From Victor Veselago and Leon Chua to Today

Abstract:

Metamaterials were first proposed in the late 60s by Victor Veselago, and have since demonstrated many novel properties, such as negative refractive indexes, superlensing, anomalous reflection, perfect absorption and advantages in many RF, microwave and antenna designs. Tunable metamaterials have also shown additional functionality as well as programmability by incorporating software control.

Since the prediction of the memristor in the early 70s by Leon Chua, many memristor designs have experimentally verified its existence. The definition of memory components has also been expanded to include memcapacitors and meminductors.

In this presentation, an introduction to metamaterials and memory components will be presented, and it will be shown how these two concepts can be merged together. The design of memory components for metamaterial applications will be shown, and this will be incorporated in a programmable metasurface design.



Tensor-Based Nonlinear Classfier for High Order Data Analysis

Abstract:

Recent advances in sensing technologies have stimulated the development and deployment of sensors that can generate large amounts of high-order data. Interdependencies between information from different data modalities can improve the performance of data classification techniques. However, exploitation of high-order data raises new research challenges mainly due to the high dimensionality of the acquired information and, depending on the application at hand, the limited number of labelled examples.

Tensor based model can tackle the dimensionality problem by (i) significantly reducing the number of training parameters that need to be estimated during training, and hence the required training samples, and (ii) by retaining the spatial structure of the input samples. In this talk we will present a tensor based nonlinear classifier for high order data analysis, which is based on a modification of a feedforward neural network, such that its weights satisfy the rank-1 canonical decomposition. Finally, we will also present an algorithm for training it.



SiLago: The Next Generation Synchoros VLSI Design Platform to enable affordable Billion Gate ASICs

Abstract:

VLSI design community faces the challenge of unscalable large engineering and manufacturing costs and 2-4 orders loss in computational efficiency compared to hardwired solutions. As a solution, SiLago raises the abstraction of physical design platform from the present day boolean level standard cells to micro-architectural level SiLago (Silicon Large Grain Objects) blocks as the atomic physical design building blocks and introduce a grid based synchoros VLSI Design scheme to compose arbitrary designs by abutting SiLago blocks to eliminate the logic and physical syntheses for the end user. The word synchoros is derived from the Greek word for space – choros. Synchoros objects discretize space uniformly with the grid, the way synchronous objects discretize time with clock ticks. The synchoros design style and micro-architectural level physical design enables SiLago method to rapidly explore the higher abstraction design space and generate valid VLSI designs at GDSII level corresponding 10-100 million gate complexity in minutes with an engineering effort comparable to programming. The SiLago method also holds the promise to eliminate the mask engineering cost.



Crosstalk-aware Routing Spectrum Assignment and WSS Placement in Optical Networks

Abstract:

Due to crosstalk-induced interactions among different connections, malicious high-power jamming signals can potentially spread widely in a transparent optical network. Moreover, due to imperfect port isolation in wavelength selective switches (WSSs), present within optical switching nodes, crosstalk also affects the quality of the transmitted signal. Therefore, it is necessary to design an optical network in a way that the effect of the crosstalk is minimized, while at the same time keeping the cost of the network low. In this talk, we present an appropriate WSS placement and crosstalk-aware Routing and Spectrum Assignment (RSA) algorithm in flexible grid optical networks, in the form of an Integer Linear Program (ILP) formulation and a heuristic algorithm analogous to vertex coloring in order to minimize the impact of the crosstalk effect.



Collimated light propagation: The next frontier in underwater wireless communication

Abstract:

Traditional underwater communication systems rely on acoustic modems due their reliability and long range. However their limited data rates, lead to the exploration of alternative techniques. In this talk, we briefly go over the potential offered by underwater wireless optical communication systems. We then summarizes some of the underwater channel challenges going from severe absorption and scattering that need to be surpassed before such kind of systems can be deployed in practice. We finally present some of the on-going research directions in the area of underwater wireless optical communication systems in order to (i) better characterize and model the underwater optical channel and (ii) design, develop, and test experimentally new suitable modulation and coding techniques suitable for this environment.



Resilient Architectures and Algorithms for Generation Control of Inertial-Less AC Microgrids

Abstract:

This talk discusses the problem of frequency regulation in islanded ac microgrids with no inertia, i.e., those consisting entirely of generators interfaced through power electronics. The control architecture we propose to achieve this is designed to drive the average frequency error to zero while ensuring that the frequency at every bus is equal and that the operating point that results is stable. We also introduce a distributed implementation of the proposed control architecture that relies on a combination of several distributed algorithms. Collectively, these algorithms eliminate the need for acentralized entity with complete knowledge of the network, its topology, or the capabilities or properties of the generators and loads therein. Moreover, the distributed implementation we propose relies on minimal measurements, requiring only that the power injection at each bus be measured. Additionally, by eliminating the need for a centralized processor and a communication network connecting it to each generator, these distributed approaches can achieve higher system-level reliability, adaptability, and resilience.