http://www.doe.carleton.ca/seminars/index.php The seminars from the leading academic/industry personalities are conducted regularly by the Department of Electronics. en-us webmaster@doe.carleton.ca (Scott Bruce) http://www.doe.carleton.ca/seminars/index.php http://www.doe.carleton.ca/images/cu_logo1.gif http://www.doe.carleton.ca/seminars/show_one.php?id=77<p>New renewable energy technologies, combined with a broad suite of energy-efficiency advances, are spreading widely all over the world to reduce greenhouse gas emissions and to relieve energy crisis caused by exhaustion of fossil fuels. With the development trend of energy, it is expected that distributed energy resources (DER) including distributed generation (DG), energy storage and demand response (DR) will be increasingly introduced into electric power systems in the near future. As a radically different way to deliver electricity, DER technologies are expected to make fundamental changes to the current power system structure.nnOne of the most difficult challenges that DER interconnection is facing is to provide state-of-the-art techniques for DER control and protection. This talk will highlight my research effort on anti-islanding protection for grid-connected DG systems, which provides a better understanding of how to design optimal islanding detection method. In the first part, I will present the investigation results of the positive feedback anti-islanding scheme for inverter-based DG. Then a systematic application guideline for the studied anti-islanding scheme will be introduced, and the analytical analysis and comparative case studies will be provided to show the effectiveness of the designed guideline. At the end of the seminar, I will share my vision of the future innovation and applications of energy technologies in renewable energy and smart grid fields.</p> Fri, 02 Mar 2012 11:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=77http://www.doe.carleton.ca/seminars/show_one.php?id=76<p><span style="font-size: small;">Semiconductor technologies exhibited explosive growth in complexity and speed over the last two decades. Since the early 1980s, the device sizes have scaled down from few micrometers to tens of Nano-meters and the operating frequencies have increased from a few megahertz to several gigahertz. Also, the spacing between devices and interconnect have dramatically decreased due to the continuous scaling down of the technology feature size. These trends have led to issues and challenges in the design and analysis of high performance integrated circuits that previous generations did not exhibit. Most of these issues are at the circuit and interconnect (physical) levels. Also, these issues are expected only to increase in importance in future generations of integrated circuits. This talk will overview the most important challenges for electronics design in the nano-scale.</span></p> Tue, 28 Feb 2012 12:30:05 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=76http://www.doe.carleton.ca/seminars/show_one.php?id=75<p><span style="font-size: small;">Sacling as we know it is taking a different direction from the last three decades. Chips with tens of billions of transistors and hundreds of cores are expected to be the future of scaling. These chips will achieve performance through parallelism and application specific optimized cores. This trend will use superior technologies to integrate more cores on a chip rather than to push the frequency envelope as in the past. It is expected that every aspect of design and analysis will need to be modified to accommodate this new platform and trend. There is a clear need for new CAD tools and design methodologies that are very different from existing tools in both their focus and scope. This talk will delve into the specific challenges with respect to both design and CAD that is required for these many core chips. The talk will also provide an overview into the market and technology factors guiding and driving this trend. Attendees will be provided with insight into both present and future research vectors to support this nascent exponential.</span></p> Tue, 28 Feb 2012 11:30:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=75http://www.doe.carleton.ca/seminars/show_one.php?id=74<p>Energy is one of the top priorities in the world and smart grids are the centerpiece of this energy focus. Grid design, control and stability are the main objectives of smart grid technology in order to enhance the voltage stability of electric power distribution systems during faulty conditions and disturbances. Analysis and benefits of implementing smart grids based on multi-agent systems (MAS) show that it is a suitable technology for the complex and highly dynamic operation of a power system network.nThe existing power grid suffers from the lack of pervasive and effective communications, monitoring, fault diagnostics, and automation control, which further increase the possibility of a region-wide system breakdown due to the cascading effect that can be initiated by a single fault.nCurrently, for the power system, voltage control systems are centralized and operated through a central computer which supervises the output of all generators and adjusts optimally the voltage set points of these generators. This centralized regulation algorithm must know the whole network configuration and therefore for a large-scale power system, it may become difficult to perform a centralized control system. This motivates us to study and find efficient and secure voltage control mechanism in a power system by identifying the most appropriate controls based on decentralized and distributed control.nThis presentation firstly presents a definition and vision of the smart grid and its key areas including: Sensing and Measurement, Advanced Control Methods, Advanced Components and Integrated Communication. Secondly, an optimal electrical network graph partitioning technique is presented that divides a power network into appropriate regions to eventually prevent the propagation of disturbances and minimize the interaction between these regions. The optimized number of partitions is found based on the bus voltage sensitivity to the disturbances being applied to the loads in each region. A number of representative buses which are labeled as pilot buses are established and these are identified in each region displaying the critical point for secondary voltage control. The graph theory applied to this situation has the ability to simplify and decompose large connected power networks.</p> Fri, 20 Jan 2012 13:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=74http://www.doe.carleton.ca/seminars/show_one.php?id=73<p><span style="font-size: small;"><span style="font-size: 10pt; line-height: 115%;">A PLL is perhaps the most widely used mixed-signal circuit block in a system-on-chip (SoC). Advancements in process technology offer advantages as well as challenges for the design of PLLs. This tutorial will highlight the challenges and present design techniques to overcome them. A brief discussion of PLL basics is presented first. Noise transfer functions from various points in the loop to the output are shown and the importance of having a low oscillator-gain is highlighted. Optimization of loop parameters to minimize phase-noise and area is emphasized. Challenges posed by nanometer CMOS technology are then discussed. Circuit techniques to overcome these are discussed for critical building blocks. Process and temperature compensation techniques to minimize VCO gain, capacitance multiplication techniques to minimize loop-filter area, and a rail-to-rail output swing charge-pump with matched up/down currents to minimize spurs are a few of the circuit techniques that will be discussed.</span></span></p> Thu, 08 Dec 2011 18:30:40 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=73http://www.doe.carleton.ca/seminars/show_one.php?id=72<p>This presentation discusses the basic concepts and development of multiportn interferometer techniques (six-port in particular) and demonstrates theirn applications in the design of both ultra-wideband and carrier-basedn transceivers at microwave and millimeter-wave frequencies. Such emergingn techniques can be extended to the design of transceivers over terahertz andn optical ranges. Various architectures of the multiport circuits aren discussed with respect to different applications. Practical implementationsn under different technological platforms are described with simulated andn measured results for QAM and UWB applications. In particular, then equivalence between the multiport junction and mixing circuit is highlightedn for direct frequency conversion and translation. It is shown that then proposed multiport technique is very promising and flexible for the low-costn design of microwave and millimeter-wave systems such as software-defined and cognitive radio architectures.</p> Mon, 25 Jul 2011 12:00:28 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=72http://www.doe.carleton.ca/seminars/show_one.php?id=71<p class="MsoPlainText"><span style="font-size: 11pt; font-family: 'Arial','sans-serif';">This talk will present the current Mobile Telecommunication Systems in use and Next Generation Mobile Telecommunication Systems to be used in future in China. The key technologies and challenges are given in the end.</span></p> <p> </p> Tue, 07 Jun 2011 14:00:30 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=71http://www.doe.carleton.ca/seminars/show_one.php?id=66<h1>Topic 1: Cross-Layer Design of Smart Antenna Systems</h1> <p>Smart Antenna Systems use the additional degrees of freedom offered by their multiple antennas to exploit, among other things, multipath in the propagation environment. Therefore, by construction, antenna design of smart antenna systems cannot be assessed by simple performance metrics such as gain, polarization and efficiency alone. At a minimum, performance has to be considered in the context of the nature and degree of the multipath. Capacity, the maximum possible throughput, is an appropriate performance metric when the antennas are properly combined with their propagation environment but nothing more is known about the system. When, additionally, the specific Link and Media Access Control (MAC) layer characteristics of the system are taken into account, the actual throughput of the communication link becomes a more appropriate performance metric. A Cross-Layered design approach of Multiple Input Multiple Output (MIMO) antenna systems is presented in this talk. An electromagnetics exact formulation from baseband-to-baseband of a Smart Antenna System is given. The formulation consists of full wave analyses of the antenna arrays involved on both sides of the link and a plane wave decomposition for the propagation environment. Subsequently, the baseband signals are fed into link simulators, specific for each system of interest, to provide estimates of the Bit Error Rate (BER) and throughput. Calibration and Channel estimation algorithms are described for Time Division Duplex (TDD) systems, such as the IEEE 802.16 (WiMAX) and TDD LTE. The state of the art in designing antennas for terminals and for base stations is outlined. Examples of actual product designs for WiMAX and IEEE 802.11n are also given. </p> <p> Finally, the talk ends with some recommendations on research topics to further the state of the art.</p> <h1>Topic 2: Electromagnetic Design for Wireless Applications and Multidisciplinary Optimization</h1> <p>This presentation starts with several specific electromagnetic design examples for wireless applications. These examples include antennas for cellular handsets, RFIDs as well as electromagnetic interference solution concepts. Various characteristics of advanced design methods are then examined. The case is made that multidisciplinary design methods need to be developed and employed for efficient solution of complex problems. At present, multidisciplinary issues encountered at the design of feature rich products are solved by intense communications between the design groups of interacting disciplines. The design of today’s challenging products demands the same and higher degree of communications between the tools used by interacting disciplines. An electromagnetic and structural design example of a cell phone drop to the floor (impact analysis) is used to elucidate the concepts discussed. Additionally, an outline of a framework capable of addressing concurrent optimization of multiple disciplines and of complex products is presented. The seminar ends with a list of proposed problems that need to be solved so that maximum efficiency can be achieved in solving the complex problems of the future.