Quantum mechanics in energy and signal transfer processes

When : Thursday, December 4, 2014 - 15:30
Speaker : Prof. Elisabetta Collini
Affiliation : Dipartimento di Scienze Chimiche, Università di Padova
Where : Aula Magna ‘A. Lepschy’
Description :

One of the most surprising and significant advances in the study of the photosynthetic light- harvesting process is the discovery that the electronic energy transfer might involve long-lived electronic coherences, also at physiologically relevant conditions. This means that the transfer of energy among different chromophores does not follow the expected classical incoherent hopping mechanism, but that quantum-mechanical laws can steer the migration of energy. The implications of such quantum transport regime, although currently under debate, might have a tremendous impact in our way to think about natural and artificial light-harvesting and suggest new directions for the development of artificial devices for the efficient capture and re-use of solar energy. Central to these discoveries has been the development of new ultrafast spectroscopic techniques, in particular two-dimensional electronic spectroscopy, which is now the primary tool to obtain clear and definitive experimental proof of such effects. The possibility of engineer quantum mechanics in artificial systems not only has inspired new guidelines for the design of artificial photosynthetic systems, but also has opened a revolutionary way for the effective use of biological systems and conjugated polymers as quantum devices and quantum resources for signal processing. For example, we have proven that 2D maps can be used to generate multivariate and parallel logic processes.

Future Internet Security and Privacy (challenges)

When : Friday, October 31, 2014 - 14:15
Speaker : Mauro Conti and Moreno Ambrosin
Affiliation :
Where : Aula Magna "Antonio Lepschy"
Description :

The Internet is an amazing success story, connecting hundreds of millions of users. However, in the last decade, there has been a growing realization that the current Internet Protocol is reaching the limits of its senescence. In fact, the way people access and utilize it has changed radically since the 1970-s when its architecture was conceived.This has prompted several research efforts that aim to design potential next-generation Internet architectures. In particular, Content-Centric Networking (CCN) is an emerging networking paradigm being considered as a possible replacement for the current IP-based host-centric Internet infrastructure. CCN focuses on content distribution, which is arguably not well served by IP. Named-Data Networking (NDN) is an example of CCN. NDN is also an active research project under the NSF Future Internet Architectures (FIA) program. FIA emphasizes security and privacy from the outset and by design. To be a viable Internet architecture, NDN must be resilient against current and emerging threats.In this talk, we highlight the main security and privacy issues we identified in NDN. Then, as a representative case, we discuss interest flooding, a possible denial-of-service attack that exploits key architectural features of NDN. We show that an adversary with limited resources can implement such attack, having a significant impact on network performance. We then introduce Poseidon: a framework for detecting and mitigating interest flooding attacks. Finally, we report on results of extensive simulations assessing proposed countermeasure.

Quantum random flip-flop and its applications (RINVIATO)

When : Thursday, October 16, 2014 - 15:30
Speaker : Dr. Mario Stipčević
Affiliation : Ruđer Bošković Institute, Zagreb, Croatia
Where : Aula Magna "A. Lepschy"
Description :

We define a new non-sequential logic element: the random flip-flop (RFF) and show how it can be realized experimentally by using quantum randomness present in a photonic system. By definition RFF operates similarly to the conventional flip-flop except that its clock input functions with probability of 1/2 and otherwise the flip-flop does nothing. Even though the shift from deterministic to random flip-flop may seem minimal, the latter shows an unexpectedly rich behavior and possible areas of application because now not only logic states but also time becomes important. We discuss several digital and analog applications of the novel device in cryptography, over-Turing computing, randomness preserving frequency division, random frequency synthesis and random pulse computing.

