Bruce E. Dunne

Analysis of gradient algorithms for TLS-based adaptive IIR filters

Steepest descent gradient algorithms for unbiased equation error adaptive infinite impulse response (IIR) filtering are analyzed collectively for both the total least squares and mixed least squares-total least squares framework. These algorithms have a monic normalization that allows for a direct filtering implementation. We show that the algorithms converge to the desired filter coefficient vector. We achieve the convergence result by analyzing the stability of the equilibrium points and demonstrate that only the desired solution is locally stable. Additionally, we describe a region of initialization under which the algorithm converges to the desired solution. We derive the results using interlacing relationships between the eigenvalues of the data correlation matrices and their respective Schur complements. Finally, we illustrate the performance of these new approaches through simulation.

Low-cost EEG-based sleep detection

A real-time stage 1 sleep detection system using a low-cost single dry-sensor EEG headset is described. This device issues an auditory warning at the onset of stage 1 sleep using the “NeuroSky Mindset,” an inexpensive commercial entertainment-based headset. The EEG signal is filtered into low/high alpha and low/high beta frequency bands which are analyzed to indicate the onset of sleep. Preliminary results indicate an 81% effective rate of detecting sleep with all failures being false positives of sleep onset. This device was able to predict and respond to the onset of drowsiness preceding stage 1 sleep allowing for earlier warnings with the result of fewer sleep-related accidents.

Software Defined Radio for direction-finding in UAV wildlife tracking

Two Software Defined Radio (SDR) approaches for assisted wildlife tracking with a multi-rotor Unmanned Aerial Vehicle (UAV) are described. The approaches are Pseudo Doppler (PD) and Yagi Rotation (YR), where each method corresponds to a common Radio Direction-Finding (RDF) technique. Pseudo Doppler utilizes the Doppler Effect by digitally sampling four (or more) monopole antennas in a sequential and circular order, simulating the rotation of a single monopole antenna. This approach can be complex with SDRs and suffers from reduced signal reception from the omnidirectional monopoles. The Yagi Rotation technique utilizes the directional Yagi antenna, rotating 360° about the antennas center of mass. As the antenna is rotated, the SDR System calculates the Signal to Noise Ratio (SNR) and Angle of Arrival (AoA) of the incoming radio signal. The Yagi Rotation was implemented and field tested; numeric results shown in this paper indicate that system accuracy is limited to the azimuth beamwidth of the Yagi antenna. Due to complexity and reduced signal reception from omnidirectional monopoles, the Pseudo Doppler approach was not implemented; however, a system design is presented. Finally, the SDR methodology and UAV control implications of each approach are discussed in this paper.

A hands-on Bioelectric Potentials course for Electrical Engineering majors

We describe a new course in Bioelectric Potentials, offered to seniors in a 4-year Electrical Engineering program, that integrates bioelectric mechanisms, numerical methods and simulation, hardware design, and technical writing. The course focused on mechanisms governing the generation of bioelectrical signals together with measurement and analysis techniques in muscle and nerve electrophysiology. Topics covered included the Nernst potential; development of action potentials; impulse propagation and nerve conduction; membrane biophysics; excitation-contraction coupling; and the development and management of cardiac arrhythmias. The course incorporated quantitative modeling and simulation learning exercises for data analysis and outcome prediction. Students were exposed to numerical methods for curve fitting, integration, differentiation, and solving ordinary differential equations. The course also incorporated hands-on activities including the building of an ECG amplifier and QRS detector circuit, and the measurement of their own EMG, ECG, and EEG signals using the ADInstruments PTB 32 Teaching System. Technical communication skills, both oral and written, were reinforced through three project reports and in-class presentations of pertinent journal articles. Assessment of student learning and achievement of course objectives will be presented.

Eye gaze detection system for impaired user GUI control

Gaze point, or point of regard, refers to the point in space where an individuals visual attention is focused. Estimating gaze point at a given time can provide a host of information pertaining to intention and perception of a scene. This information can be applied toward enhancement of Human-Computer Interaction (HCI), making such interfaces more smooth and efficient. For the population with limited or no mobility, a gaze point estimation system that accurately selects components of a computer application is extremely beneficial. Herein, a custom gaze point detection HW/SW system intended to allow the mobility hindered population the ability to control selection in a computer interface via gaze is presented. The principles of the image difference method for pupil detection, coupled with glint detection and calibration were implemented for an accurate, occlusion-immune estimation of gaze point in real-time. Under the ideal scenario of static head pose and lighting environment, the system was accurate to 1.05°. The gaze estimator tolerated small 1.5 inch head translations and over two orders of magnitude change in ambient illuminance, while sacrificing less than 1° of accuracy.

The effect of photodiode shape on dark current for MOS imagers

The effect of photodiode (PD) shape was studied in the attempt to reduce the dark current in MOS imagers. In such imaging systems, each pixel ideally produces a voltage directly proportional to the intensity of light incident on the PD. Because of non-idealities, the PD performance is compromised by the presence of dark current, which becomes the most significant source of noise degrading overall image quality, particularly for low light environments. Unfortunately, due to the statistical variability of dark current, it is not possible to simply correct the readout voltage error via subtraction. To minimize the effect, recent research suggests that PD shape and features have an influence on dark current levels. We test that assertion by considering PDs with different corners while maintaining high fill-factor rates, along with rectangular and triangular shapes to exploit charge transfer characteristics. In all, five PD geometries were built to test the influence of PD shape on the dark current signal including a traditional square shape, two square shapes with increasingly rounded corners (135 and 150 degrees), a triangular design with sharp corners and finally, a triangular design with 120 degree corners. Results indicate the PDs with a square shape and 90 degree corners exhibit the lowest dark current and highest readout voltage. Furthermore, the triangular shape suggests improved charge transfer characteristics; however, this improvement appears to be negated by an increase in dark current response. Therefore, our findings indicate that the traditional PD square shape is the preferred design.