(1) Introduction

Signal Processing:

1. Acquisition

2. Representation

3. Manipulation

4. Transformation

5. Storge

Sampling: is used to convert a continuous-time signal into a discrete-time signal.

Quantization: is used to convert a discrete-time signal into a digital signal.

Nyquist Condition: Fs ​ ≥ 2Fmax 

​

Discrete-Time Signals
x[n]is a sequence of numbers defined for every value of the integer variable n
x[n]is NOT defined for non-integer values of n
The duration of length of a discrete-time signal Lx : the number of samples from the first non zero sample x[n1 ]to the last
​ ​

non-zero sample x[n2 ] ​

Lx = n2 − n1 + 1​ ​ ​

the range n1 ​ ≤ n ≤ n2 is donated [n1 , n2 ]is called the support of the sequence
​ ​ ​

Energy of a sequence
∞
Ex = ∑ ∣x[n]∣2
​ ​

n=−∞

Power of a sequence
Ex
There’s a problem with the power calculations as Px ​ = T , if
​

​ Ex → ∞and T → ∞
​

The power will be undefined, so we calculate the power among an “observation window” defined from −Lto L

Px = lim [( ) ∑ ∣x[n]∣2 ]
L
1
2L + 1
​ ​ ​ ​

L→∞
n=−L

In case of periodic signal, we defined these limits around one cycle of the period

1 N −1
∴ Pav =
​

N ​ ∑n=0 ∣x[n]∣2 
​

(1) Introduction 1
 Elementary Discrete-Time Signals
Unit sample or unit impulse

1 n=0
δ[n] = {
0 =0
n
​ ​

δ[n] = u[n] − u[n − 1]

Unit step

1 n≥0
u[n] = {
0 n<0
​ ​

n
u[n] = ∑ δ[k] ​

k=−∞

Running Summation

Periodic sequence

x[n] = A cos(ω0 n + ϕ) ​

x[n]is called periodic with fundamental period N , if — x[n + N] = x[n] ∀n

N is the number of samples per period
Condition of periodicity: NTs ​ = kT0 ​

For one cycle of a signal (T0 ), there must be integer numbers of samples (Ts )
​ ​

Complex exponential sequence

x[n] = Aejω0 n = A cos(w0 n) + A sin(ω0 n)

​

​ ​

Discrete-Time Systems
A discrete-time system is a computational process or algorithm that maps an input sequence x[n]into another output sequence
y[n].

System Properties
Causality
A system is called causal if the present value of output does NOT depend on future values of the input/output

if x[n] = 0for n < n0 —> for y[n] = 0when n < n0 

​ ​

A constant term in the output equation will make the system non-causal as the pervious condition has been violated

Stability
A system is said to be stable in the bounded-input bounded-output (BIBO) sense

if every bounded input signal results in a bounded output signal

∣x[n]∣ ≤ Mx < ∞​ → ∣y[n]∣ ≤ My < ∞

​ ∀n

(1) Introduction 2
 Linearity
H
x1 [n] → y1 [n]
​ ​

H
x2 [n] → y2 [n]
​ ​

H
ax1 [n] + bx2 [n] → ay1 [n] + by2 [n]
​ ​ ​ ​

Time-Invariance
Time-invariance means that the system (properties) does NOT change over time.

A time shift in the input results in a corresponding (i.e. same) time shift in the output.

y[n] = H{x[n]}
y[n − n0 ] = H{x[n − n0 ]}
​ ​

Block Diagrams & Signal Flow Graphs

For LTI systems:

Feedforward System Feedforward System

Output depends only on the inputs Output depends only on the inputs and pervious outputs

EX: y[n] = x[n] − x[n − 1] EX: y[n] = x[n] − x[n − 1] + y[n − 1]
FIR Filters IIR Filters

System Classifications
FIR Systems IIR Systems
The impulse response h[n] of an FIR system has a finite The impulse response h[n] of an IIR system has an infinite
number of nonzero samples. number of nonzero samples.

Always stable May or may not be stable

(1) Introduction 3
 LTI Systems
Impulse Response
replace each x[n]with δ[n]—&— each y[n]with h[n]

The output sequence y[n]of an LTI system can be computed using the linear convolution sum of its input sequence x[n]and
the system’s impulse response h[n].

x[n]has Lelements
h[n]has M elements
y[n]has L + M − 1elements

y[n] = x[n] ∗ h[n]

∞
y[n] = ∑ x[m]h[n − m]
​

m=−∞

∞
y[n] = ∑ h[m]x[n − m]
​

m=−∞

📌 This formulae acquire the system to be: LTI SYSTEM ONLY

System Causality
An LTI system with impulse response h[n]is causal if h[n] = 0for n < 0

System Stability
An LTI system with impulse response h[n]is stable in the BIBO sense —

∞
iff the impulse response is absolutely summable: (∑n=−∞ ∣h[n]∣) ​
< ∞

Convolution Steps Matrix Form

Change the axis Put matrix y[n]= matrix h[n]matrix x[n]

Shift h[n]and then mirror Put y[n]in a column matrix of (L + M − 1) × 1

Scan Put x[n]in a column matrix of size L × 1

Put h[n]in a matrix of size (L + M − 1) × L

Response of LTI Systems to Simple Test Sequences

Impulse Response:
H
x[n] = δ[n] → y[n] = h[n]

Step Response:

(1) Introduction 4
 H
x[n] = u[n] → y[n] = s[n]

Complex exponential sequence:

H
x[n] = z n → y[n] = z n H(Z)

z n : Eigen Function of the LTI system

H(z): is the system transfer function — Eigen value

LCCDE: Linear Constant-Coefficients Difference Equation

N M
y[n] = − ∑ ak y[n − k] + ∑ bk x[n − k]
​ ​ ​ ​

k=1 k=0

This is an IIR because it has feedback “Recursive system”

ak : feedback constant coefficients

​

bk : feedforward constant coefficients

​

N : order of the system

NOTE:

1. feedback summation starts from k = 1, while feedforward summation starts from k = 0
2. if ak ​ = 0—> NO FEEDBACK - it implies FIR system “Non-recursive system”

Section Notes
x[n] = [1 , 2, 3, 4]; this underset hyphen indicates the x[0]term
−
​

e.g.: if x[n] = [1, 2, 3 , 4]

​

∴ x[0] = 3and then x[1] = 4, x[−1] = 2& x[−2] = 1

if h[n]starts from 0 ≤ nthen the system is causal
Step response in terms of impulse response is running summation (LTI)
n
s[n] = ∑k=−∞ h[k] ​

h[n] = s[n] − s[n − 1]

(1) Introduction 5