Modeling the Pathogenesis, Prevention and Reversal of Type 2 Diabetes

When : Wednesday, October 15, 2014 - 11:00
Speaker : Artie Sherman
Affiliation : Laboratory of Biological Modeling, NIH/NIDDK (Bethesda, USA)
Where : Sala riunioni 201
Description :

It is widely, but not universally accepted, that type 2 diabetes results from the failure of beta cells to compensate for insulin resistant.  Objections have been raised because insulin rises long before glucose (Starling's Law of the Pancreas), which can be interpreted to mean that hyperinsulinemia causes hyperglycemia rather than the reverse.  Moreover, even if one accepts that hyperinsulinemia is a compensatory response, it is not clear what the signal for compensation is if glucose remains normal.   We address this by extending the pioneering mathematical model of Topp et al (J. Theor Biol.  2000 206:605) for regulation of beta-cell mass and diabetes progression.  The revised model includes, among other things, the hypothesis of Dor and colleagues that beta-cell proliferation is governed by insulin secretion rate, not directly by glucose.  We show that the system exhibits threshold behavior resulting from bistability between health and disease.  This accounts for clinical observations that prevention is easier than cure and for the remarkable ability of bariatric surgery to rapidly reverse diabetes. * Joint work with Joon Ha

Modeling Pancreatic Alpha Cells and Interactions with Other Islet Cells

When : Tuesday, October 14, 2014 - 12:00
Speaker : Artie Sherman
Affiliation : Laboratory of Biological Modeling, NIH/NIDDK (Bethesda, USA)
Where : Sala riunioni 201
Description :

There has been much debate about whether glucagon secretion is controlled by intrinsic properties of alpha cells or by paracrine suppression by beta and delta cells because it has been difficult to observe appropriate reduction of secretion in isolated alpha cells. We have previously published a model for the intrinsic response (Biophys. J. 2014 106:741) and confirmed the hypothesis of Rorsman and colleagues that glucose sensing can occur by closure of KATP channels, which inactivates Na+ and Ca2+ channels.  However, the response to glucose is bell-shaped, so either suppression or stimulation can be exhibited depending on KATP expression levels in each cell.  We show now with a tri-hormone islet model that this problem can be solved by paracrine interactions: glucose stimulates insulin and somatostatin secretion, which suppress those alpha cells that would otherwise be active at high glucose.  The model suggests that somatostatin plays the more critical role as it always suppresses secretion, whereas insulin may suppress or stimulate, depending on the state of the individual cell.  The model thus accounts for both the heterogeneity observed in isolated alpha cells and the appropriate stereotypical behavior of the intact islet. * Joint work with Margaret Watts

Analysis of Randomized Work Stealing with False Sharing

When : Friday, June 20, 2014 - 14:30
Speaker : Vijaya Ramachandran
Affiliation : University of Texas at Austin
Where : Sala Riunioni DEI/G (III piano)
Description :

ABSTRACT   We analyze the overhead due to false sharing when parallel tasks are scheduled using randomized work stealing (RWS). We obtain high-probability bounds on the cache miss overhead, including the overhead due to false sharing, for several parallel cache-efficient algorithms when scheduled using RWS. These include algorithms for fundamental problems such as matrix computations, FFT, sorting, basic dynamic programming, list ranking and graph connected components. Our main technical contribution, from which these results follow, is the derivation of nontrivial high-probability bounds on the number of steals incurred by these algorithms in the presence of false sharing, when using RWS. This is joint work with Richard Cole, NYU

Programming Distributed Machine Learning Applications / Frameworks

When : Tuesday, June 3, 2014 - 16:30
Speaker : Vijay Saraswat
Affiliation : IBM TJ Watson Research Center, Yorktown Heights, NY USA
Where : Sala Riunioni DEI/G
Description :

