Random signals. Histogram of the random signal Continuous Time Sinusoidal signals.

Random signals

Histogram of the random signal

Continuous Time Sinusoidal signals

CT Signal (Periodic)• A simple harmonic motion is described

mathematically by

.int

,

sin

.

)(sin

.,1

)()(

),2cos()(

2

),cos()(

ervaltimegiven

ainincludedareperiodstimemorethatsensetheinsignal

theofnoscillatioofratetheinresultsFfrequencythegIncrea

distinctthemselvesaresfrequencie

differentdistinctwithsignalusoidaltimeContinuous

relatedinverselyarefrequencyandperiodtimeF

Twhere

txTtx

periodicbetofunctionafor

tFtAtx

F

ttAtx

p

apa

a

a

Complex Signal• A complex exponential signal is

.log

.exp

sin22

)cos()(

.int

sincos

)(

)()(

)(

tofromissignalanarangeFrequency

signalsonentialconjugatecomplexamplitude

equaltwoaddingbyobtainedbecansignalusoidalA

eA

eA

tAtx

sfrequencienegatveroduce

toneedweeconveniencalmathematicforOnly

je

Aetx

tjtja

j

tja

Discrete time sinusoidal signal

).2

1(

2

1)(

sin.3

.2int

sin.2

.

sin.1

:Pr

./12/1

.6/sin

.dim

)2cos()(

2

)cos()(

expsin

for

flyequivalentororwhenobtainedis

usoidtimediscreteainnoscillatioofratehighestThe

identicalareofmultipleegeran

byseparatedaresfrequenciewhoseusoidstimeDiscrete

numberrational

aisfrequencyitsifperiodicisusoidtimeDiscrete

operties

samplecyclef

frequencywithusoid

timediscreteashowsslidenexttheinfigurethe

samplepercyclesofensionsthehasf

nfnAnx

f

nnAnx

asressedissignalusoidaltimediscreteA

Discrete-time sinusoidal signal

Condition of Periodicity of DT signal

)(2

1

2

1,2

.......3,2,1,0),cos()(

sin

)cos()2cos(])2cos[(

sec

.)int

exp(

22

)2cos(])(2cos[

)()(

00

00

0

identicalorishableindistinguare

forkwhere

knAns

sequencesusoidalallresultaAs

nnnn

propertyondtheproveTo

egers

twoofratioaasressedbecanfN

kf

kNf

nfNnf

nxNnxperiodicbeTo

k

kk

ko

oo

oo

Effect of Sampling

Harmonically related complex exponential

• Continuous time exponentials

.tan

)()(

exp.

.int

.

.1

.....,3,2,1,0)(

0

00

0

00

2

tsconscomplexarbitraryarecwhere

ectsctx

followsasdconstrucrebecanansionseriesFourierdistinct

aresettheofmembersallandvalueanytaketoallowedisFkeger

positiveforanyTk

Tkperiodwithperiodicalsois

k

Tpperiodwith

periodicisthatsignalaSincekFfrequencylfundamentaork

T

kF

periodlfundamentawithperiodicaresignalsabovetheofall

keets

k

k

tjkk

kkka

pp

p

tkFjtjkk

Discrete time exponentials

Discrete time complex exponential is periodic if its relative frequencyIs a rational number. We choose f0=1/N. We define a set of harmonically

related complex exponentials by

Nffrequencylfundamentawithsetrelatedlyharmonical

formtoNnktonkfromsayonentials

complexNutiveconanychoosecanWesamplesNof

periodcommonahavesettheofmembersAllonentials

complexperiodicdistinctNonlyaretherethatmeansThis

nseeeens

thatnotewetimecontinuousthetostcontraIn

kens

knkfjnkfjnjNNknj

Nk

nkfjk

1

1,,exp

sec.

.exp

)(.)(

.........3,2,1,0)(

0

00

222/)(2

2

00

0

Negative Frequencies

Analog to Digital Converter

ADC: To process analog signals by digital means, it is first necessary to convert them into digital form, that is , to convert them into a sequence of numbers having finite precision. The procedure is called analog to digital conversion and the devices used is called AD converters.

1. Sampling: This is the conversion of continuous time signal into discrete signal obtained by taking samples of continuous time signal at discrete time instants. If xa(t) the input to the sampler, the output

xa(nT)=x(n) where T is the sampling interval.

2. Quantization: This is the conversion of discrete time continuous valued signal into a discrete time discrete valued (digital) signal. The value signal sample is selected from a finite set of possible values. The difference between the unquantized sample x(n) and the quantized output xq(n) is called quantization error.

3. In coding process each quantized output xq(n) is represented by a b-bit number.

In speech processing the digital signal is converted back into analog form again.

This process is called DA conversion. The DA conversion is done by connecting all the dots by some kind of interpolation.

• The figure shows a form of DA conversion called stair-case approximation

or zero-order hold.• Other approximations are possible, connecting successive points linearly

(linear interpolation) or fitting a quadratic through three successive points (quadratic approximation).

• For signals having finite frequency content (limited band width)sampling theorem specifies the optimum form of interpolation.

• Analog signal does not lose any information provided the sampling frequency is sufficiently high. Otherwise there is a problem called aliasing

Zero-order hold

Sampling of an analog signal