Teaching Digital Communications by Building a BPSK System
R·I·T
Miguel Bazdresch
Rochester Institute of Technology, ECT ET
Rochester, NY
Complete BPSK Communications System as Implemented by Students
data
Framer
Mapper
bit 1 -> 1
bit 0 -> -1
SQR Raised
cosine
filter
+
X
Sound
Card
Cable
Adaptive
Bit Timing
Recovery
SQR Raised
cosine
filter
Envelope
Detector
Sound
Card
data
Deframer
Acos(2πfcn)
Reception
Signal Design
Solution
Choose dynamic, interpreted languages such as Octave and Matlab.
I Use a computer’s sound card as an analog front-end.
I Emphasize the importance of signal orthogonality.
I State the problem as learning to design, evaluate and optimize signals.
I Measure and evaluate the channel.
I Solve real-world effects such as timing and synchronization.
I
After filtering and demodulation, the students:
Transmitted Signal’s Eye Diagram
100
Get an open eye diagram.
I Achieve initial bit synchronization (the sound card’s
sampling clock is free-running and out of our control).
I Track bit timing to compensate transmitter-receiver clock
differences.
I
50
0
−50
Received Eye Diagram
−100
0.65
0
10
20
30
40
Sample index
Channel
Students learn to find and
understand a channel’s response:
Benefits
1.2
·10−4
0.5
0
−0.5
−1
Orthogonal Signaling
Emphasizing orthogonal signaling simplifies exposition. The problem of
transmitting numbers x, y is reduced to generating
R ∞a voltage
v (t) R= xs1(t) + ys2(t). Reception is done by x̂ = −∞ v (t)s1(t)dt = x and
∞
ŷ = −∞ v (t)s2(t)dt = y .
0.55
0
10
20
30
40
0
2
4
Sample index
6
8
10
12
14
Sample number
Rx Constellation With No Bit Tracking
Rx Constellation Over 500 Bits
0.8
0.6
0.5
0.5
0.2
0
−2
−1
0
1
2
Frequency (Hz)
·104
Amplitude
0.4
Amplitude
It is possible for students to design and build a full communications
system in one semester. Students learn the principles in depth while
applying engineering methods to make theory agree with practice.
The most common objection: This course lacks breadth – not a lot of
textbook pages are covered.
Spectrum magnitude
Using white-noise or an impulse
as input.
I Identify largest band where
narrowband transmission is
possible.
Every pair of computers will have a
different channel.
0.6
0.5
Channel Transfer Function
1
I
Cluster Variance per Sample
1
Signal amplitude
Students learn to design
orthonormal signals that:
I Have a specified bandwidth.
I Meet a specified data rate.
I Meet the specifications of the
computer’s sound card.
Students also learn to use eye
diagrams to evaluate a signal’s
quality, both in transmission and
reception.
Signal amplitude
Theoretical and simulation-based courses are less attractive and arguably
less useful to industrial-oriented programs. How to enable undergraduate
students to build, test and evaluate a complete digital communications
system in one semester?
Cluster variance
Introduction
0
−0.5
−0.5
0
100
200
300
Bit number
Synchronization
0
400
500
0
500
1,000 1,500 2,000 2,500
Bit number (decimated)
Carrier synchronization is difficult. The effects can be observed and
then solved (inefficiently) using AM DSB-LC modulation.
Rochester Institute of Technology – Telecommunications Engineering Technology
Mail: [email protected]
WWW: http://2pif.info