</p> Fri, 25 Mar 2011 13:30:45 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=66http://www.doe.carleton.ca/seminars/show_one.php?id=65<p>The power system is in transition to a grid structure capable of delivering optimal electric power by incorporating enabling technologies such as advanced control strategies, integrated communication, and improved components and power electronic interfaces. Simultaneously, the power system will need to fulfill more stringent operational constraints.</p> <p>To this end, the next generation power system, sometimes referred to as the ``smart grid,'' aims to realize the full potential of the infrastructure. However, this will cause the power system apparatus to operate closer to their operational limits, which may transiently result in violation of dynamic operational limits and even instability. This problem is exacerbated in a small scale power system such as a microgrid or an active distribution system (ADS), in which resources such as power generation capacity are relatively limited. In these systems, due to relatively frequent changes in the in-service status of components, a controller designed and tuned for one system configuration may not perform well for another configuration. Thus, the problem of effective and fast mitigation of transients, especially when the initial operating point is close to the limits, is of more significance.</p> <p>This presentation focuses on two control strategies to address this problem. First, it develops a hierarchical framework for the microgrid control to devise intermediate set points subsequent to a major change or disturbance. Second, it proposes a distributed control strategy for transient response shaping of the controllable devices of the microgrid using local measurements. The proposed control strategies and their performance are discussed and evaluated, and their technical feasibility is established through a number of case studies.</p> Tue, 14 Dec 2010 09:30:43 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=65http://www.doe.carleton.ca/seminars/show_one.php?id=64<p>The demand for high speed and high bandwidth of data transmission in communication systems is rapidly growing. Electronic processing of optical signals currently used in communication networks requires optical-to-electrical and then back to optical (OEO) conversion, which involves a lot of bulk optical components, separately fabricated, tested, packaged, and then assembled and aligned. The role of photonic integration at present is to replace the expensive bulk optics in transmitters and receivers by integrated optical circuits containing all the necessary optical components on a single chip. Practical realization of integrated optics requires a single robust material platform for integrating multiple optical components, including lasers, modulators, detectors, and passive optical devices, on a single chip.</p> <p>The projects that I am working on as a postdoc at the University of Toronto are aimed at developing integrated optical devices in Aluminum Gallium Arsenide (AlGaAs). It is a very promising material for integrated optics due to its mature fabrication technology, multifunctionality, and flexibility in tailoring integrated optical devices. AlGaAs can serve as an active medium for integrated lasers and detectors, and can also be made passive. In addition, AlGaAs can exhibit efficient nonlinear interactions with a negligible multiphoton absorption, which is especially important for wavelength conversion devices that are used in wavelength division multiplexing (WDM) communication networks. It is thus possible to accommodate all types of integrated optical devices on a single chip based on a single material platform.</p> <p>In my talk, I will discuss wavelength conversion by cross-phase modulation and four-wave mixing in AlGaAs with a specially designed wafer composition. I will also talk about our recent work on integrated pulse shapers. These devices have a broad range of applications, including communication networks, nonlinear photonics, signal processing, optical computing, coherent molecular excitation, and many more. My research is mostly focused on AlGaAs material platform, but we also look into other promising integrated technologies. I will briefly overview our work on pulse shaping devices in femtosecond-laser-written glass Bragg gratings. Finally, I will discuss the promise and future directions of integrated optics based on AlGaAs that could potentially become a material platform for all-optical communication systems of the future. </p> Thu, 09 Dec 2010 15:00:15 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=64http://www.doe.carleton.ca/seminars/show_one.php?id=63<p>With the recent upswing in the economic activities in countries such as India and China , global focus on educational needs of such countries has significantly increased. There has been a significant policy shift in the education related governing bodies of these countries in the recent years, to facilitate global education partnerships.</p> <p>In the last few years a number of initiatives have come up in Ontario and in Canada to support international academic collaborations. For example, these include the MITACS Globalinks (for example, for internships from Indian Institute of Technologies (IITs), BITS, and other selected universities), Canadian Commonwealth Scholarships (for graduate students only) and the recent OMG Student Exchange Program (for undergraduate and graduate students) which was established in 2007 by the Government of Ontario as an initiative for the development of post-secondary educational linkages between Ontario and the Indian states of Maharashtra and Goa with the aim of promoting the reciprocal flow of knowledge.</p> <p>In this talk, emerging global educational trends with focus on student demands and career opportunities will be discussed. Global education partnerships in the context of Canadian institutions will also be covered.</p> Tue, 30 Nov 2010 13:00:18 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=63http://www.doe.carleton.ca/seminars/show_one.php?id=62<p>Bragg Gratings (BGs) are perfect candidates for the study of dynamic nonlinear optical effects: the (one-dimensional) photonic bandgap, in combination with a Kerr-type medium, leads to ultrafast (10s of femtoseconds), all-optical functions such as switching, intensity modulation and limiting. A drawback of all-optical nonlinear switching is the high intensity requirements due to the small nonlinearity of conventional, silica optical fibers. This disadvantage can be overcome using (1) filed-enhanced BG designs such as a grating with a phase-shift in the middle, or/and (2) highly-nonlinear materials such as chalcogenide or bismuth-oxide glasses. Benefits and drawbacks of both techniques will be discussed in the presentation, supported by the experimental demonstrations.</p> <p>Two of my latest research works involving the applications of a tilted fiber Bragg grating (TFBG) in the microwave waveform generation and ultrafast optical signal processing are reported in this presentation. In the first work, a TFBG as an optical spectral shaper to generate a chirped microwave waveform in a spectral-shaping and wavelength-to-time mapping system is proposed and demonstrated. </p><p> Two TFBGs with tilt angles of 10o and 12o are fabricated and employed to generate two chirped microwave waveforms with chirp rates of 0.00700 and 0.00756 GHz/ps, respectively. In the second work, a tunable temporal photonic fractional differentiator implemented based on a TFBG is proposed and demonstrated. The fractional differentiation of an optical Gaussian pulse with a bandwidth of 40 GHz is demonstrated, in which the fractional order is continuously tuned from 0.81 to 1.42.</p> Mon, 01 Nov 2010 10:00:34 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=62http://www.doe.carleton.ca/seminars/show_one.php?id=61<p>Photonic crystal based devices are gaining popularity in optical communication systems. The photonic crystals are realized by introducing periodicity in the material properties like refractive index. Light is trapped in a photonic crystal due to photonic band gaps. A 2-D photonic defect can be used as an optical waveguide and point defects inside the waveguide can be used to alter the energy flow inside the waveguide. Generally, time consuming computational methods like FDTD, are used for modeling the photonic defects. Developing of electrical models for the photonic defects will help is designing photonic devices using circuit theory and consequently will be computationally more efficient. Electrical models for the photonic defects and their application like optical filters, sensors, etc will be discussed. Use of photonic band gap in realizing various dispersion and energy confinement characteristics in an optical fiber will also be discussed.</p> Tue, 12 Oct 2010 15:00:18 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=61http://www.doe.carleton.ca/seminars/show_one.php?id=60<p> Stringent power requirements in the chips by the International Technology Roadmap for Semiconductors (ITRS) dictate replacement of silicon dioxide as it has already reached the direct tunneling regime. Therefore, for high speed and low power applications high-k dielectric materials are being integrated into standard CMOS technologies. At present, reliability requirements of advanced gate stacks with high-k dielectrics are of intensive research interests as these high-k dielectrics needs to meet the silicon dioxide standards. In this talk some of the inherent asymmetry on breakdown characteristics of interfacial layer (IL) and high-k layer in the overall gate stacks breakdown will be discussed. Gate stack’s response to many degradation mechanisms such as charge trapping and defect generation, soft breakdown, progressive breakdown and finally hard breakdown will be evaluated as a function of interfacial layers, grown on various process conditions. Correlation of stress-induced leakage current (SILC) with the breakdown behavior will be outlined. </p> Mon, 09 Aug 2010 12:15:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=60http://www.doe.carleton.ca/seminars/show_one.php?id=59<p>Over the past ten years, the threat of IEMI has been studied by many scientists and engineers throughout the world. The vulnerabilities of modern electronic systems have been evaluated and the types of weapons that can be built have also been determined. It is now possible to establish the types of protection that can be applied to reduce this threat in a cost-effective fashion. This presentation will review the history of this work and will discuss the new protection approaches that have been recommended and are being standardized by the International Electrotechnical Commission (IEC).</p> Mon, 05 Jul 2010 14:00:18 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=59http://www.doe.carleton.ca/seminars/show_one.php?id=58<p>Recent advances in computational electromagnetic tools made antenna design possible along with integration of antennas on various ground, sea and air platforms. Full wave techniques such as Method of Moments (MoM), Multilevel Fast Multipole Method (MLFMM) along with asymptotic techniques such as Physical Optics (PO), Geometrical Optics (GO) and Uniform Theory of Diffraction (UTD) will be reviewed. For many practical applications, sometimes it is necessary to study the electromagnetic behavior on a specific structure over a broad frequency band, and therefore it is important to have some benchmark data on computational resources needed for some commonly used numerical techniques. During this workshop, various antenna design applications will be examined. Also a full-size air, ground and sea platforms are considered and the frequency limit is pushed at different bands using several numerical techniques. A benchmark study of computational resources required by available computational tools for electrically large platforms will be presented.