SIGNALS AND SYSTEMS
Chapter 2: Continuous-Time Systems
(Added slides)
CLASSIFICATIONS: LINEAR AND NON-LINEAR
❖ Homogeneity Property
x(t) αx(t) αy(t)
α = const
❖ Adding Property
x1(t)
x1(t) + x2 (t)
x2 (t)
❖ A linear system satisfy the Superposition Property (including Homogeneity and
Adding properties)
CLASSIFICATIONS: LINEAR AND NON-LINEAR
Example Determine if the following system is linear
a. y(t) = x(t) +a
b. y(t) = tx(2t)
a. y(t) = x(t) +a
Homogeneity test
S x ( t ) = x ( t ) + a
y (t ) = ( x (t ) + a ) = x (t ) + a
S x ( t ) y ( t )
The system is non-linear in this case.
CLASSIFICATIONS: LINEAR AND NON-LINEAR
b. y(t) = tx(2t)
❖ Homogeneity test S x ( t ) = tx ( 2t )
y ( t ) = ( tx ( 2t ) ) = tx ( 2t )
S x ( t ) = y ( t )
System posses homogeneity test
❖ Additivity test S x1 ( t ) + x2 ( t ) = t ( x1 ( 2t ) + x2 ( 2t ) )
= tx1 ( 2t ) + tx2 ( 2t )
Let: y1 ( t ) = tx1 ( 2t )
y 2 ( t ) = tx2 ( 2t )
y = y1 ( t ) + y2 ( t ) = tx1 ( 2t ) + tx2 ( 2t )
S x1 ( t ) + x1 ( t ) = y1 ( t ) + y2 ( t )
System posses additivity test
Conclusion: The given system is linear!
CLASSIFICATIONS: TIME-VARYING V.S. TIME-INVARIANT
System Time shift y1(t)
x(t)
Time shift System y2(t)
If y1(t) equals to y2(t), then the system is time invariant
CLASSIFICATIONS: TIME-VARYING V.S. TIME-INVARIANT
Examples
Determine if the following system properties are valid.
CLASSIFICATIONS: MEMORY V.S. MEMORYLESS
Examples: determine if the systems has memory or not
a. y(t)= (t+5)x(t)
a. For signal y(t)= (t+5)x(t) -> The output only depends on “t” and the
current state of x(t) -> The system is memoryless
CLASSIFICATIONS: STABILITY
Examples: determine if the system is stable (BIBO) or not?
y(t)= 2x(t)
if x ( t ) B1 then y ( t ) B2
where B2 = 2 B1
Since y(t) will always be ≤ B2 = 2B1 for an input bounded by B1
The system is stable (BIBO)
Example: Consider an echo system (or a multipath system).
y ( t ) = 1 x ( t − 1 ) + 2 x ( t − 2 )
Determine if the system is stable (BIBO) or not?
❑ Suppose x(t) is bounded by a finite value M, or x ( t ) < M < , for all
times
Or the value of x(t) cannot exceed an envelope [ -M, M] at all times.
❑ Therefore:
y ( t ) 1 x ( t − 1 ) + 2 x ( t − 2 ) 1 + 2 M
So the corresponding output is bounded. The system is BIBO stable.
LTI: Convolution
Computing the Convolution Integral:
1. Analytically: If we have functions for 𝑥(𝑡) and ℎ(𝑡) then perform the
integration to get a function 𝑦(𝑡).
❖ Simplifications may be possible using special properties of the convolution
integral.
2. Graphically: If we have waveforms for 𝑥(𝑡) and ℎ(𝑡).
❖ Can be visually intuitive.
3. Numerically: If we have data sets (think data acquisition) for 𝑥(𝑡)
and ℎ(𝑡), then, replace the integral with a sum.
LTI: CONVOLUTION
Procedure for the graphical evaluation of the convolution integral:
1. Substitute x(t) and h(t) with x(τ) and h(τ), respectively.
2. Fold (form the mirror image of) x(τ) or h(τ) about the vertical
axis to obtain x(-τ) or h(-τ).
3. Slide x(-τ) or h(-τ) to the right a distance t to obtain x(t-τ) or
h(t-τ).
4. Multiply the two functions to obtain the product , x(t-τ) h(τ) or
x(τ) h(t-τ) .
5. Integrate this product by varying t from - to +.
Example 1
Compute y(t) = h (t) u(t) using the graphical evaluation.
y (t ) = h (t ) u (t ) = h ( )u (t − ) d = u (t − ) h ( )d
− −
t-1
y (t ) = h (t ) u (t ) = u (t − ) h ( )d
−
Signals when t=1
LTI: CONVOLUTION
t-1
y(t) =
Convolution for interval 1<t<2
y(t)
0, for t 0
2
t
− + t , for 0 t 1
Finally, we get:
y (t ) = 2
2
t
− 2t + 2, for 1 t 2
2
0, for t 2
LTI: CONVOLUTION
Example 2: Reflect-and-shift Convolution Evaluation
Evaluate the convolution integral y(t) = x(t) h(t).
h(t)
LTI: CONVOLUTION
1. Graph of x(τ) and h(t − τ): (see Fig. a)
x()
h(t-)
(a)
LTI: CONVOLUTION
Example 4
Graphical Convolution
• Revisiting the prior RC example
The impulse response of this circuit is:
−t
1 RC
h (t ) = e u ( t ) = e−t u ( t )
RC
• If we choose: 𝑅𝐶 = 1/2 = 𝜏𝑐 , hence,
• Input: 𝑣𝑖𝑛 𝑡 = 𝑢 𝑡 − 𝑢(𝑡 − 1)
h(t ) = 2e −2t u (t )
Graphical Convolution
• Analytical soln is:
• For, t < 1
t t t
vout (t ) = vin ( ) h(t − )d = (1) 2e −2(t − ) d = 2e −2t e 2 d
0 0 0
=e −2 t e 2 t = e −2t ( e 2t − 1) = 1(1−−ee−2−t2t ) ( u (t ) − u (t − 1) )
0
Graphical Convolution
• For, t > 1
1 t
vout (t ) = vin ( ) h(t − )d + vin ( ) h(t − )d
0 1
1
= h(t − )d + 0 =e −2 t e 2 1
0
0
= e −2t ( e 2 − 1) = (ee−2(−2(t −t1)−1)−−ee−2−t2t ) u (t − 1)
