Publications & Independent Research

This work will review a newly introduced framework for automated detection of the P-wave arrival for real-time earthquake early warning systems. These systems are prone to suffering from erroneous detection due to high background noise. To determine relative efficacy of the EVSSLMS algorithm, its mean square error (MSE) performance and convergence time is compared to several other common algorithms of the least mean square (LMS) variation using seismic sensor data source from Stanford University’s Earthquake Dataset.

An introduction and analysis of the Discrete Cosine Transform and its utility in image compression applications. Custom DCT/IDCT methods written in MATLAB are used to analyze compression performance, impacts of discretization, and image SNR.

This work aims to introduce the generalized Renyi entropy and show how such an entropy measure can be used to characterize signal complexity through the consideration of time-frequency representations (TFR) as pseudo probability density functions. At a high level, this topic should serve well due to its relevance to recent trends in machine learning/artificial intelligence, network traffic analysis, and cybersecurity.

This work will introduce the IEEE P370 standard and examine one of the PCB fixture de-embedding techniques detailed in this specification, the 2x Thru calibration. The Qorvo TQP3M9035 LNA and its accompanying characterization board will serve as the device under test (DUT).

A demonstration of the calcualtion of the noise figure for a real RF system using collected IQ samples and a known input signal. Understand system noise figure and not just Front End noise figure performance. Know how your system will perform!