</p> Mon, 05 Jul 2010 13:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=58http://www.doe.carleton.ca/seminars/show_one.php?id=57<p>Currently emerging biosensing and intracortical stimulation devices are promising alternative to allow studying the neural activity underlying cognitive functions and pathologies, understanding neurons interactions, locating seizures, detecting mind driven decisions, and achieve efficient treatment for these central neural systems. This talk covers low-power analog circuits and packaging techniques used for the design and integration of biosensing Microsystems. Such devices are interconnected to intracortical neural tissues, and include high-reliability wireless links used to power up such implanted devices and bidirectionally exchange data with external base station. Global view of typical devices altogether with corresponding multidimensional challenges will be described. Special attention will be paid to report two case studies: 1) Automatic detection of action potentials from massively parallel channels, and 2) Epilepsy seizures monitoring and onset treatment. </p> Wed, 19 May 2010 11:00:33 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=57http://www.doe.carleton.ca/seminars/show_one.php?id=56<p>Ultra Wideband (UWB) technology has been gaining attention from both industry and academia since the Federal Communications Commission (FCC) authorized the unlicensed use of its 0-900MHz or 3.1-10.6GHz spectrum. A UWB system is defined as any system whose signals have a -10dB fractional bandwidth at least 20% of its center frequency, or a -10dB absolute bandwidth greater than 500MHz. Given the large bandwidth of UWB systems, in order to avoid interference to other wireless systems, the FCC has set an extremely low power spectral density mask for UWB systems, the maximum equivalent isotopic radiated power (EIRP) spectral density of which is -41.3dBm/MHz. In this presentation, we will provide an introduction to the UWB impulse radio systems, their developments and the recent emerging code-shifted-reference transceiver technology.</p> Thu, 25 Feb 2010 12:00:23 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=56http://www.doe.carleton.ca/seminars/show_one.php?id=55<p>This talk will give and overview on the design, fabrication and characterization of optical devices. It will provide research examples in the field of optical devices and nano-fabrication,and discuss current work in this area at Carleton University,MIT and Harvard University.</p> Thu, 19 Nov 2009 11:30:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=55http://www.doe.carleton.ca/seminars/show_one.php?id=54<p> Asynchronous-logic is an emerging methodology where the microprocessor will be increasingly asynchronous – 40% by 2020 from the present 11% (ITRS roadmap). It is an alternative design approach (as opposed to prevalent synchronous-logic) to ultra low power digital circuits and it could be robust to the process, voltage and temperature (PVT) variations for biomedical applications and the like. In this talk, an overview from the following perspectives will first be given: some pertinent requirements of biomedical applications, the current-art technology roadmap and its technology challenges, and a review of asynchronous-logic and synchronous-logic. Thereafter, the completed and on-going asynchronous-logic projects will be presented, in part, for the design of low power biomedical applications. These projects include a Fast Fourier Transform processor, an Intel-based 8051 microcontroller, a Motorola-based 24-digital signal processor, and an asynchronous electronic design automation tool. Finally, some potential projects by adopting asynchronous-logic will be discussed. </p> Tue, 29 Sep 2009 12:30:02 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=54http://www.doe.carleton.ca/seminars/show_one.php?id=53<p>Smart grids and sustainable energy will be powering our economy in the future, and will help meet energy challenges such as climate change and fossil fuel crisis. Real time power system simulation is a core smart grid technology that provides unprecedented observability & controllability in power systems and thus can greatly improve the system security, reliability and efficiency.</p> <p>Electromagnetic transient programs (EMTP) can most closely simulate the real power system dynamics by continuously tracing the evolution of the system states in arbitrary multi-phase networks with lumped or distributed parameters. Therefore EMTP-type simulators are very appealing to be used for real-time dynamic assessment. However, their capability in real time simulation of power systems is compromised due to the small time step used and relatively slow simulation speed.</p> <p>A Shifted Frequency Analysis (SFA) method has been proposed by the speaker to accelerate EMTP solutions for real-time simulation of power system dynamics. The SFA allows the use of large time steps in the EMTP solution environment to accurately simulate power system dynamics within a band centered around a fundamental frequency. The following topics will be discussed:</p> <ul> <li> How to transform the original system into a shifted frequency system,</li> <li> How to couple SFA models into an EMTP-type simulator, </li> <li>The development of a SFA-based EMTP simulation tool, and </li> <li>Future research on applications in smart grid, renewable energy, and associated hardware and software innovations </li> </ul> <p>The SFA provides dynamic phasor results, which help power engineers gain better insight into the system dynamics. The proposed SFA method is the first practical step towards unifying power system slow and fast dynamic analysis methods using the EMTP.</p> Thu, 10 Sep 2009 09:30:33 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=53http://www.doe.carleton.ca/seminars/show_one.php?id=52<p>Adequate provision of reactive power is essential in power systems in order to ensure their secure and reliable operation. Reactive power is tightly related to bus voltages throughout a power network, and hence reactive power services have a significant effect on system security. Insufficient reactive power supply can result in voltage collapse, which has been one of the main reasons for some major blackouts worldwide. In the erstwhile vertically integrated power system structure, provision of reactive power by electric utilities was embedded within the electricity supply to customers. However, in the deregulated power system structure, reactive power is managed and charged for separately as an ancillary service.</p> <p>Currently, most power system operators procure reactive power services from available providers based on operational experience and expected voltage problems in the system. In real-time, most system operators use power flow programs to dispatch reactive power from the already contracted generators. However, there are several issues and concerns associated with the current procurement practices and pricing policies for reactive power which call for further systematic procedures to arrive at more efficient service management and sufficient reactive power support for a more reliable power system. Some of these issues are technical limitations associated with power system operation, whereas others are policy issues related to the rules under which the electricity market operates in a certain jurisdiction.</p> <p>An integrated framework for reactive power ancillary services management in competitive electricity markets is presented in this research work using a two-settlement model approach. The proposed scheme works at two hierarchical levels and in different time horizons- the first level is the reactive power procurement model which works in a seasonal time horizon, and takes into consideration system security aspects in determining an optimal set of generators and zonal price components. This would form the basis for seasonal contracts between the system operator and the selected reactive power service providers. The second level is the reactive power dispatch model which works in a one-hour to 30-minute window, and takes into account both the technical and economical aspects of operation in a market-based environment. The developed framework addresses the main issues associated with reactive power ancillary service management post-deregulation, proposing appropriate policy solutions that suit the requirements of system operators in such a competitive market environment. The framework is generic in nature and designed to be adopted by system operators in any electricity market structure, be it a bilateral contract market or a pool market.</p> <p>Motivated by the above research work and the recent recommendations and incentives of the government of Canada to move towards clean sources of energy such as wind and solar generation, directions for future research are then presented. A potential research approach is to investigate the contribution of Distributed Generation (DG) resources to ancillary service provision, their optimal pricing in case of such ancillary service support, and grid connection agreements and technical requirements. This is particularly important for wind generators which have a significant potential as a renewable source with possibility for providing ancillary services. Other proposed research directions include: considering reactive power from capacitor banks and FACTS devices as ancillary services that are eligible for financial compensation, and developing price and demand forecast models for ancillary service providers to provide them with information about demand and price spikes.</p> Fri, 26 Jun 2009 13:30:05 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=52http://www.doe.carleton.ca/seminars/show_one.php?id=51<p>Introducing Distributed Generation (DG), traditional and renewable technologies, into distribution systems has both positive and negative impacts on system operation. One of the negative effects of implementing DGs is related to system protection, such as the disturbance caused to the existing relay coordination and possibility of islanding occurrence in looped and radial distribution systems respectively. In this talk, the solutions for restoring the original relay coordination and preventing islanding formation are explored for distribution-systems equipped with traditional/wind-turbine DGs.n  n In looped distribution systems, a directional over-current relay mis-coordination problem arises from installing DGs. A new approach involves the implementation of a fault current limiter (FCL) to locally limit the DG fault current, and thus restore the original relay coordination. The proposed restoration approach is carried out without altering the original relay settings or disconnecting DGs during faults to ensure reliable power continuity. Therefore, it is applicable to both the current practice of disconnecting DGs, and the emergent trend of keeping DGs running during fault. The process of selecting FCL impedance type (inductive or resistive) and its minimum value will be illustrated by examining the implementation of DGs with and without FCL. Various simulations are carried out for both single- and multi-DG operation, and different DG and fault locations. The obtained results will be discussed.nn On the other hand, in radial distribution systems, the major protection issue is the possibility of islanding formation following a disturbance event and loss of the main grid. Therefore, local distribution companies require disconnection of DGs as soon as disturbance events occur to prevent sustainable islanding and the damage of DGs and other equipments to protect public safety. Thus, real-time laboratory testings were carried out to evaluate existing anti-islanding detection schemes to assess the selection of a fast and reliable scheme for a practical radial distribution system equipped with a group of wind-turbine (WT-DGs). Two anti-islanding schemes (passive local detection and communication-based) were examined on various commercial relays to disconnect WT-DGs following a disturbance event and loss of the main grid. A dedicated fiber optic connection and a radio-frequency communication-based scheme were evaluated and compared with passive local detection scheme performance. Various WT-DG penetration levels were simulated and different available commercial relay functions were tested. The obtained results, recommendations, and economic evaluations for this choice will be discussed in this talk.