+ All Categories
Home > Documents > Chapter 2. Signals Husheng Li The University of Tennessee.

Chapter 2. Signals Husheng Li The University of Tennessee.

Date post: 11-Jan-2016
Category:
Upload: russell-hill
View: 217 times
Download: 1 times
Share this document with a friend
38
Chapter 2. Signals Husheng Li The University of Tennessee
Transcript
Page 1: Chapter 2. Signals Husheng Li The University of Tennessee.

Chapter 2. SignalsHusheng LiThe University of Tennessee

Page 2: Chapter 2. Signals Husheng Li The University of Tennessee.

Homework 2Deadline: Sept. 16, 2013

Page 3: Chapter 2. Signals Husheng Li The University of Tennessee.

Spectrum

Physically, the signal is transmitted in the time domain.

It is more convenient to study the signal in the frequency domain.

The frequency domain description is called the spectrum.

The frequency description of signal can be obtained from Fourier transform:

Page 4: Chapter 2. Signals Husheng Li The University of Tennessee.

Example: Rectangular Pulse

Time domain

Frequency domain

Page 5: Chapter 2. Signals Husheng Li The University of Tennessee.

Signal EnergyRayleigh’s Theorem: The signal energy is

given by

Integrating the square of the amplitude spectrum over all frequency yields the total energy.

|V(f)|^2 is called the energy spectral density.

Page 6: Chapter 2. Signals Husheng Li The University of Tennessee.

Band Limited Signals

A signal should not use all bandwidth. Hence, we have to limit its band.

Sinc function is a band limited one

A band limited signal is infinite in the time, which is impossible in practice.

Page 7: Chapter 2. Signals Husheng Li The University of Tennessee.

Frequency Translation

We need to transform a baseband signal to much higher frequency one. (Why?)

It is equivalent to multiplying a sinusoidal signal having the carrier frequency.

Page 8: Chapter 2. Signals Husheng Li The University of Tennessee.

RF Pulse

time

frequency

Page 9: Chapter 2. Signals Husheng Li The University of Tennessee.

ConvolutionWhen a signal is passed through a linear time

invariant (LTI) system, the output is the convolution of the input signal and the system impulse response.

In the frequency domain, the convolution is equivalent to multiplication:

Page 10: Chapter 2. Signals Husheng Li The University of Tennessee.

Transfer Function

Each LTI system can be represented by its transfer function.

Page 11: Chapter 2. Signals Husheng Li The University of Tennessee.

Signal Transmission: Distortionless Case The output is undistorted if it differs from the

input only by a multiplying constant and a finite time delay:

In the frequency domain, it is equivalent to

In practice, the signal is always distorted.

Page 12: Chapter 2. Signals Husheng Li The University of Tennessee.

Linear Distortion: Amplitude

Linear distortion includes any amplitude or delay distortion associated with a linear transmission system, which is easily descried in the frequency domain.

The amplitude could be distorted.

Low frequency attenuated High frequency attenuated

Page 13: Chapter 2. Signals Husheng Li The University of Tennessee.

Linear Distortion: Phase If the phase shift is not linear, the various

frequency components suffer different amounts of time delay, called phase or delay distortion.

The delay is given by

Page 14: Chapter 2. Signals Husheng Li The University of Tennessee.

Two Waveforms: Example

Page 15: Chapter 2. Signals Husheng Li The University of Tennessee.

EqualizationLinea distortion is theoretically curable through

the use of equalization networks.

Digital transversal filter

Page 16: Chapter 2. Signals Husheng Li The University of Tennessee.

Multipath in Wireless

The multiple paths in wireless communications cause different delays along different paths, thus causing inter-symbol interference.

For example, consider two paths:

Page 17: Chapter 2. Signals Husheng Li The University of Tennessee.

Destructive Interference (two-path)

Page 18: Chapter 2. Signals Husheng Li The University of Tennessee.

