Lecture 1: Applications Noise Cancellation

Hands free
• noise suppression/cancellation (handsfree in microphone
cars, airplanes, headphones)
• echo cancellation
Noise source +
Adaptive
• channel equalization microphone -
filter

• channel estimation or system identification

• Goal:
• target tracking/radar : velocity and acceleration • Why adaptive filter?
estimation from position measurements
Examples:
• engineering/medical/biological/financial • Active noise cancelling headphones
applications etc. • Reduction of 50Hz interference in electrocardiography
• ... • Reduction of noise in speech, e.g., in the cockpit of
(military) aircrafts

1 2
Magnus Jansson, Signal Processing, KTH Magnus Jansson, Signal Processing, KTH

Adaptive Equalization Training sequence

n 𝑡 𝑥 𝑡
𝑥 𝑡 𝑥 𝑡 + e 𝑡
ℎ 𝑤 -

Communication channel Receiver

• Goal: 𝑥 𝑡 𝑥 𝑡

Examples:
• Digital communication systems
• Audio processing
• Image enhancement

3 4
Magnus Jansson, Signal Processing, KTH Magnus Jansson, Signal Processing, KTH
 System Identification Tracking (state estimation)
System n 𝑡
Observation:
𝑦 𝑡
ℎ 𝑦 𝑡 ℎ 𝑥 𝑡 𝑣 𝑡
𝑥 𝑡 e 𝑡

𝑦 𝑡 - Goal: To estimate 𝑥 𝑡
𝑤

Model State model:

𝑥 𝑡 1 𝑓 𝑥 𝑡 ,𝑢 𝑡 𝑤 𝑡
• Goal: 𝑤 𝑛 ℎ 𝑛 or 𝑦 𝑡 𝑦 𝑡

Examples: Examples:
• Channel estimation • GPS/GNSS positioning
• Modeling of dynamic systems • Target tracking
• Cell tracking from image
sequences

5 6
Magnus Jansson, Signal Processing, KTH Magnus Jansson, Signal Processing, KTH

Advices from last year Chapter 2: Stationary Signals

students
• “Take the course project seriously and it would help.”
• “Try to understand what is going down behind the
equations”
• “Go for the bonus points.”
• ”To do the homeworks.”
• ”Start doing exercises sooner. The exercise sessions
are few.”
• “Do the homeworks properly and start early with the
projects.”

7 8
Magnus Jansson, Signal Processing, KTH Magnus Jansson, Signal Processing, KTH
 Spectrum Jointly stationary

9 10
Magnus Jansson, Signal Processing, KTH Magnus Jansson, Signal Processing, KTH

Filtering Filtering, cont’d

11 12
Magnus Jansson, Signal Processing, KTH Magnus Jansson, Signal Processing, KTH