</p> Tue, 16 Jun 2009 13:30:36 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=51http://www.doe.carleton.ca/seminars/show_one.php?id=50<p>This presentation provides an overview of current issues and challenges involved in integrating renewable energy sources with the electric power network while introducing a future perspective on improving the power quality and the accuracy of the metering systems towards the smart grid. Research directions on power quality improvement and future research phases leading to economic operation of the power grid at high power quality level will be presented. Other topics that are relevant to my experience such as teaching philosophy and online teaching of power engineering courses will be highlighted.</p> Mon, 08 Jun 2009 15:30:39 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=50http://www.doe.carleton.ca/seminars/show_one.php?id=49<p>This talk will provide an overview of WiSense, a research and development project that partners industry and university research groups with the mission of defining and studying design challenges, scalability and compatibility issues for the development of wireless heterogeneous sensor networks.</p> <p>The WiSense project targets wide-spread applications, such as biomedicine, home security, intelligent vehicular systems, and emergency preparedness and response. This talk will discuss the technologies that are deployed in advanced wireless sensor networks, the challenges of designing scalable and cost-effective networks, and the importance of collaborations between academic and industrial partners.</p> Wed, 13 May 2009 12:00:35 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=49http://www.doe.carleton.ca/seminars/show_one.php?id=48<p><strong>Subtitle 1: Large-Scale Electromagnetic Systems Solving Techniques</strong></p> <h3>Subtitle 2: A 3-D High performance EM Simulation Tool: GEMS software and System</h3> <p>Given the recent trend whereby the designer of an EM device or a complex system thinks that it is imperative to carry out a “sanity check” before entering into the fabrication stage, a reliable and general-purpose EM solver has become highly-desired as well as invaluable part of the engineer’s toolkit. However, it is not uncommon for the designer to run into a roadblock when attempting to simulate many real-world problems, due primarily to the inability of the available EM software modules to handle problems that require large number of degrees of freedom (DoFs), which can often run into billions. GEMS, a unique computer code which is recent entry into the EM software scenario, was originally developed under the sponsorship of the U.S. Navy, who needed to solve large and complex shipboard antenna problems, such as phased arrays with thousands of elements sharing a common platform, e.g., the mast of a ship, but found that the existing commercial softwares were highly inadequate in their ability to solve large problems that required a large number of DoFs to model them.</p> <p>GEMS is a based on the parallel FDTD algorithm, and it has been designed to efficiently simulate large and complex EM problems using a group of distributed computers. Another desirable and unique feature of GEMS is that it provides a combination of EM software and hardware including network, operating system, I/O and storage, and optimized processors and memory. The talk will describe some of the features of GEMS, illustrated by numerous real-world examples that involve the simulation of<span>  </span>large and complex EM problems,<span>  </span>including antennas; antenna arrays; electronic packaging; RF and microwave circuits; bio-electromagnetics; and EMC/EMI, to name just a few.</p> Tue, 05 May 2009 11:00:49 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=48http://www.doe.carleton.ca/seminars/show_one.php?id=47<p>Impedance matching is an essential part of an RF system that is used to maximize the power transfer from a generator to a load and to minimize the reflections from the load. When the  load impedance is fixed the impedance matching is not a difficult task, especially for narrowband systems. However in wireless communication systems variations in antenna input impedance can occur. In a portable cellular handset, for instance, this impedance will undergo significant deviation from interaction with the user.  This varying impedance can lead to unacceptable degradation in performance of the power amplifier, affecting its output power, efficiency and phase characteristics. Indeed, the load mismatch conditions may result in the presence of over voltage in the output amplifier and may eventually lead to permanent failure of the output transistor due to avalanche breakdown. The overvoltage problem can be solved through different approaches, such as  the utilization of an isolator in the output of amplifier or an active feedback, functioning as a voltage limiter.</p> <p>These approaches protect the amplifier, but do not avoid the reduction of output power due to the load impedance mismatching. An automatic impedance matching system may be a solution  either to avoid overvoltage or to prevent the reduction of output power. This talks describes the constituent blocks of an automatic impedance system , especially a systematic design approach of tunable impedance matching networks. Some practical realizations will be also be presented and discussed.</p> <p> </p> <pre>n</pre> Thu, 19 Mar 2009 13:00:55 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=47http://www.doe.carleton.ca/seminars/show_one.php?id=44<p>The next wave in the information revolution will consist of bringing intelligence to the information and communication technology (ICT) sector, allowing seamless and intelligent networking and communication between different users using different services and operators. This will lead to the convergence of communication technologies, aiming at the development and deployment of cooperative and ubiquitous networks that involve existing and future wireless and satellite communications systems.</p> <p>A critical element in enabling the convergence of different communication systems is the development of software defined radio (SDR) systems that can be used across different frequency bands and for multi-standard applications. This SDR has to be developed to support different frequency carriers and modulations schemes concurrently, in addition to being power- and spectrum-efficient, in order to be able handle high data rates, while being less energy-hungry and more environmentally friendly.</p> <p>The design of power amplifiers as critical components in any SRD based communication terminal has to be considered closely together with the system architecture, in order to ensure optimal system level performances in terms of linearity and power efficiency. This implies the use of adequate transmitter architectures that convert the analog baseband information to architecture dependent amplifier driving signals, such as sigma-delta, EE&R, Polar and LINC architectures. This talk lays out the principles behind SDR systems and examines the design of software-enabled linear and highly efficient RF/DSP co-designed power amplifiers/transmitters for multi-standard and multi-band applications. Recent advances and practical realizations will also be presented and discussed.</p> Fri, 06 Mar 2009 15:30:57 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=44http://www.doe.carleton.ca/seminars/show_one.php?id=46<p><strong></strong><strong></strong>The design of advanced components for space and terrestrial telecommunications requires both sophisticated design methodologies and manufacturing technologies.  The challenge is to design integrated high performance components while using low cost manufacturing processes.  This presentation reports on technologies and design methodologies intended to surpass conventional approaches. Stereo lithography and micro stereo lithography are rapid prototyping processes based on a space resolved laser polymerization.  Ceramic suspensions can be used together with these processes for prototyping complex microwave components. A thin layer of ceramic suspension is repetitively deposited, hardened in the defined cross sectional pattern and bonded to the previous layer by laser polymerization.  Original microwave components (waveguides, resonators, filters… operating at a few GHz to 180 GHz) have been fabricated using these processes.  On the other hand, shape optimization can lead to new components.  Several approaches have been developed in the context of computer aided design (CAD), and some of them can be adapted to electromagnetic inverse problems.  For instance, the topology gradient approach is applicable in 3 dimensions for optimizing the distribution of dielectric (ceramic) material.  Several components designed and manufactured using these techniques have been tested, providing original solutions for the development of microwave components.<strong>n </strong></p> Wed, 04 Mar 2009 13:30:58 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=46http://www.doe.carleton.ca/seminars/show_one.php?id=45<p>Thanks to rapid advances in computing hardware, computer simulations have become a viable alternative for real-life controlled physical experiments. Unfortunately the computational cost of computer simulations remains a barrier, one simulation may take many minutes, hours, or days. As a result visualization, optimization, and design space exploration or the simulation response quickly becomes prohibitively expensive. Consequently, surrogate modeling methods (= response surface models, metamodels) such as Neural Networks, Kriging models, Support Vector Machines, rational functions, etc. have become a standard technique for tackling this problem. This talk will discuss the problems involved in surrogate modeling and present a fully automatic approach (and associated software implementation) to surrogate model generation, applicable to a wide range of domains. In concrete terms the goal is to provide an automated answer to the following question from a domain expert: "I need an approximation model for my problem with 5% accuracy, using a minimal number of computer simulations".</p> Fri, 27 Feb 2009 13:00:44 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=45http://www.doe.carleton.ca/seminars/show_one.php?id=43<p>The talk will provide a short overview of the evolution of the fiber optics communications in the last decade. The emphasis will be on the DWDM (dense wavelength division multiplexing) and other key technology enablers. The author will also analyze some of the major research, engineering and business challenges and trade offs. The final part of the presentation will focus on the most recent trends in the DWDM-based NGN (new generation networks) research and development.