Challenges and Opportunities of Connected Context Computing

Size: px

Start display at page:

Download "Challenges and Opportunities of Connected Context Computing"

Emerald Webb
5 years ago
Views:

1 Keynote I Challenges and Opportunities of Connected Context Computing Yen-Kuang Chen Ph.D., IEEE Fellow Principal Engineer, Intel Corporation Associate Director, Intel-NTU Connected Context Computing Center Adjunct Professor, Dept. of Electrical Engineering, National Taiwan University We are seeing a paradigm shift, called connected context computing, where everything can sense information, communicate the data with one another, collaboratively analyze the context, and take proper actions to help humans. This new paradigm shift is enabled by the maturity of a few game changing capacities, such as, sensors, internet, wireless communication, cloud computing, and machine learning algorithms. When the system can sense intelligently and respond proactively, the service can be provided before the human asks for it. This will bring user experiences of the computing systems to a completely different level. However, in order to fulfill the requirement of this new paradigm, there are many challenges in terms of smartness, simplicity, security, scalability, and sustainability. For example, how can we efficiently analyze the sensor data from different sources, how can we easily deploy and manage sensor systems, how can we provide end-to-end security, how can we seamlessly connect everything, and how can we provide perpetual intelligent sensors and objects? This talk will discuss the challenges and opportunities of connected context computing. Yen-Kuang Chen is a Principal Engineer at Intel Corporation and Associate Director of Intel-NTU Connected Context Computing Center at National Taiwan University. His research interests include Internet of Things, machine-to-machine, multimedia signal processing, video compression, parallel processing, and computer architecture. He is a key contributor to SSSE3 and AVX instruction sets on Intel Core processor family. He has 30+ US patents, 20+ pending patent applications, and 85+ technical publications. He has served as an editorial board member of 5+ journals and a program committee member of 40+ international conferences, on multimedia, video compression/communication, image/signal processing, VLSI circuits and systems, parallel processing, and software optimization. He is an invited participant to 2002 Frontiers of Engineering Symposium (National Academy of Engineering) and to 2003 German-American Frontiers of Engineering Symposium (Alexander von Humboldt Foundation). He received his Ph.D. from Princeton University, and is an IEEE Fellow. xliv

2 Keynote II Multimedia Sensing, Processing and Applications in Internet of Things. Huadong Ma Professor and Ph.D School of Computer Science, Beijing University of Posts and Telecommunications, The Internet of Things (IoT) can enable the interconnection and integration of the physical word and the cyber space, and has been widely considered as the key technology for sensing and understanding the physical scene. As we know, 90% of our sensory information is video and audio, thus we need to study multimedia computing in IoT in order to have complete and accurate understanding of the physical scene. In this talk, we first introduce the challenges of multimedia computing in IoT. Combing our on-going works, we present some researches on multimedia sensing, processing, and applications in the IoT environment. Finally, we analyze the open issues and outline the future works on multimedia computing in IoT. Dr. Huadong Ma is a Chang Jiang Scholar Professor and Executive Dean, School of Computer Science, Beijing University of Posts and Telecommunications (BUPT), China. He is also Director of Beijing Key Lab of Intelligent Telecommunications Software and Multimedia, BUPT. He is Chief Scientist of the project Basic Research on the Architecture of Internet of Things supported by the National 973 Program of China. He received his PhD degree in Computer Science from the Institute of Computing Technology, Chinese Academy of Science in From 1999 to 2000, he held a visiting position in the Department of Electrical Engineering and Computer Science, The University of Michigan, Ann Arbor, Michigan, USA. He was a visiting Professor at The University of Texas at Arlington from July to September 2004, and a visiting Professor at Hong Kong University of Science and Technology from Dec to Feb His current research focuses on multimedia computing, sensor networks and Internet of things, and he has published over 150 papers in journals (such as ACM/IEEE Transactions) or Conferences (such as IEEE INFOCOM, ACM MM) and 4 books on these fields. He was awarded National Funds for Distinguished Young Scientists in xlv

Keynote III Energy-Proportional Networked Systems Dejan Kostic Assistant Professor and Phd School of Computer and Communications Sciences at EPFL CH - 1015 Lausanne Switzerland Energy consumption of

3 Keynote III Energy-Proportional Networked Systems Dejan Kostic Assistant Professor and Phd School of Computer and Communications Sciences at EPFL CH Lausanne Switzerland Energy consumption of the Internet is already substantial and it is likely to increase as operators deploy faster equipment to handle popular bandwidth-intensive services, such as streaming and video-ondemand. Further, the CMOS technology that is predominantly used in the ICT equipment is reaching its energy efficiency limits. The net effect of these issues is that the Internet will shortly hit the power delivery and cooling limits while the hardware is trying to sustain the exponentially increasing traffic demand. In this talk, we will discuss the efforts in designing energy-proportional networked systems that mitigate this problem by trying to use the minimal amount of energy to satisfy each offered load (traffic). The goal here is to compute the minimal set of system elements that can accommodate the current load, while letting remaining equipment enter a power saving mode. A similar issue arises when trying to map workloads in the cloud to the minimal set of virtualized resources that can satisfy the performancerelated service level objectives. The key challenge here is being able to quickly adapt to frequent traffic/workload variations, as this task often translates to a computationally intensive problem. We use several simple, deployable techniques that include precomputation of network routing, and memoization (caching) of previously seen cloud workloads. We have managed to approximate energy-proportionality in networks both for the access as well as the network core and datacenter networks. We also apply some of our energy-proportional networking techniques to the problem of resource allocation in virtualized environments, and demonstrate large savings and fast adaptation times to changing workloads. Dejan Kostic obtained his Ph.D. in Computer Science at the Duke University. He spent the last two years of his studies and a brief stay as a postdoctoral scholar at the University of California, San Diego. He received his Master of Science degree in Computer Science from the University of Texas at Dallas, and his Bachelor of Science degree in Computer Engineering and Information Technology from the University of Belgrade (ETF), Serbia. Since 2006, he has been working as a tenure-track assistant professor at the School of Computer and Communications Sciences at EPFL (Ecole Polytechnique Fédérale de Lausanne), Switzerland. In 2010, he received a European Research Council (ERC) Starting Investigator Award. His interests include Distributed Systems, Computer Networks, Operating Systems, and Mobile Computing. xlvi

