When

Monday, October 27, 2025 • 3:00–4:00 PM

Where

DMF 461, Bridgewater State University

Speaker

Paul Fredette, MSEE (Bridgewater State University)

Format

In-person with Zoom access; recording available. 

View recording ↗

Abstract

The presentation investigates the behavior of linear difference equations near the Shannon
sampling limit. It began as a practical question about the “bandwidth” of a smoothing filter
for noisy data from an Atomic Physics experiment at BSU. After a brief review of digital
signal processing, the discussion extends to continuous signal analysis, statistics, and
modeling. The first-order filter is a simple recursive model with exponential decay, widely used for
real-time smoothing (from sensor noise reduction to internet protocols). When driven by
sampled sinusoidal inputs, an unexpected response emerges as sampling approaches two points
per cycle—the threshold associated with the Shannon limit. We present solutions as the
sampling rate nears this boundary, highlighting the initially puzzling behavior and the
eventual explanation behind the observed “reflection” in the response.

Bio

Paul Fredette is an adjunct professor in his third year at BSU following a career in Electrical
Engineering. He co-founded two electronic communications equipment companies and served as
technical lead for the Audio/Visual systems created for the first national SANE TeleNursing
center, now used by the Massachusetts Department of Public Health. He currently teaches Physics
(PHYS 181), Electronics (PHOE 342), Computer Networking (COMP 430), and Mathematics (MATH 105).
Paul holds an MSEE from MIT (1975) and a BSEE from UMass Dartmouth (1971).

Questions? Contact an organizer:
Vignon Oussa,
Xiangfei (Fei) Chen,
Mahmoud El-Hashash.