</p> Thu, 22 Jan 2009 13:00:06 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=43http://www.doe.carleton.ca/seminars/show_one.php?id=9<p>Radon is a naturally occurring radioactive gas produced by the decay of uranium in the earth’s crust. Radon may become trapped in buildings, posing a serious health hazard. Exposure of lung tissue to the alpha particles produced in the decay of radon may cause lung cancer. It is estimated that the mortality due to radon exposure across North America is comparable to that due to traffic accidents. In some regions (including the Ottawa area) mortality from radon can be much higher. There is therefore a strong need for an inexpensive radon monitor that can be widely deployed. This talk describes the development of a simple MOS IC to detect the alpha particles produced in radon decay, and the incorporation of this chip with an electrostatic radon progeny concentrator to form a complete working radon monitor. The new monitor is direct-reading and capable of detecting hazardous levels of radon in approximately one day. This represents a significantn improvement over competing consumer-level techniques for radon detection. It should be possible to produce the monitor at low cost in high volume.n</p> Fri, 28 Nov 2008 01:00:28 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=9http://www.doe.carleton.ca/seminars/show_one.php?id=10<p><strong></strong>As the semiconductor industry moves beyond the 45nm node and as systems become more heterogeneous, System on Chip (SoC) solutions are facing major barriers due to technical and business related reasons. This is leading to the development of new technological solutions such as System on Package (SoP). SoP, a technology being pioneered by Georgia Tech, allows for integration of functions in the package. Higher levels of integration are possible by embedding functions in the substrate and merging the package and board level technologies into one. SoP enables the integration of digital, RF, opto-electronic and sensor electronics in the package leading to system miniaturization with micro, nano and bio convergence. This talk will focus on some of the technologies and CAD methods for SoP being developed at Georgia Tech.</p> Mon, 10 Nov 2008 17:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=10http://www.doe.carleton.ca/seminars/show_one.php?id=11<p>The past thirty years have been an exciting period for tremendous advances in the development of interior permanent magnet (IPM) technology. During this time, IPM machines have expanded their presence in the commercial marketplace from a few specialized niche markets such as machine tool servo drives to mass-produced applications including high-efficiency electric traction drives for the latest generation of hybrid-electric vehicles (HEV). Power ratings of available IPM motor drives have dramatically expanded by approximately three orders of magnitude during this period, now reaching power levels up to 1 MW. What are the factors that made such impressive progress possible? Closer examination reveals that several different knowledge-based technological advancements and market forces have combined, sometimes in fortuitous ways, to accelerate the development of the impressive IPM technology that we find available today. The purpose of this talk is to provide a broad explanation of the various factors that lead to our current state-of-the-art IPM technology in HEV. This highly efficient energy conversion technology has enormous impacts on the world electrical energy supply and demand utilizing conventional fossil fuel sources. Examples will illustrate commercial successes such as Toyota's PRIUS hybrid electric vehicles. The talk will highlight the developments to expect in electronics-based, highly-efficient and green automobiles now and in the near future.</p> Fri, 07 Nov 2008 11:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=11http://www.doe.carleton.ca/seminars/show_one.php?id=13<p>In this talk, an improved drain-source current (I-V) model forn <span class="SpellE">HEMTs</span> is proposed.  The model is simple, easy to extract, and convenientn for implementation in simulation tools.A single modeling equation is developed, allowing accurate prediction of both static and dynamic I-V characteristics. The model parameters can be extracted to match the measured data closely for a wide bias range withoutn sacrificig accuracy. It is validated through DC as well as power measurements using <span class="SpellE">GaAs</span> HEMT transistors. Also included in the presentation, will be a brief overview of the new IEEE Section recently created in Chengdu, China. Chengdu is becoming one of the fastest growing hubs for electronic research and development in China. As a founding member of the IEEE Section and the founding Chair of the Chengdu Chapter of IEEE Electronic Device Society, Professor Ma will give an <span class="GramE">insider ’s</span> perspective of the new IEEE Chengdu Section, and the electronic research and education in the University of Electronic Sciences and Technology of China.</p> Tue, 07 Oct 2008 12:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=13http://www.doe.carleton.ca/seminars/show_one.php?id=12<div>For over 60 years, artificial man--made satellites have been providing diverse, highly available services, worldwide. The Sun is the lifeline of majority of satellite space segments, providing to satellites a thermal equilibrium, and, via solar cells, the electric energy. When the Sun becomes obscured by the Earth or by the Moon, a solar eclipse occurs. A satellite's lifeline becomes vitally reduced or cut and its thermal equilibrium disrupted. Different measures have to be taken to reduce and/or avoid potential degradations and/or disruptions of services. The worst case scenario, an unavailability of service, is also called an outage.nn Direct exposure to the Sun by a receiver's antenna main beam would cause an increase in the receiver's system noise temperature, which, consequentially, may cause a degradation of service and even an outage.nn We focus on satellites in the GeoStationary Orbit (GSO). A simplified link budget calculation for a hypothetical GSO satellite is presented, and a component in this budget, which is impacted by the Sun, is pin-pointed. Geometrical aspects of satellite earth station antennas, satellite space segments, the Sun, the Earth, the Moon and related natural phenomena, including equinoxes in particular, are described and illustrated with a number of examples. Three major alinement scenarios are analyzed and illustrated: Earth Station-GSO-Sun alinement also known as the Sun transit of the antenna main beam, GSO-Earth-Sun alinement and GSO-Moon-Sun alinement. The last two scenarios are solar eclipses.nn The Moon caused solar eclipses (partial, total, annular) are the least frequent of the three; the total eclipse could last minutes, an annular eclipse lasts approximately 10 minutes and a partial one still longer. The Earth caused solar eclipses (partial, total) occur regularly around equinoxes; at equinoxes, they last approximately up to 72 minutes around midnight of the local time, and repeat every day with shorter and shorter durations for approximately up to +/-23 days. The timing of the Sun transit of the antenna main beam depends on the geographic location of a particular earth station antenna. It occurs on equinoxes for stations at equator and shifts away from equinoxes for stations at higher latitudes. The exposure lasts for approximately 10 minutes for an average size antenna at its peak worst time instant. It is gradually reduced within a number of days. Higher gain antennas are exposed over shorter periods of time but with higher intensity, and vice versa.n</div> Mon, 22 Sep 2008 19:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=12http://www.doe.carleton.ca/seminars/show_one.php?id=14<p>The anticipated presentation will cover the current status and future trends of millimeter-wave MMICs, including those using III-V compound (GaAs, InP, GaN, etc.) and Si-based (CMOS, SiGe HBT and BiCMOS) MMIC technologies. Millimeter-wave MMICs used to be applied to military and astronomy systems for long time and started to be utilized for civil applications in the decade, such as communications and automotive radars. The evolution of IC technologies has enabled the performance of Si-based MMICs over 100 GHz, even in standard bulk CMOS processes. This is believed to have a major impact in the future development of millimeter-wave systems. Since low-cost mass-production potential pushes forward the technology, a very high integration of circuit functions on a chip, such as RF, base-band circuitry, automatic-control for a steady operation, and maybe even the antenna, etc. should be included, and thus the system on chip (SOC) issues should be addressed, especially in MMW regime. Moreover, millimeter-wave packaging cost always dominated in the module development. In order to simplify the assembly and reduced cost, the concept of system in package (SIP) has been proposed. This presentation will also survey the current technologies for SOC and SIP and discuss related issues and challenges.</p> Mon, 23 Jun 2008 16:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=14http://www.doe.carleton.ca/seminars/show_one.php?id=15<p>Modern microwave systems are designed in a top-down hierarchical process, with specifications starting at the system level, propagating down towards the subsystem, module, integrated circuit, and finally to the level of the transistor, resistor, and other fundamental electronic building blocks. A complimentary bottom-up process combines accurate representations of the building blocks at one level of abstraction to create or verify a functional block at the next higher level of design complexity. At a low level in the design hierarchy is the nonlinear device, or transistor. A detailed model, involving the simulation of the many coupled partial differential equations of physics is often needed to design such a device. However, one cannot simulate an entire IC at this physically detailed level. The complexity of the problem is overwhelming in terms of computer resources and time. Instead, for the purpose of integrated circuit design, transistor terminal (behavioral) characteristics can be abstracted into compact nonlinear models (SPICE models) and their interaction simulated at the circuit level. Analogously, modern communication systems are sufficiently complex to preclude their complete simulation at the compact transistor model level of description. There are simply too many nonlinear equations to solve to make this practical. Instead, the input-output behavior of the ICs can be abstracted into functional block behavioral models, and the simulations done at the next higher abstraction level.</p> <p>This lecture introduces general concepts and specific techniques for effective (efficient, general, and accurate) nonlinear behavioral modeling of microwave semiconductor devices and functional circuit blocks. A behavioral model is a simplified but accurate model of a lower-level component in the design hierarchy that simulates efficiently at the next higher level of abstraction. A unified treatment at both the device and functional block level is a distinguishing feature of this presentation. So too is the application to behavioral models constructed from real measurements and also from simulations starting from a detailed (complex) model. The emphasis is placed on the combination of nonlinearity and dynamics. Nonlinearity includes harmonic and inter-modulation distortion, clipping, etc. Dynamics includes frequency-dependence and long-term memory effects from a variety of physical origins. In the realm of dynamic nonlinearities, insight from linear analysis cannot always be applied. Superposition is not generally valid, the Fourier domain is less useful, and Green functions don?t exist. No fully general or overarching theories of nonlinear dynamical systems exist that are comparable to what exists for linear systems. Nevertheless, great progress has been made recently in nonlinear behavioral modeling. In fact, this lecture suggests we are at the threshold for full interoperability of large-signal measurement systems, modeling approaches, and simulation algorithms for nonlinear hierarchical behavioral modeling. This means we can begin to do for driven nonlinear microwave systems what small-signal S-parameters enable for linear systems.