ABSTRACT We propose to use X10 (http://x10-lang.org) to program distributed machine learning applications. Designed for the productivity of Java-like languages, and the performance of MPI-like libraries, at scale, X10 is built around a few core programming constructs for concurrency and distribution. X10 is a statically typed, garbage-collected, parallel, object-oriented programming language that supports dynamic, asynchronous, state-full computations running in a single global address space, partitioned over a cluster of tens of thousands of cores. Programs can run either natively (single binary, running on each node in a cluster), or in a managed environment (cluster of JVMs), either on commodity networks, or high performance networks (e.g. infiniband), or supercomputers such as Blue Gene/Q, IBM P775, or the Japanese K computer. X10 is particularly useful for building libraries / application frameworks involving distributed/global data-structures. In earlier work we have developed high performance, in-memory implementations of application frameworks such as Hadoop Map Reduce, a global matrix library, and a library for globally load-balancing highly irregular computations. Application and libraries such as a non-resilient version of Pregel, or distributed version of the DBScan clustering algorithm, an expectation-maximization algorithm for HMM training etc have been developed in a few hundred lines of code. Additionally, X10 is being used to develop distributed SAT solvers (e.g. Sat X10, DDX10 based on decision diagrams) and adaptive solvers using local search.

Brain-Computer Interface and Motor Learning in Stroke Rehabilitation

When : Friday, May 16, 2014 - 11:00
Speaker : Prof. Junichi USHIBA
Affiliation : KEIO University, Japan
Where : Aula Magna A. Lepschy
Description :

Abstract: Brain computer interface technology (BCI) is of great interest as a potential therapeutic measure for people with severe neurological disorders. We forms an interdisciplinary research team to promote its clinical transfer and to explore underlying neurological mechanisms. In this talk, I will address the tips how neural plasticity for functional recovery from stroke is induced by means of BCI. Some neuroscientific theories and practices related to BCI rehabilitation will be given.Outside the BCI study, our laboratory recently launched robotic rehabilitation and MRI study for visualizing structural plasticity of the brains. The brief summary of these ongoing projects will be also introduced at the latter half of my talk. Bio:2001 Graduation at the Faculty of Science and Technology, Keio University, Japan. March 2003 - Aug 2003  Guest Researcher at the Center for Sensori-Motor Interaction, Aalborg University, Denmark. 2004 received Ph.D. from the School of Science and Technology, Graduate School of Keio University, Japan. 2004 - 2007 Research Associate at the Department of Biosciences and Informatics, Faculty of Science and Technology, Keio University, Japan. 2007- 2012 Assistant Professor at the Department of Biosciences and Informatics, Faculty of Science and Technology, Keio University, Japan. 2013- Present Associate Professor at the Department of Biosciences and Informatics, Faculty of Science and Technology, Keio University, Japan. 2013- Principal Investigator at the Keio Institute of Pure and Applied Sciences.  

Micro-LEDs for Applications in Optogenetics

When : Monday, May 5, 2014 - 14:30
Speaker : Prof. Ulrich T. Schwarz
Affiliation : Università di Friburgo e Fraunhofer - IAF
Where : Aula Magna "A. Lepschy"
Description :

Abstract:In the emerging field of Optogenetics, nerve cells are made sensitive to light, to allow selective stimulation of nerve pulses by light pulses. This technology is used as tool in fundamental neuroscience to explore the mechanisms active in the brain, and to new medical devices which may allow deaf people to hear or blind people to see. In an interdisciplinary approach we develop a new class of optical probes (»optrodes«) by direct integration of tiny light-emitting diodes with edge lengths of a few ten micrometers (μLED) with micro-engineered probes. The actual device is small enough to fit into the cochlea (the innear ear) of a rodent.  This device will be able to stimulate nerve cells in the cochlea in order to generate a hearing sensation excited by light pulses, potentially with a much higher frequency resolution than conventional CIs. In this development, we employ our experience with high power LEDs, now standard in solid-state-lighting, with experience in fabrication of neuro-sensors based on an integration of metal lines with flexible polymers. In particular the thin-film LED is the key enabling technology for the miniaturization of the optical probes.Bio:Ulrich T. Schwarz received his Ph. D. degree and habilitation in physical science from the University of Regensburg, Germany, in 1997 and 2004, respectively. Since 2009 he is full professor (W2) at the Institute for Microsystems Engineering (IMTEK) at Freiburg University. At the same time he is group leader at the Fraunhofer Institute for Applied Solid State Physics (IAF) in Freiburg. His research is in the field of optoelectronic devices based on group-III-nitrides, i.e. light emitting diodes and laser diodes in the violet to green spectral region. Another focus is on singular optics, in particular polarization singularities. Awarded by the Alexander von Humboldt foundation with a Feodor Lynen scholarship he spent two postdoctorial years (1997–1999) at Cornell University, Ithaca, NY, with research on intrinsically localized modes. In 2001 he joined the group of Prof. R. Grober at Yale University, New Haven, CT. In 2006/2007, he visited Kyoto University, Kyoto, Japan, with an invited fellowship (long-term) awarded by the Japanese Society for the Promotion of Science. He is senior member of the Optical Society of America (OSA), and member of the American Physical Society (APS), the International Society for Optical Engineering (SPIE), and the German Physical Society (DPG).

