Technologies

Virtual EM focuses on four broad technology areas:

  1. Computational Electromagnetics
  2. Antennas and RF Circuits
  3. Wireless Sensor Networks
  4. Machine Learning

1. Computational Electromagnetics

Both full-wave and asymptotic techniques are covered. Mature computational engines have been developed utilizing full-wave methods such as Finite Element Method (FEM) and the Method of Moments (MoM) along with acceleration algorithms such as Fast Multiple Method (FMM) and Domain Decomposition. Further algorithmic and hardware accelerations are implemented through Fast Fourier Transform (FFT), multi-core CPUs, Graphics Processing Units (GPUs) and special purpose processors such as MD-GRAPE. Asymptotic Methods such as multi-bounce Physical Optics (PO) have been hybridized with the full-wave methods for treating electrically large platforms. In addition to using commercially available hardware for acceleration, Virtual EM has been funding internally the development of a special purpose processor.

2. Antennas and RF Circuits

Both large and small, high and low power, active and passive antenna solutions are being developed for both defense and civilian applications. Reconfigurable aperture antennas are being developed for composite aircraft platforms that are capable of both receive and high power transmit. Antennas include adaptive impedance matching circuits managed by specialized firmware installed on microcontrollers and use innovative printing technologies for conformal installation. Some of the prototypes utilize the Self-Structuring Technology licensed from Monarch Antenna, Inc. (Ann Arbor, Michigan) offering self-healing and self-adapting features. Low-power passive designs embedded into PCBs have been developed for ZigBee, RFID and WiFi applications.

3. Wireless Sensor Networks

Virtual EM has a broad program on wireless sensor networks involving a variety of sensors and communication architectures. While the sensors range from those utilizing cutting-edge chemical and biological sensing modalities to off-the-shelf solutions such as temperature, humidity and pressure, the communication architectures include RFID, ZigBee, 6loWPAN and WiFi. Analog and digital sensor interfaces as well as the associated firmware are designed and prototyped in house.

4. Machine Learning

Machine learning algorithms such as Support Vector Machines (SVMs) are being developed for target detection in clutter (for detecting submarines) and for picking out targets from a sea of decoys (for differentiating missiles from decoys). Same algorithms are being adopted for civilian applications to improve the sensitivity of a variety of sensors.

The above technologies are supported by past and present SBIR projects described on the Projects page.