</p> Mon, 31 Mar 2008 12:00:56 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=15http://www.doe.carleton.ca/seminars/show_one.php?id=16<p>Quantum processing of information requires the development of quantum systems which are at the same time coherent and quantum in nature, and yet easily manipulated to process and extract classical information. To meet this challenge we have embarked on the development of technologies which would allow us to design and build nano-scale scalable and coherent solid state systems using elementary building blocks such as single electron spins, single excitons, and single photons using semiconductor quantum dots. We show how gated quantum dots allow to localize individual electrons, control their spin properties by their number, form of confinement, and the magnetic field, enabling nanospintronics. The spin can be probed and exploited by connecting quantum dots to spin polarized reservoirs. The resulting spectroscopic tool, the spin blockade spectroscopy, will be described as well as a prototype nano-spintronic device, the “single spin transistor”. By combining the single spin transistors into coherently coupled devices one is attempting to build an electron spin-based quantum computer. I will describe double and triple quantum dots and extrapolate to the exciting physics such new capabilities enable. In order to combine the control over spin with the control over photons we need to confine both electrons and valence holes. This is done by transferring the gated technology to self-assembled quantum dots. I will review progress in our understanding of the electronic and optical properties of InAs-based self-assembled quantum dots emitting at 1.5micron. By combining lithography with self-assembly single InAs dots can be positioned on InP nanotemplates. This control allows integration of quantum dots with photonic cavities and opens up possibility of “manufacturing” a single photon gun for quantum cryptography and communication. Finally, building on the newly acquired capabilities with quantum dots we will venture into combining information processing and storage using quantum dots containing both electrons and magnetic ions, a step toward control of magnetism on nanoscale. *in collaboration with A.Sachrajda, M.Korkusinski,and R.Williams.</p> Tue, 04 Mar 2008 13:30:24 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=16http://www.doe.carleton.ca/seminars/show_one.php?id=17<p>As process technologies and supply voltages shrink, designers are faced with a pressing need to address systematic and random sources of variation in a more deliberate and thorough way. Accounting for variation within the flow of design has not progressed commensurate with the process technologies. We still rely on best-, worst- case corners, mismatch plots and maybe a Monte Carlo verification if there is enough time. It is time for a new approach. This talk will begin with a brief review of the physical phenomena and industry standard device models for variation sources, including random local and global variations and systematic proximity effects. New techniques to accelerate, increase accuracy and derive more information from statistical variation analysis will be presented.</p> Tue, 12 Feb 2008 13:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=17http://www.doe.carleton.ca/seminars/show_one.php?id=18<p>There is a considerable push to place silicon as the technology of choice for the implementation of millimetre-wave integrated circuits. This presentation outlines the design challenges present in silicon at millimetre-wave frequencies and compares the two major system solution approaches: System-On-Chip (SOC) vs. System-in-Package (SiP). While SOC solutions integrate all circuit sections into one IC reducing power, area and cost of fabrication, a System-in-Package solution provides the most efficient use for each technology: CMOS for digital and analog control circuit sections, SiGe and III-V for front-end circuits and packaging technologies that can absorb passives and antenna elements. A System-in-Package implementation for millimeter-wave circuits at 60 GHz and beyond will be discussed using Low-Temperature Co-fired Ceramics (LTCC) and Integrated Packaging Devices (IPD) package technologies. This presentation will emphasize the use of III-V and silicon technologies. An introduction to package technologies suitable for millimeter-wave integration will also be presented. The design, implementation and testing efforts for a 60 GHz radio transceiver in silicon will be discussed.</p> <p> </p> Mon, 21 Jan 2008 11:00:11 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=18http://www.doe.carleton.ca/seminars/show_one.php?id=19<p>Conventional active electronic devices are based on crystalline semiconducting materials, predominately silicon. Although this dominance of silicon continues with its limits being continually expanded, other technologies on the horizon have significant promise for complementary functions and applications unsuitable to silicon. Active soft materials such as conducting polymers, organic/inorganic composites and biological materials have interesting properties that can be used to provide unique functionality. These materials also have the advantage of being compatible with flexible substrates and simple printing techniques for the fabrication of devices. In this talk, two areas of interest are presented to highlight the functionality possible. Current work using conducting polymers to fabricate complex interacting systems is used to illustrate organo-electronic technology and recent research into the application of the protein bacteriorhodopsin (bR) to optical switching provides a demonstration of organo-photonic technology.</p> <p> </p> Fri, 11 Jan 2008 11:00:52 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=19http://www.doe.carleton.ca/seminars/show_one.php?id=25<p>Energy efficient system design requires systematic optimization at all levels of the design abstraction ranging from devices and circuits to architectures and algorithms. The design of micro-power systems will enable operation using energy scavenging. A major opportunity to reduce the power dissipation of digital circuits is to scale the power supply voltage below the device thresholds (i.e., sub-threshold operation). The opportunities and challenges associated with sub-threshold design will be presented. This includes variation-aware design for logic and SRAM circuits, efficient DC-DC converters for ultra-low-voltage delivery, and algorithm structuring to support extreme parallelism. A number of integrated circuit examples that demonstrate sub-threshold operation will be presented. Other power management techniques such as ultra-dynamic-voltage scaling, fine-grained power gating and 3-D integration will be discussed. The use of highly digital architectures for wireless communication circuits can also significantly reduce system energy dissipation. Specific examples of power management will be presented, focusing on wireless sensor networks and impulse based ultra-wideband communications as drivers.</p> Thu, 20 Dec 2007 14:00:39 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=25http://www.doe.carleton.ca/seminars/show_one.php?id=26<p>Electrostatic discharge (ESD) is a process in which a finite amount ofncharge is transferred from one object (i.e., human body) to the other (i.e.,nmicrochip). This process can result in a very high current passing throughnthe microchip within a very short period of time, and more than 35% of chipndamages can be attributed to such an event. As such, designing robustnon-chip ESD structures to protect microchips against ESD stress is a highnpriority in the semiconductor industry. An overview on the ESD sources,nmodels, and protection schemes will first be given in this talk. This isnfollowed by the examples of robust ESD solutions for protecting datancommunication transceivers, high voltage IC, gas-sensor microchips, and lownvoltage RF IC.</p> Mon, 19 Nov 2007 12:30:58 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=26http://www.doe.carleton.ca/seminars/show_one.php?id=27<p>1. Minimizing Power Amplifier Memory Effects</p> <p>Abstract: Memory effects are changes in an amplifiers non-linear characteristics resulting from the past history of the input signal. Predistortion linearization depends on a stable non-linear response, and can be significantly degraded by memory effects. This presentation will begin with an overview of linearization techniques and the application of digital signal processing (DSP) to the correction of distortion in power amplifiers. The problems caused by memory effects will be introduced, different sources of memory effects will be discussed and techniques for their suppression presented.</p> <p>2. Advances in Millimeter-wave Linearization</p> <p>It is now more than ten years since the introduction of the first millimeter-wave (MMW) linearizers. During this time the capabilities of MMW linearizers have advanced tremendously. Today's MMW linearizers operate at higher frequency (> 45 GHz), provide wider bandwidth (> 10 GHz), and give greater performance (> 20 dB NPR at 4 dB OPBO). They also operate with a wider variety of MMW power devices (TWTAs ­ helix and coupled cavity, Klystrons and EIKs, MMPMs and SSPAs). More than 1 kW of MMW linear power can be produced today with high efficiency in a relatively small package. This paper discusses the state-of-the-art in MMW linearization and provides a road map of what to expect in the future.</p> <p> </p> <p> </p> Mon, 22 Oct 2007 13:00:19 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=27http://www.doe.carleton.ca/seminars/show_one.php?id=28<p>This talk presents management of waste in urban municipalities through RFID and GPRS-UMTS technologies.nnThe waste collection for municipalities in Western Europe can be described with some facts:nn-           the costs of waste treatment (collection, treatment, etc) are explodingnn-           waste is getting a luxury articlenn-           technological developments will lead to efficiency improvements in the waste collectionnn-           municipalities ask from the waste collectors for cost transparencynn-           there is a general demand from the citizens for systems which facilitate fair invoicing systemsnnThese aspects have lead to the development and the implementation of identification and weighing systems, with which the above aspects can be controlled or realized. For the identification of the waste collection bins transponders are used based on RF-ID technology. The weighing of the bins is performed by a dynamic weighing system which use the vibrating string technology. Future developments are GPRS-UMTS data communication from the truck to a host system.n nn</p> Tue, 11 Sep 2007 13:00:32 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=28http://www.doe.carleton.ca/seminars/show_one.php?id=31<p>As operating frequency increases and device sizes shrink, the complexity of current state-of-the-art designs has increased dramatically. One of the main contributors to this complexity is high speed interconnects. At high frequencies, interconnects become dominant contributors to signal degradation, and their effects such as delays, reflections, and crosstalk must be accurately simulated. Time domain analysis of such structures is however very difficult because, at high frequencies, they must be modeled as distributed transmission lines which, after discretization, result in very large networks. In order to improve the simulation efficiency of such structures, model order reduction has been proposed in the literature. Conventional model order reduction methods based on Krylov subspace have a number of limitations in many practical simulation problems. This restricts their usefulness in general commercial simulators.nn The focus of this presentation is the development of new reduction techniques to address the key shortcomings of current model order reduction methods. Firstly, a new approach for handling macromodels with a large number of ports will be presented.  