Biological Network Analysis

When : Tuesday, March 11, 2014 - 15:00
Speaker : Simona Rombo
Affiliation : Universita' di Palermo
Where : Aula Magna "A. Lepschy"
Description :

ABSTRACT For many years now, the analysis of biological sequences associated to cellular components (e.g., proteins, DNA and RNA) have played a key role in understanding the mechanisms inside the cell. After that the genome coding of several organisms has been completed, a large deal of attention is turning towards studying how cellular components interact with each other to accomplish the biological functions of the cell. As an example, to fully understand cell activity, proteins cannot be analyzed independently of the other proteins of the same organism, because they seldom act in isolation to perform their tasks.The interactions among cellular components can be modelled by graphs, usually called biological networks, where nodes represent components and edges their interactions. In the past few years, biological networks received great attention, also due to the increasing amount of interactions among cellular components retrieved by experimental, computational and high-throughput techniques. Aligning, querying and partitioning such graphs is useful in many applications, for example to infer knowledge about conservation across species or to predict the function of unknown proteins.In this talk, we illustrate the main computational problems involving biological networks and some of the most recent techniques proposed to solve them, highlighting the open issues and the new directions in this context.           

Systems, Security, and Privacy Issues in Mobile Networking

When : Wednesday, February 26, 2014 - 11:00
Speaker : Prof. Alessandro Mei
Affiliation : Sapienza University
Where : Aula Magna “A. Lepschy”
Description :

Smartphones are killing PCs. Smartphone's market vastly outperformed. PC's market in the last few years. Not only that, more that 50% of cellular phones sold today are smartphones, pushing classical mobiles into a corner. And more than 50% of Facebook connections come from a smartphone. Smartphones are a modern revolution that is changing the way we live.In this talk, we will present some of the work in systems, security,and privacy  in mobile networking that is being performed at Sapienza in our group. We will see how modern mobile operating systems are actually becoming hybrid cloud-mobile operating systems. We will see some of the new privacy and security concerns of mobile networking, and, lastly, how we can get incredibly reliable information on the socio-economic status of groups of people just looking at part of the "background radiation" coming from our smartphones.

Present and Future of GaN Power Devices

When : Monday, February 24, 2014 - 11:00
Speaker : Dr. Daisuke Ueda, Ph.D.
Affiliation : Advanced Technology Research Laboratories & Device Solution Center, Panasonic Corp., Japan
Where : Aula Magna "Antonio Lepschy"
Description :

GaN-based power devices moving toward industrialization. Normally-offcharacteristics has been realized by introducing p-type AlGaN gatestructure, where holes can be injected into the channel reducing theon-resistance by conductivity modulation. Experimentally fabricated GaNGIT inverter system attained the world-highest conversion efficiencyover 99.3%. AlGaN/GaN Natural Super Junction was also developed toovercome the trade-off between the blocking voltage and theon-resistance. Further, we propose a fusion of microwave and powerdevices as Drive-by-Microwave technology, which will enable a new matrixconverter system in a simple manner. Direct liquid-immersion packagetechnology, where GaN chip are mounted inside the heat-pipe, ismentioned. The technology contributes to the dramatic reduction ofjunction temperature, which had been a problem caused by the increasedpower density of GaN devices.