms. They can monitor the environment and classify events, e.g., license plates, chemical readings, etc. They generate metadata that can be used to resolve Insurance claims after minor road accidents or even assist Law Enforcement Agents in the forensic investigation of crimes. The Service notion suggests that the VANET can be viewed as a Mobile Computing Cloud (MCC) where vehicles interact and collaborate to provide Mobile Services not available from the Internet Cloud. This vision is corroborated by two emerging realities: the vehicles pick up too much multimedia information from the environment to possibly upload it all to the Internet Cloud. Moreover, time sensitive safety applications require the data to be processed locally in the vehicles. In this talk we revisit VANET applications and services and propose a Mobile Cloud vehicle services platform. We then discuss a specific example based on content creation, search and dissemination. We conclude by making the case for a uniform services platform that can support a broad range of vendors and applications.
Dr. Mario Gerla is a Professor in the Computer Science Dept at UCLA. He holds an Engineering degree from Politecnico di Milano, Italy and the Ph.D. degree from UCLA. He became IEEE Fellow in 2002. At UCLA, he was part of the team that developed the early ARPANET protocols under the guidance of Prof. Leonard Kleinrock. He joined the UCLA Faculty in 1976. At UCLA he has designed network protocols for mobile, challenged environments including ad hoc wireless clustering, multicast and Internet transport (TCP Westwood). He has lead the ONR MINUTEMAN project in 2000-2005, designing the next generation scalable airborne Internet for tactical and emergency scenarios. He is now leading an NSF sponsored Vehicular Research and Testbed project for safe navigation, content distribution, urban sensing and intelligent transport. Parallel research activities are wireless medical monitoring using smart phones and cognitive radios in urban environments.Dr Gerla has is active in the organization of wireless conferences and workshops, including MedHocNet and WONS. He serves on the IEEE TON Scientific Advisory Board. He was recently recognized with the MILCOM Technical Contribution Award and the IEEE Ad Hoc and Sensor Network Achievement Award, both in 2011.
On-demand video streaming dominates today’s Internet traffic mix. For instance, Netflix constitutes a third of the peak time traffic in the USA. Nearly half of UK online households have accessed BBC’s shows through its on-demand streaming interface, BBC iPlayer. Using UK-wide traces from BBC iPlayer as a case study, this talk will characterise users’ content consumption at scale and discuss techniques that can be deployed at the edge by users to substantially decrease the load on the Internet. We will survey both well-known techniques such as peer-assisted VoD, studying whether it works at scale, as well as new edge-caching mechanisms that can potentially be deployed today. We will conclude by exploring new directions for content-centric network architectures, to address the roots of the pain points observed in our user workload, in a “clean” fashion.
Nishanth Sastry is a Lecturer (roughly equivalent to a US-style tenure track Assistant Professor) at King’s College London. He holds a PhD from the University of Cambridge, UK, a Master’s degree from The University of Texas at Austin, and a Bachelor’s degree from Bangalore University, India, all in Computer Science. He has over six years of experience in the Industry (Cisco Systems, India and IBM Software Group, USA) and Industrial Research Labs (IBM TJ Watson Research Center). His honours include a Yunus Innovation Challenge Award at the Massachusetts Institute of Technology IDEAS Competition, a Benefactor’s Scholarship from St. John’s College, Cambridge, a Best Undergraduate Project Award from RV College of Engineering, a Cisco Achievement Program Award and several awards for his work at IBM. His work has ranged over several layers of the network stack and he is currently involved in building better networked systems by harnessing user-level and social network information.