This approach is based on taking advantage of prior information about the loads. Specifically, resistive, capacitive, transmission line loads as well as nonlinear loads such as diodes and inverters are considered. Secondly, a multi-level reduction based on singular value decomposition will also be presented for regular as well as parametric macromodels. The macromodel obtained using the presented approach is typically about one third the size of the macromodel obtained using the traditional parametric model order reduction method. In addition, a new parametric formulation will be discussed to allow the sparsification of parametric networks. Finally, a time domain reduction method will be presented for the macromodeling of nonlinear parametric systems. Using these approaches, CPU speedups of 1 to 2 orders of magnitude are obtained.</p> Fri, 27 Jul 2007 12:30:07 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=31http://www.doe.carleton.ca/seminars/show_one.php?id=30<p>  Though Radio Frequency Identification (RFID) systems have been in use for various applications in the past, currently they are gaining  popularity due to their application to retail supply chain management systems. Compared to low-frequency (LF) and high-frequency (HF) RFID systems (which operate through near-field inductive coupling and thus have relatively short read range), ultra high frequency  (UHF) RFID systems operate through farfield backscattering, have larger read range, and have been widely used in supply chain  management and inventory control.  However, very often the electromagnetic (EM) performance of the reader/tag systems could be  significantly degraded due to the complex physical environments.  With the aid of computational electromagnetic (CEM) tools, such  situations can be analyzed and optimized to improve the performance of RFID systems. This talk presents options for the EM  characterization of such systems with the aid of full wave or hybrid numerical methods. Analysis of RFID tags, readers, tag placement,  tag/reader coupling, and tag/reader systems in complex environments will be addressed.</p> Tue, 24 Jul 2007 12:20:20 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=30http://www.doe.carleton.ca/seminars/show_one.php?id=29<p>A new method is presented to calculate state space realizations of a three-dimensional (3D) image set. It is based on interpreting the image set as the impulse response of a 3D separable system. The proposed realization algorithm consists of two parts:nn1.) Decomposition of a 3D image set into the product of three 1D components;nn2.) Balanced state space realizations of finite 1D sequences.nnThe proposed method can be used for realizing the given 3D images exactly or approximately. The advantage of the method is noise can be reduced with little degradation on the image quality. It has been successfully applied in noise reduction of various 3D image sets of fluorescently labeled cells acquired by a fluorescent microscope. It can be used for noise reduction in a 3D image set or for noise suppression of a point spread function (PSF) which is an essential component in any 3D deconvolution algorithms.</p> Mon, 04 Jun 2007 11:00:17 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=29http://www.doe.carleton.ca/seminars/show_one.php?id=24<p class="MsoNormal"><span style="font-family: Times New Roman; font-size: small;"><span style="font-size: 12pt;"><span style="font-family: arial,helvetica,sans-serif;">The presentation will start with an overview of nanofabrication facilities available at the University of Alberta, including the Micromachining and Nanofabrication Facility (NanoFab – open to external users), and at the NRC National Institute for Nanotechnology.  Next the nanostructure fabrication work of the Brett group will be introduced, detailing the glancing angle deposition (GLAD) process capable of fabricating sub-micrometer arrays of posts, chevrons, helices, and square spirals.  Recent developments will be highlighted including indirect and direct fabrication of self organized organic microstructures and 3D periodically structured semiconductors in a square spiral architecture.   Applications to sensors, chiral optic devices, luminescent devices, and photonic crystals will be described. </span></span></span></p> <p class="MsoNormal"><span style="font-family: Times New Roman; font-size: small;"><span style="font-size: 12pt;"> </span></span></p> Mon, 19 Mar 2007 13:00:37 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=24http://www.doe.carleton.ca/seminars/show_one.php?id=32<p>Advances in the microchip industry and nanotechnology call for electromagnetic simulation of structures that are very complex, but are still a tiny fraction of a wavelength. The demand for electromagnetic simulation of antennas on cars, aircraft, and radar scattering, calls for electromagnetic simulation of structures that are many wavelengths long. This talk describes recent advances in solving Maxwell’s equations for complex structures which are a tiny fraction of a wavelength to ultra large structures involving hundreds of wavelength using integral equation methods derived from first principle electromagnetics. Modern electromagnetic simulations often require many degrees of freedom to describe the geometry. Hence, we will describe fast and efficient methods to solve large and dense matrix systems that follow from integral equations.</p> <p>A comparison between various kinds of numerical methods will be discussed. Then a brief overview of the fast algorithm, the multilevel fast multipole algorithm will be given. Extension of such fast algorithm to layered media and to very low frequencies will be presented. In order to capture quasi-static physics (circuit physics) and wave physics, a numerical solver has to work reliably in the very low frequency regime as well as the wave regime and the quasi-optical regime. Ways to overcome numerical instabilities associated with integral equations as well as acceleration techniques will be discussed. We will show large-scale simulation examples from scattering, subsurface probing, antennas mounted on cars, and complex structures as encountered in computer circuits and chips.</p> <p>Complex Coordinate Stretching for Absorbing Boundary Condition Berenger proposed the perfectly matched layer as an absorbing boundary condition for differential equation solvers. However, the phenomenon of absorption as proposed by Berenger can also be interpreted as complex coordinate stretching. This talk will describe the use of complex coordinate stretching in designing absorbing boundary condition for differential equation solvers for acoustic, electromagnetic and elastic wave simulations. Due to the simplicity of complex coordinate stretching concept, it can be easily extended from one set of partial differential equation to another set of partial differential equation. Hence, it can generalize the perfectly matched layer absorbing boundary conditions to cylindrical, spherical, and curvilinear coordinates quite easily. We will also discuss other applications of complex coordinate stretching.</p> <p> </p> <p> </p> Mon, 06 Nov 2006 10:00:41 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=32http://www.doe.carleton.ca/seminars/show_one.php?id=33<p>The decade of the nineties is highlighted with truly remarkable progress in our ability to carry out simulations, not only for large scale problems, but also in terms of hybridization and integration of passive and active RF circuits for a variety of applications. These developments have allowed for broadband antenna design, simulations of large multilayered and multifunctional antennas with embedded frequency selective surfaces (FSS), metamaterial substrate designs, MEMS analysis and design, large finite arrays and full scale aircraft scattering analysis using first principle methods, electromagnetic coupling and interference of systems involving passive and active components, magnetic resonance imaging (MRI) simulations, indoor propagation and evaluation of wireless systems, etc. What is probably so remarkable is that a decade ago (early 90s), we had just started looking at three-dimensional applications and the development of practical simulation tools was seemingly far away. Today, we have access to robust and fast three dimensional algorithms for modeling composite materials and have also demonstrated that simulations of practical vehicles or large finite antenna arrays and possibly RF integrated systems can be carried out on a desktop PC. In addition, we have delved into topology optimization/design. The latter holds promise for novel antenna and microwave circuit design, RF filters and RFICs for mixed signal applications, and others. This presentation will provide an overview of frequency domain developments, with particular focus on hybrid formulations and fast methods and their integration with formal design methodologies for antenna applications.</p> Tue, 24 Oct 2006 13:00:21 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=33http://www.doe.carleton.ca/seminars/show_one.php?id=23<p>We are developing a cell selection capability (from a cell library) that aims to enable the design of digital circuits that have the lowest possible power consumption for any desired speed goal, and whose leakage power is maintained below the specified limit. It is anticipated that this approach will yield circuits that have substantially lower total power and lower leakage for the same delays achieved today. The global optimization is performed over all the options available in the cell library (drive strengths, beta ratios, channel lengths, threshold voltages, etc.).</p> <p>The classical method of logical effort has the limitation that it only specifies how to achieve the fastest possible delay for a circuit. The method has no capability to minimize power for a delay goal other than the fastest possible. In addition, the classical method of logical effort does not distinguish between rising and falling delays for a path. Hence the method cannot optimize the beta ratio (pull-up strength versus, or divided by, the pull-down strength) of a cell. This is critical since many high performance cell libraries have (or should have) multiple beta ratios for each cell. The existence and effective use of multiple beta ratios can result in very significant power savings. Since rising/falling delays cannot be distinguished, it is also not possible to make effective use of multiple channel lengths and/or multiple threshold voltages. This is particularly serious due to the strong need to control static power today.</p> <p>In recent work we have formally extended the method of logical effort to distinctly handle rising and falling delays for a path. We have also extended the method of logical effort such that power can be minimized for a specific delay (less than or equal to the optimal delay). The cell selections (extracted from a library) utilized to implement a circuit are obtained by our new logical effort formulation by means of a robust Lagrangian relaxation technique. Now that we can optimally assign cell sizes, beta ratios, channel lengths, threshold voltages, etc., for both the rising and falling versions of a given path, we have further devised a novel scheme that optimizes the selection of all the cells in the entire circuit. The method appears efficient and should be able to handle circuit blocks of beyond a million cells.</p> <p>Our flow will also enable designers to target a specific power limit and then obtain the fastest possible circuit that satisfies the power limit. We also can determine what the optimal library content is. For example, our flow can determine which logic functions should be provided in the library. Similarly, we can determine which sizes and beta ratios (how many and which values) should be provided in the library. This is done by comparing the power vs. delay results for particular library content versus a continuous library assumption. We can also study which channel lengths and threshold voltages should be available in the library (also how many and which ones).