Search

GaN smart power chip technology has been realized using a GaN-on-Si HEMT platform, featuring monolithically integrated high-voltage power devices and low-voltage peripheral devices for mixed-signal functional blocks. In particular, this brief presents the imperative analog functional block-voltage reference generator for smart power applications with wide-temperature-range stability. The circuit is capable of proper functions within a wide temperature range from room temperature up to 250°C , illustrating the unique advantage of the wide-bandgap GaN in high-temperature operation. The voltage reference generator was designed with an AlGaN/GaN HEMT and Schottky diodes, and the devices were operated in the subthreshold regime to obtain low power consumption. The voltage reference generator achieved an average drift of less than 0.5 mV/°C and can be used as a reference voltage in various biasing and sensing circuits.

In this paper, a zero-bias mixer using a lateral field-effect diode fabricated on standard GaN-on-Si AlGaN/GaN high-electron-mobility-transistor wafers is demonstrated. The diode features strong nonlinearity near zero bias, enabled by a threshold-voltage modulation using a fluorine-plasma-treatment technique. The maximum change in conductance was adjusted to ~0 V, leading to optimal conversion loss (CL) of the mixer at zero bias and eliminating the need for any dc supplies. The mixer is characterized from room temperature (RT) to 250degC . At 2.5 GHz and at RT, the CL and third-order intermodulation intercept point are 12.9 dB and 17.64 dBm, respectively. The operation of the proposed diode is modeled by a physical equivalent circuit, with the element values extracted from the measured S-parameters. The voltage-biasing dependence of the CL can be explained by the model. The high-temperature operation of the mixer shows that the proposed mixer can perform well in high-temperature and ultralow-power applications.

We present a novel and efficient motion segmentation and tracking algorithm that follows the shift and align paradigm. We introduce two statistical tests to evaluate the similarity of aligned image pixels or patches and we use them to determine the spatial extend of each segment. The one statistical test is fast and accurate when the noise is moderate and the other employs a sophisticated noise model involving the Mahalanobis distance to handle correlated noise. Direct computation of the Mahalanobis distance is prohibitively expensive so we apply the Sherman-Morrison-Woodbury identity and amortization to reduce the cost by several orders of magnitude. We tested both versions of the algorithm on a variety of image sequences (indoor and outdoor, real and synthetic, constant and varying lighting, stationary and moving camera, one of them with known ground truth) with very good results.

The finite-difference time-domain (FDTD) technique was extended to analyze the frequency response of surface acoustic wave (SAW) filters. In this method, the partial derivatives of quasi-static Maxwell's equations and the equation of motion are discretized to centered finite differences. In addition, the perfectly matched layer boundary condition was applied to reduce spurious reflections. Two structures are considered in this paper. First, the model was applied to analyze the influence of the number of electrodes on the frequency response of a SAW filter fabricated on a zinc oxide (ZnO) substrate. Then, the proposed method was further extended to analyze the frequency response of a ZnO/diamond/Si-layered SAW filter. The simulated results are in a good agreement with the existing experimental data, indicating that the FDTD method was an appropriate approach for modeling SAW devices.

Summarization: Recent growth in the number of digital images available motivates the development of image/video databases for the effective management of these ever-increasing images. A common image retrieval task requires retrieving all images in the database similar in image content to an example query image. In this paper, we develop a simple, fast and robust motion segmentation algorithm to separate images into independently moving objects and an indexing scheme that uses the trajectories, shapes and image flow vectors of the independently moving objects to insert/query the sequences in a database. In order to test our ideas, we developed a video database prototype and experimented with real images from scenes showing moving cars. Παρουσιάστηκε στο: Vision Interface'99