Keynote IV Smart (i)cards and ithings Pierre Paradinas Chair of Embedded and Mobile Systems CNAM (Conservatoire National des Arts et Métiers) 292 rue Saint Martin 75141 Paris CEDEX 03 France The talk

4 Keynote IV Smart (i)cards and ithings Pierre Paradinas Chair of Embedded and Mobile Systems CNAM (Conservatoire National des Arts et Métiers) 292 rue Saint Martin Paris CEDEX 03 France The talk will present and introduce the smart card technology and its ecosystem. The relation of smart card and internet will be covered. A focus on specific features of the card and the relation with the IoT will be discuss. Pierre Paradinas is Professor at Cnam (Paris) and owned the «Embedded and Mobile Systems chair». He received a PhD in Computer Science from the University of Lille (France) in 1988 on smart cards and health care application, and was founders of Biocarte SA (design, development and distribution of patient cards). Pierre Paradinas joined Gemplus in 1989, and was successively researcher, internal technology audit, Advanced Product Manager and launched the card based on Data base engine (CQL), Director of a common research lab with universities and National Research center (RD2P). He set up the Gemplus Software Research Lab in He was also appointed technology partnership Director in 2001, based in California until June He was the Gemplus representative at W3C, ISO/AFNOR, Open Card Framework and Java Community Process, co-editor of the part 7 of ISO7816, director of the European funded project Cascade project where the first 32-Risc microprocessor with Java Card was issued. He served as Director of Technology Development at Inria : the national institute for research in computer science and control, operating under the dual authority of the Ministry of Research and the Ministry of Industry, is dedicated to fundamental and applied research in information and communication science and technology (ICST). He owns 27 years of experience in smart cards technology and is deeply involved in the technology and research of smart cards, as program committee member (Chair of the CARDIS 04) and expert for European Commission and French government. xlvii

5 Keynote V Cyber-Physical Systems: (Some) Opportunities and Challenges Peter Marwedel Professor and Ph.D, TU Dortmund, Dortmund, Germany The recently introduced term cyber-physical systems (CPS) describes systems integrating computations with the physical environment. The talk will start with a view onto a representative set of future CPS applications. In this way, we will emphasize the enormous potential of CPS. This introduction will be followed by a description of some key challenges. After a brief overview of such challenges, we will provide a closer look at two of the challenges: timing predictability and education. For timing predictability, we will demonstrate how a reconciliation of timing analysis and compilation can provide a level of predictability which cannot be achieved be each of these techniques in isolation. Furthermore, we will present requirements for education in cyber-physical systems and possible solutions which cross the boundaries of the areas involved. Solutions include specializations of existing curricula as well as new integrated CPS curricula. Peter Marwedel studied physics at the University of Kiel, Germany. He received his PhD in physics at this university in As a Post-Doc, he was one of the pioneers in high-level synthesis based on the hardware description language MIMOLA. In 1987, he received the Dr. habil. degree (a formal qualification for becoming a professor) in computer science from the same university. In 1989 he became a full professor at the faculty of informatics, TU Dortmund, Germany. He also was a visiting professor at UC Irvine (USA) in Since 1997 he is CEO of Informatik Centrum Dortmund, which is a spin-off of TU Dortmund. Dr. Marwedel was a cluster leader of the European network of excellence on embedded and real-time systems, ArtistDesign and is the chair of the follow-up organization EMSIG. The collaborative research centre SFB 876 at TU Dortmund selected him as vice chair. Dr. Marwedel is the author or co-author of over 120 books and papers and is the author of a popular text book entitled Embedded System Design: Embedded Systems Foundations of Cyber-Physical Systems. His research interests include embedded software, low-power software, compilers for embedded processors, timing predictability, high-level synthesis, handling of faults in software and test program generation for processors. Dr. Marwedel won the teaching award of TU Dortmund in 2003 and is Dean of Studies at the Department of Informatics at TU Dortmund and chairs the education committee of the academic senate at TU Dortmund. He is an IEEE fellow and a fellow of the DATE conference. xlviii

Durham Research Online

Durham Research Online Deposited in DRO: 29 August 2017 Version of attached le: Accepted Version Peer-review status of attached le: Not peer-reviewed Citation for published item: Chiu, Wei-Yu and Sun,