</p> <p> </p> Mon, 02 Oct 2006 14:00:46 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=23http://www.doe.carleton.ca/seminars/show_one.php?id=34<p>  Retrodirective arrays are unique phased array systems that are capable of providing power gain in the direction of a received source with no prior knowledge of the source's location. When compared to fully adaptive arrays that adjust their beam patterns to the signal environment based on the received spatial signature, retrodirective arrays have small computational burdens for antenna pattern adaptation. Traditionally, retrodirective arrays have been designed around analog methods such as doubled local oscillator phase conjugation and heterodyning for use in line of sight non-fading channels. In this talk, we present a novel implementation of a retrodirective array for duplex digital communications and show that the retrodirective array can enhance system performance in a multipath environment. Measurements of the retrodirective array using flexible uplink and downlink modulation schemes are presented. They prove experimentally that the designed system is capable of tracking a signal source accurately under duplex communications conditions.</p> Fri, 15 Sep 2006 10:30:13 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=34http://www.doe.carleton.ca/seminars/show_one.php?id=35<p>Ferromagnets are important in numerous areas of electronics, including antennas and microwave devices, but their characteristic hysteretic behaviour is widely misunderstood. It has led to an extraordinary variety of different empirical models, most of which are only tenuously related to the underlying physics. In contrast, this talk introduces a simple theory that is based solely on quantum mechanics and classical physics. It provides a quantitative analytical description of numerous aspects of ferromagnetic phenomena. The theory introduces the concept of the domain-size function, which provides close links between the magnetization process at the quantum-dynamical scale, the behavior of the domains at the mesoscopic scale, and the measured characteristics of macroscopic samples. The theory is valid for materials ranging from extremely soft magnets to the hardest permanent magnets, with coercivities ranging over 6 orders of magnitude, and accounts for the relationships between the spatially-averaged domain size, the observed shapes of hysteresis loops, and the nature of the initial magnetization curve. Recent work indicates that the theory can also account for complex minor-loop behaviour, including first-order and second-order return curves (FORCs and SORCs). Examples of correlations between theory and measured data will be included.</p> <p> </p> Fri, 15 Sep 2006 09:30:17 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=35http://www.doe.carleton.ca/seminars/show_one.php?id=36<p>The LEST carries out research activities on physical component and signal processing for communication systems. These include theoretical modeling, low-cost technologies and materiel processes for applications to components at frequencies up to the millimeter range. Technologies are based on different processes: silk-screen and etching, membrane and foam. After a brief presentation of global activities carried out at the LEST, we shall focus on the various technologies and, in particular, on foam technology that offers some interesting potential up to the millimeter range. The foam (polyméthacrylic imid) can be shaped, machined and provides low-loss permittivity medium with a relative dielectric constant close to unity. Hence, it can contribute to increase antenna efficiency and provide low-cost solutions by using spray metallization. Recent work realized at the LEST with this technology will be presented.</p> Thu, 20 Jul 2006 13:00:36 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=36http://www.doe.carleton.ca/seminars/show_one.php?id=22<p>This talk describes the various university–industry linkage programs existing in the Indian higher education system, in the context of the technological revolution that’s sweeping the country. India has a very vast higher education system. There are over 330 Universities, 1400 professional institutions offering degree programmers in Engineering, Sciences, Humanities, Pharmacy, and many other disciplines. Knowledge has become a critical resource in the present global economy. Collaboration between industry and educational institutions is very important for the creation and dissemination of knowledge and thereby benefiting all stake holders.</p> <p>The talk covers the mode and scope of industry linkages among the premier institutions of India, including, IIT (Indian Institute of Technology), Delhi & Chennai, IISc.( Indian Institute of Sciences), Bangalore, BITS (Birla Institute of Technology & Science), Pilani, JNU (Jawahar Lal Nehru University), Delhi, Pune University, Hyderabad University, and Jadavpur University. It also highlights the emerging aspects of university-industry collaborations in India, in the context of the technological revolution that’s sweeping the country, with emphasis on electrical and information technology segments.</p> <p> </p> Wed, 28 Jun 2006 13:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=22http://www.doe.carleton.ca/seminars/show_one.php?id=21<p><strong></strong>There are many popular numerical techniques for electromagnetic waven modeling. These techniques can be divided into the frequency-domain andn time-domain methods. Since no one single method is suitable for all areas ofn application, a general purpose EM wave modeling software package shouldn include as many popular methods as possible. This seminar presents an objectn oriented modeling framework for building such a multiple-engine EM modelingn package. The framework captures the essence of various EM modelingn techniques; specialized methods can then be derived from the base model ofn the framework.<strong></strong></p> Thu, 11 May 2006 14:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=21http://www.doe.carleton.ca/seminars/show_one.php?id=20<p>The tremendous growth in broadband access services has propelled the development and deployment of photonics technology to a new frontier for which the efficient and cost-effective delivery of optical bandwidth to the end user (residential homes and the business premises) has become the focus of intensive research and development. In this respect, fiber-based solutions provide practically unlimited bandwidth. Fiber to the premises (FTTP) will enable not only converged voice, data, and video applications, but also a range of other value-added services. Among the various optical access technologies being developed and evaluated, passive optical networks (PONs based on different protocols appear to be the most promising architectures in terms of overall performance and cost-effectiveness. Several schemes and versions of the PON networks have been proposed, standardized and implemented to address the needs of different markets. To deploy a commercially viable PON system, low-cost and high-performance optical components such as transceivers are key elements.</p> <p>In this presentation, I will give a general review about the status and trend in market demand and technology status for the fiber-to-the-premises (FTTP) applications. In particular, the network requirements and the enabling technologies for the optical transceivers used for EPON and GPON will be discussed. Finally, some of the recent research and development activities on integrated FTTP transceivers at McMaster University (Canada) are reported.</p> <p> </p> Fri, 21 Oct 2005 15:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=20http://www.doe.carleton.ca/seminars/show_one.php?id=37<p>The solution of power electronic circuits has traditionally been accomplished by equation solvers such as Matlab or by circuit simulators such as PSpice and EMTP. The talk covers the fundamentals and trends in modeling and analysis of power electronic circuits. The talk also focuses on the related tools and methodologies, such as EMTP (including the latest version of EMTP RV (restructured version – with completely re-written code and with easy to use GUI). Examples on the use of EMTP RV from a simple two-switch DC-DC Chopper to a highly detailed High Voltage DC Transmission system are presented.n</p> Fri, 23 Sep 2005 13:00:41 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=37http://www.doe.carleton.ca/seminars/show_one.php?id=38<p>The enormous capability of the present and future advanced CMOS technologies encourages increasingly large and complex chips. On the other hand, today’s design strategies, methodologies, system architectures, and circuit techniques are not sufficiently effective to support the provided device capability. Particularly, the increasingly large power consumption, on-chip interconnect delay, and on/off chip communication bandwidth are among the major challenges. This presentation is intended to discuss several high-speed, low power circuit techniques, including: (i) A velocity of light limited on-chip interconnects design technique, utilizing transmission-line properties of higher metal layers, (ii) A multi-GHz skew-tolerant on-chip communication link with synchronizing ports (elastic FIFOs) that absorb all link latencies and clock skews, (iii) A low power, multi-GHz on-chip resonant clocking technique as an alternative to the mature but power hungry conventional clocking, and (vi) an embedded high density 128K cache memory utilizing a 5-Transistor single bit-line SRAM cell.</p> Fri, 29 Jul 2005 13:00:54 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=38http://www.doe.carleton.ca/seminars/show_one.php?id=39<p>As an important link between process technologies and circuit design, compact modelling has become more critical than ever as advanced process development enters into nano-scale and wireless era.  Novel physical effects and new electrical behavior appear in nano-scale devices. It is very challenging to describe these complicated physics in compact models for nano-scale devices while maintaining other modeling requirements such as scalability, continuity, accuracy and simplicity. Also, comprehensive modeling efforts are needed to make compact models suitable for RF applications where accurate and physical predictions in DC, AC small signal, noise, and linearity etc.  In this talk, we will discuss the modeling issues in nano-scale technology and overview further modeling challenges to develop compact models for RF applications.n n</p> Mon, 11 Jul 2005 13:00:18 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=39http://www.doe.carleton.ca/seminars/show_one.php?id=40<p>      An overview of circuit techniques dedicated to design reliable low-voltage (1-V and below) analog functions in deep submicron standard CMOS processes. The challenges of designing such low-voltage and reliable analog building blocks are addressed both at circuit and physical layout levels. State-of-the-art circuit topologies and techniques (input level shifting, bulk and current driven, DTMOS), used to build main analog modules (operational amplifier, analog CMOS switches) are covered with the implementation of MOS capacitors.</p> Thu, 26 May 2005 09:30:05 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=40http://www.doe.carleton.ca/seminars/show_one.php?id=41<p>Embedded random access memories can occupy up to 70% of the total area ofnmodern System on Chips (SoCs). Embedded SRAMs are the most popularly usedndue to their  robustness compared to DRAMs. Owing to a number ofnconstraints, embedded SRAMs have a significant impact on power, performance,ntestability and yield of complex SoCs. In this presentation, some of thesenissues will be discussed.</p> Tue, 24 May 2005 13:00:00 GMThttp://www.doe.carleton.ca/seminars/show_one.php?id=41