EM Simulation Company with Deep Roots in R&D
MWJ: How did Remcom get started?
SL: Remcom was started in the early 1990s by Dr. Raymond Luebbers, a professor of electrical engineering at the Pennsylvania State University, and two of his graduate students, Christopher Penney and myself (Scott Langdon), in State College, Pennsylvania. Dr. Luebbers had a research group working in computational electromagnetics with a focus on the finite difference, time domain (FDTD) method. At that time the work was primarily concerned with radar cross section analysis and improvements to the FDTD method for performing these calculations in an efficient manner on existing computer resources. There was a strong unfilled need for a tool which would provide better problem definition, visualization, and result post-processing. Around 1993, a project was presented to the group to develop a software tool for the analysis of mobile telephones and specifically for computing the Specific Absorption Rate (SAR) exposure from the phones on humans. That project grew into a software tool that came to be known as "XFDTD," since the original program ran under the X-Windows system on Unix machines. A decision was made to market the program commercially, though we were researchers with no commercial experience. Fortunately, the response was quite positive and we decided to keep improving the product.
MWJ: How do you differentiate your products from those of other EM companies?
SL: As the computational EM market has grown so has the size and complexity of the problems to be solved. We make particular effort to enable XFdtd to generate and manipulate complex geometries and run on large and fast hardware. For example, we were the first in our industry to fully support fine-grain MPI with our solver (2003) and the first, according to our customers, to import, modify, and solve a complete mobile phone (2002). We enabled XFdtd calculations to leverage graphics processing units (GPUs) using NVIDIA’s CUDA, and later Fermi, architectures beginning in the summer of 2010. We’ve also ported our solver to IBM Blue Gene.
Many tools available today have very similar features on paper. Having a tool is only one part of the solution though. Most users are not experts with the product out of the box, so we try to provide quality support to our customers to get them running the software effectively as quickly as possible. This starts with training courses that we offer, but continues through personal assistance for specific tasks, the development of custom scripts for performing tasks, and the addition of new features when required. We realize that our customers’ success with their products wraps back in part to how useful our software is in meeting their needs, and we want to help as much as possible.
MWJ: What is the state of simulation technology today? Could you give us an example of something you can solve today that would have been impossible five years ago?
SL: Quite simply, the problems people are simulating today were not even dreamed about when we started. Years ago we were trying to perform radar scattering analysis of canonical shapes like spheres and plates or doing an antenna analysis of a dipole or monopole antenna. The market now is very broad with some customers simulating devices like MRI coils in the MegaHertz range all the way up to others working with photonics in the light spectrum. We are able to do simulations of entire aircraft now that are perhaps hundreds of wavelengths long, or work on very small scales with PCB analysis on micron scales. The amazing improvement in computer processing power, especially the introduction of high performance computing and GPU availability, has created tremendous progress in the scale of problems possible. Also, new technologies in the simulation methods such as sub-cell refinements that allow modeling of detailed features while using significantly less computer resources have also helped.
MWJ: Remcom seems to be well connected with universities and R&D. What’s the nature of this relationship?
SL: We recognize that many of our future customers are students now and that in academic environments there are frequently limited resources for purchasing scientific software. We provide our products for very little cost for academic use and at significant discount for funded research. Several of our people have ties with universities in various forms and we often collaborate with research efforts.
MWJ: In February, Remcom was selected by the US Department of Homeland Security (DHS) to develop an advanced software tool to model and assess the effects of wind turbines on radar systems used throughout the United States. What kinds of challenges will this kind of problem present to your developers and modelers?
SL: For more details on the DHS contract, we would like to refer readers to our , which was approved by DHS; however, we can comment on the challenges of modeling wind turbine effects on radar. The problem is actually quite complex. Radar pulses encounter Doppler shifts that vary along the lengths of the blades and change with wind speed and direction. At large wind farms, this is complicated by a vast amount of multipath from signals that reflect and diffract between wind turbines. These effects result in a great deal of interference to radars. The key challenge is to handle all of these effects with a large number of wind turbines in motion, and do so in a reasonable amount of time.
MWJ: EM simulation has been used for years by RF and signal integrity designers. Does your Wireless InSite software, which performs analysis of site-specific radio propagation and wireless communication systems play to a significantly different type of engineer? If so, what type of user are we talking about and do they require different support?
SL: Engineers have been using RF propagation models to simulate radar, broadcast signals, and communications for many years. The Wireless InSite modeling suites are used for similar types of analysis, but they extend analyses to detailed terrain modeling or to urban and indoor environments, where multipath and shadowing substantially affect the received signals. Engineers and analysts tend to use this software for DOD mission planning, design of communications links or wireless networks, or a variety of other scenarios where buildings or complex terrain are a significant issue.
The type of support varies widely depending on the user and application, but generally focuses on helping users to adapt the software to their applications. Our customers are constantly challenging Remcom to model more complex scenarios and are applying the models to new applications that we, as developers, never considered when the models were first introduced. The accuracy and diversity of Wireless InSite's propagation suite enable it to handle this wide variety of applications. We have customers interested in just about every aspect of propagation phenomena, from atmospheric attenuation to ground wave propagation. Wireless InSite gives them the tools to simulate, visualize, and understand the propagation in their scenario. Customers can take this understanding and maximize the performance of their RF systems.