Professor Costas D. Sarris
   Home  /  Research  /  People  /  Publications  /  Courses  /  Contact  /  Links  /  Misc
Click images to enlarge

Figure 1: Power map of an indoor floorplan determined through the S-MRTD technique (f=900 MHz, wall conductivity 0.05 S/m).

Figure 2: Adaptive mesh tracking of an optical pulse propagating in a dielectric waveguide junction.

Figure 3: Vertical electric field amplitude of a mode in a negative refractive index transmission line unit cell (periodic FDTD analysis).

Research Overview

For many years, research on Computational Electromagnetics has focused on improving the modeling of fundamental building blocks for complex, real-world systems. The results of this effort are reflected in a whole generation of simulators, which are now indispensable high-frequency design and analysis tools for academic and industrial researchers and engineers alike. Despite its impact and huge commercial success though, Computational Electromagnetics research still encounters an ever mounting set of new challenges, in accordance with Edward Teller's prediction that ''A state-of-the-art calculation requires 100 hours of CPU time on the state-of-the-art computer, independent of the decade.''

Moreover, translating our detailed knowledge of component parts to their impact on the behavior of the systems they belong to, defines a new frontier for computational physics. This frontier of multiscale modeling involves the integration of potentially heterogeneous models over the wide range of scales present in most physical problems. Our research is inspired by these challenges to address fundamental questions, devise novel techniques and investigate their application in critical areas of electromagnetic engineering.

For example, the fundamental component of any multiscale simulation scheme is adaptive mesh refinement, or the variability of the distance (in space and time) between the unknown quantities determined by a numerical solver. Our group has developed novel algorithms in this area which combine large mesh refinement ratios with numerical stability. Moreover, we are interested in the stochastic modeling, design and optimization of microwave and optical structures, whereby statistical variations of the geometric and material properties of a structure are embedded in the modeling process.

Along with fundamental research on computational algorithms, several active projects focus on exciting applications of numerical electromagnetics on communications, sensing and biomedicine. These projects are:
  1. Realisting assessment of small antenna MIMO system performance (with RIM Inc.).
  2. Precursor fields in ultrasound and applications to brain imaging (with DRDC Toronto).
  3. Electromagentic compatibility of ultra-wideband radio systems (Early Researcher Award).
  4. Modeling and optimization of ultra-wideband systems in railway tunnel environments (with Thales Rail Signalling Solutions).
  5. Design and optimization of sub-wavelength focusing microwave and plasmonic meta-screens and exploration of wireless energy transfer applications.
  6. Analysis of statistical uncertainty in periodic structures and stochastic surface impedance boundary conditions for rough surfaces.
  7. Multi-physics modeling of graphene-based RF devices.
  8. Design of meta-arrays for RF ablation in image-guided radiotherapy (with the STTARR Innovation Center at the Princess Margaret Hospital).
  9. Diagnostic modeling of coaxial cable based home networks for broadband access (with EXFO).

For sample results from completed past projects, please follow the links below:
  1. Fast and wideband wireless channel modeling
  2. Modeling of wave propagation in dispersive media and periodic structures, with emphasis in microwave and optical metamaterials
  3. Electromagnetic compatibility / interference (EMC/EMI) in wireless and wireline systems
  4. Fundamental research in time-domain numerical techniques

Our research is currently being supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), the Ontario Ministry of Research and Innovation through and Early Researcher Award, Thales Rail Signalling Solutions, Research in Motion, Defence Research and Development Canada (DRDC)-Toronto, the Eugene V. Polistuk Chair, Ultra Electronics, EXFO Corp.

Our research is greatly facilitated by a 48-node Linux cluster, funded by the Canada Foundation for Innovation and the Ontario Innovation Trust and powerful GPU-enabled workstations funded by NSERC.


   Home / Research / People / Publications / Courses / Contact / Links / Misc
© Costas D. Sarris  
Link: ECE
- UofT Link: UofT