Laboratory Explorations Project
Laboratory
instruction for communications and signal processing
Abstract
The Laboratory Explorations Project aims to
develop and disseminate hands-on guided learning explorations for digital communication
and real-time signal processing. The
goal is to provide high quality and low-cost laboratory experiences for
students at the high school, undergraduate and graduate levels. Curricular materials support both traditional
group instruction in a laboratory setting and guided self-study. The use of
mobile device Apps and ultra low-cost platforms make the delivery of
instruction widely available and support the weLab international initiative.
Why
a new approach?
Why re-invent teaching and learning for laboratory
experiences in digital communications or digital signal processing? Motivated
by The Ohio State University’s transition to a semester-based academic
calendar, we evaluated our traditional instruction and found it standard and
stale. A bottom-up assessment of goals for student outcomes and costs of
instruction has led to this
project. Six principles guide our approach.
Cost. Radio frequency equipment is expensive. Books are expensive. But, mobile
devices are widely available.
Sensory
perception. Many
students are tactile learners. Acoustic frequency operation allows students to
use both visual perception of graphed signals and auditory perception of
modulated signals to build intuition and understanding.
Systems level
design. In most
existing lecture and laboratory instruction, students work with a “black-box”
approach that never requires them to face system-level considerations and
motivations for design decisions; here, students are asked to be their own
systems engineer.
Concepts
versus tools. In a lab
experience of short duration, students may invest time in learning tools at the
expense of learning concepts, yet not master even the tools. For student
success in a short sequence of single-session exercises, we provide a path with
a short learning curve on design and implementation tools. Motivated by success
with concepts and working engineering solutions, students can then expand their
experience to master additional tools.
Extensibility The acoustic approaches present
channel impairments that mirror radio frequency engineering challenges; indeed
the acoustic solutions developed by students are directly successful as the IF
and baseband processing for RF channels.
Optional lab explorations are provided for extension to RF
implementations.
Sustainability The software-defined solutions
developed in our approach are effortlessly maintained, in contrast to the short
life-cycles of approaches centered on hardware solutions.
High
school STEM education, too
Taylor Williams is applying elements of the Laboratory
Explorations approach in his Pre-Engineering Technologies course and AP
Computer Science course taught to Seattle high school students. One of four
mentored projects is digital communications, and all projects
are grounded in the idea of collaborative learning through a pedagogical model
called Complex Instruction. In the digital communications project, students
explore the math and science behind how a cell phone sends picture
messages. This is done through a guided
lab where students send a digitized and encoded picture using sound, then
record and reconstruct the image from that sound. This is actually a simple
modem, which is a great platform to explore different techniques used in signal
processing, even though cell phones use EM waves, not acoustic waves. Students
first learn about digital representation strategies and digital image
representation. Then, they explore the basics of acoustics learning concepts of
frequency, bandwidth, and filtering.
Finally, students evaluate and choose from a menu of options for how to
encode and decode their information, working to balance the speed and accuracy
of the overall communication system. Mr.
Williams piloted this unit with a group of eighth graders at the Summer Program
in Mathematical Problem Solving (SPMPS) camp in New York in 2013.
Digital
Communication Laboratory
A draft
outline for the Digital Communications Laboratory Explorations is available in pdf. Any computing device with audio-frequency
recording and playback capabilities can suffice to meet the hardware
requirements. For example, we have
cross-compiled Matlab code to employ an iPhone as a transmitter.
Real-time
DSP Laboratory
For real-time DSP instruction, we aim to invert the traditional teaching strategy. Since the introduction of the TMS320 series from Texas Instruments over 30 years ago, standard laboratory instruction for real-time signal processing has been to teach assembly language programming and the associated requisite review of the dual Harvard computer architecture. Then, students learn to write interrupt service routines, with all other considerations abstracted away. Students emerge from the experience competent with indirect addressing, but unable to field an embedded system. In contrast, we propose to leverage the recent advances in cross-compilers to focus on fielding an embedded system for real-time implementation of signal processing algorithms. Matlab and Simulink support this vision for many low-cost platforms, including Raspberry Pi, BeagleBoard and several Texas Instruments platforms. Exercises with the Raspberry Pi include frequency-selective filtering, angle of arrival estimation using an acoustic array, LMS adaptive interference cancellation, system identification, linear predictive speech coding, audio effects, and real-time image processing.
Support
Grant support for a graduate teaching
assistantship was generously provided by the Mathworks Academic Support
Program.
Contacts
Please contact the authors with questions and
comments.
© 2015
LC Potter