Nonlinear Distortion

Many devices could have nonlinear transfer characteristics.

The nonlinear transfer characteristic may arouse harmonics.

Page 19: Chapter 2. Signals Husheng Li The University of Tennessee.

Transmission Loss

Power gain: g=P_out / P_in

dB scale: g_dB = 10 log_10 g

For linear system of communication channel, we have

Page 20: Chapter 2. Signals Husheng Li The University of Tennessee.

Typical Values of Power Loss

Page 21: Chapter 2. Signals Husheng Li The University of Tennessee.

Example: Radio TransmissionFor the case of free-space transmission, the

loss is given by

Consider the antenna gains, the received power is given by

Page 22: Chapter 2. Signals Husheng Li The University of Tennessee.

Example: Satellite Communication

Page 23: Chapter 2. Signals Husheng Li The University of Tennessee.

Doppler ShiftA passing automobile’s horn will appear to

change pitch as it passes by.

The change in frequency is called Doppler shift.

When the moving speed is v and the angle is ϕ, the Dopper shift is

Page 24: Chapter 2. Signals Husheng Li The University of Tennessee.

Homework

Deadline: Sept. 9, 2013

Page 25: Chapter 2. Signals Husheng Li The University of Tennessee.

Ideal FilterAn ideal bandpass filter is given by

Page 26: Chapter 2. Signals Husheng Li The University of Tennessee.

Filtering

Perfect bandlimitiing and timelimiting are mutually incompatible.

Rise time is a measure of the ‘speed’ of a step response:

Page 27: Chapter 2. Signals Husheng Li The University of Tennessee.

Quadrature FilterA quadrature filter is an allpass network that

merely shifts the phase of the positive frequency components by -90 degrees.

The output of a quadrature filter is called the Hilbert transform of the input.

Page 28: Chapter 2. Signals Husheng Li The University of Tennessee.

Properties of Hilbert Transform

Page 29: Chapter 2. Signals Husheng Li The University of Tennessee.

Bandpass Signals and SystemsA bandpass signal has the following frequency

domain property:

The time domain bandpass signal can be written as

Page 30: Chapter 2. Signals Husheng Li The University of Tennessee.

Spectrum and Waveform of Bandpass Signal

Page 31: Chapter 2. Signals Husheng Li The University of Tennessee.

Quadrature-Carrier Description of Bandpass SignalA bandpass signal can be decomposed to in-

phase and quadrature components:

Page 32: Chapter 2. Signals Husheng Li The University of Tennessee.

Frequency Domain of Bandpass SignalThe frequency domain of a bandpass signal is

given by

The in-phase and quadrature functions must be lowpass signals:

Page 33: Chapter 2. Signals Husheng Li The University of Tennessee.

Lowpass Equivalent Signal In the frequency domain, we have the low pass

equivalent spectrum:

In the time domain, we have the lowpass equivalent signal:

In the frequency domain, we have

Page 34: Chapter 2. Signals Husheng Li The University of Tennessee.

Lowpass-to-bandpass transformationThe connection between and is

given by

In the frequency domain, we have

Page 35: Chapter 2. Signals Husheng Li The University of Tennessee.

Bandpass TransmissionWe can work on the lowpass equivalent spectra

directly:

Page 36: Chapter 2. Signals Husheng Li The University of Tennessee.

Carrier and Envelop Delay If the phase shift is nonlinear, we can

approximate it by using the Taylor’s expansion:

Page 37: Chapter 2. Signals Husheng Li The University of Tennessee.

Bandwidth and Carrier FrequencyA large bandwidth requires high carrier

frequency.

Page 38: Chapter 2. Signals Husheng Li The University of Tennessee.

Bandwidth: DefinitionAbsolute bandwidth

3 dB bandwidth

Noise equivalent bandwidth

Null-to-null bandwidth

Occupied bandwidth

Relative power spectrum bandwidth


Recommended