dept. of ee, ndhu 1 chapter one signals and spectra
TRANSCRIPT
1Dept. of EE, NDHU
Chapter One
Signals and Spectra
2Dept. of EE, NDHU
Why Digital ?
• Advantages
– Digital signals are more easily regenerated
– Digital circuits are more reliable and can be produced at lower cost
– Different types of digital signals can be treated as identical signals in transmission and
switching
– Digital techniques are naturally to signal processing functions that protect against
interference and jamming, or provide encryption
• Costs
– Very signal-processing intensive
– Need to synchronize at various levels
– Non-graceful degradation
3Dept. of EE, NDHU
Pulse Degradation and Regeneration
4Dept. of EE, NDHU
Typical Digital Communication System
5Dept. of EE, NDHU
Digital Communication Transformations
• Formatting
– Analog source: audio, speech, video signal
– Digital source: computer data, digital image
– Convert the source into a sequence of binary sequence
• Source encoding
– Efficiently convert the digital symbol into a sequence of binary digits
– Data compression: MEG encode, JPEG, Huffiman coding, MP3
• Channel encoder
– Introduce some redundancy in the binary information sequence that can be used at the
receiver to overcome the effects of noise and encounter the channel
6Dept. of EE, NDHU
Digital Communication Transformations
• Pulse modulation
– Map the binary information sequence into signal waveform
• Bandpass signaling
– Coherent: PSK, FSK, GMSK
– Non-coherent: DPSK, FSK
7Dept. of EE, NDHU
Basic Digital Communication Nomenclature
(Textual messages)
(Characters)
(7-bit ASCII)
(Symbol)
(Bandpass digital waveform)
8Dept. of EE, NDHU
Performance Criteria
• Analog communication systems
– The figure of merit is a fidelity criterion
– For example signal-to noise ratio, percent distortion, or
expected mean-square error between the transmitted and
received waveforms
• Digital communication systems
– Probability of incorrectly detecting a digit, or PE
9Dept. of EE, NDHU
Classification of Signals
• Deterministic and Random signals
– Deterministic signal means that there is no uncertainty with respect to its value at any t
ime, for example x(t)=5 cos 10t
– Random signal means that there is some degree of uncertainty before signal actually o
ccurs
– Random waveform is NOT possible to write an explicit expression, can be described b
y probabilities and statistical averages
• Periodic and Non-periodic signals
– A signal x(t) is periodic in time if there exits a constant T0 such that
– No value of T0 that satisfies equation (1.2) is called non-periodic signal
(1.2) for )()( 0 tTtxtx
10Dept. of EE, NDHU
Classification of Signals
• Analog and Discrete signals
– x(t) and x(kT)
• Energy and Power signals
– Energy signal is defined by the signal has nonzero but finite energy for all time
– Power signal is defined by the signal has finite but nonzero power for all the time
– Periodic signal and random signal are generally classified as power signals
– Both deterministic and non-periodic signals are generally classified as energy signals
(1.7) )(2/
2/
2lim
T
TTx dttxE
(1.8) )(1 2/
2/
2lim
T
TTx dttx
TP
11Dept. of EE, NDHU
Spectral Density
• Energy spectral density
– Where is defined as energy spectral density (ESD) of the signal x(t)
• Power spectral density
– The power spectral density (PSD) is
– See Example 1.1
(1.13) )()( 22
dffdfX(f)dttxE xx
(1.17) )(1 2
2/
2/
2
0
0
0
CdttxT
Pn
n
T
Tx
(1.18) )()( 02 nffCfG
nnx
)( fGx
2)()( fXfx
12Dept. of EE, NDHU
Autocorrelation
• A measure of how closely the signal matches a copy of itself as the copy is
shifted in the time
)()(
allfor )0()(
)()(
fR
RR
RR
xx
xx
xx
,)()()( dttxtxRx
13Dept. of EE, NDHU
Random Process
14Dept. of EE, NDHU
Random Process
• Stationary
– Strict-sense stationary if none of statistics are affected by a shift in the time origin
– Wide-sense stationary if
• Ergodic
– Time averages equal ensemble averages
– For example,
– The statistical properties of the process can be determined by time averaging over a sin
gle sample function
)(),( 2121 ttRttR xx
constant a )]([ xmtxE
2/
2/
)(/1limT
TTx dttxTm
15Dept. of EE, NDHU
Some Useful Probability Distributions
• Binormial Distribution
– Let X be a discrete random variable X=1 or X=0, with probability p an 1-p
• Uniform Distribution
• Gaussian (normal) Distribution
• Chi-square (exponential) Distribution
• Rayleigh Distribution
• Ricean Distribution
• Lognormal Distribution
16Dept. of EE, NDHU
Autocorrelation and Power Spectral Density
17Dept. of EE, NDHU
Autocorrelation and Power Spectral Density
18Dept. of EE, NDHU
Normalized Gaussian Probability Density Function
19Dept. of EE, NDHU
White Noise
Figure 1.8 (a) Power spectral density of white noise.(b) Autocorrelation function of white noise.
20Dept. of EE, NDHU
Linear Systems
• Frequency response
• Power spectral density
• Distortionless transmission
)}(Re{
)}(Im{)( where,
)()(
)()(
1
)(
fH
fHTanf
efHfX
fYfH fj
)()()( 2 fGfHfG xY
)()( 02 0 ttkekfH ftj
21Dept. of EE, NDHU
Ideal Filter
• Transfer function
• Impulse response
)(2sin2)()( 0
222 0 ttfcfdfeedfefHth uu
f
f
ftjftjftju
u
02)(
)(
and
for 0
for 1)( where
)()(
ftjfj
u
u
fj
ee
ff
fffH
efHfH
22Dept. of EE, NDHU
Impulse Response of the Ideal Low-pass Filter
23Dept. of EE, NDHU
Realizable Filter
24Dept. of EE, NDHU
Butterworth Filter
• Magnitude frequency response for the n-th order
1 )/(1
1)(
2
n
fffH
nu
25Dept. of EE, NDHU
RC Filtering an Ideal Pulse
26Dept. of EE, NDHU
Baseband versus Bandpass
mDSB fW 2
27Dept. of EE, NDHU
Bandwidth Dilemma
Strictly bandlimited signal
Strictly time limited signal
• For all bandlimited spectra, the waveform are not realizable,
and for all realizable waveforms, the absolute bandwidth is infinite.
28Dept. of EE, NDHU
Bandwidth Criteria
Fig. Bandwidth of digital data. (a) Half-power. (b) Noise equivalent. (c) Null to null. (d) 99% of power. (e) Bounded PSD (defines attentuation outside bandwidth) at 35 and 50 dB.