Notes
Notes
Notes
Some definitions:
Classification
T0
Causal Signal:
x(t) = 0 for t < 0
Anticausal: x(t) = 0 for t >= 0
Deterministic or Random
Deterministic can be predicted
Random cannot be predicted
P (t) = i(t)v(t)
t2
E = ∫ P (t)dt
t1
2
P (t) = |x(t)|
t2
2
E = ∫ |x(t)| dt
t1
E
P avg =
t2 − t1
If energy is finite, power is zero and is called energy signal.
Eg.
E ∞ = 1, P ∞ = 0
Eg.
2. x[n] = 2 for all n
P ∞ = 4, E ∞ = ∞
We can have neither power nor energy signal also (Ramp signal).
Eg.
3. x(t) = t
E ∞ = ∞, P ∞ = ∞
Signal Transformation
Time shifting
T units delayed
T units advanced
Eg.
Time Scaling
CT
DT:
x (at) squishes the signal, but they are losing information. For eg, x [3 n]: x[1], x[2], x[4], x[5], x[6],
etc. are thrown away.
x (at) stretches the signal. For eg, x[n/2] would stretch the signal.
However, what do we do for intermediate values? We Interpolate the signal.
Amplitude Scaling:
Elementary Signals
1. Exponential
Āe
s̄t
= Ae
jϕ
e
(σ+jω)t
Depending on sign of σ:
Anything in the left-half plane is decaying exponential
Anything on imaginary axis is neither growing nor decaying, i.e. it is pure sinusoid
Anything on right half plane is growing exponential
δ(x) = δ(−x)
Sifting property: ∫
∞
f (t)δ(t − T )dt = f (T )
−∞
r(t) = t u(t)
2
du(t) d r(t)
δ(t) = = 2
dt dt
System
x (t) y(t)
V s = x(t)
Vs − Vc dV c
i = = C
R dt
dV c Vc Vs
+ =
dt RC RC
Eg. Cart: This is similar to cart with friction: x(t) is the external force, a y(t) corresponds to friction.
dy(t)
+ ay(t) = bx(t)
dt
x[n] y[n]
n = 0, ±1, ±2, . . .
Eg. For the same cart, it can be represented as DT also, where time would be like Δn.
becomes
dy(t)
+ ay(t) = bx(t)
dt
Imp
If we come up with methods to solve one class of system, we can do same thing for other systems
in same class.
Classification of systems
Interconnection of subsystems, devices, to transform input into output
CT and DT systems
Systems with and without memory
Without memory:
Resistor system: v(t) = i (t) R
Output depends only on present value
With memory:
Depends on present/future
It "stores" energy!
t
summer: y[n] = ∑ n
k=−∞
x[k] -> computes running average
Invertible and non-invertible
Invertible: If there exists S2 for S1, such that x(t) -> S1 -> y(t) -> S2 -> x(t), S1 is said to be
invertible
Eg. y(t) = 2 x(t)
Non-invertible
Eg. y(t) = 0
Causal and Non-causal system (not signal)
Causal:
Output at any time depends on values from present or past times not future
Eg: y(t) = x(t) cos (t + 1)
Non-causal
Eg. y[n] = x[n] - x[n + 1]
Stability
Stable system:
Small input leads to responses which do not diverge
If |x(t)| < B < ∞ implies |y(t)| < B < ∞
x y
x(t) -> S -> Delay -> y(t - T0 ) = x(t) -> Delay -> x(t - T0 ) -> S -> y(t - T0 )
Eg. y(t) = sin (x(t))
No explicit time dependence
Time variant
Eg. y(t) = n (x(t))
Explicit Time dependence
Linearity
Linear system:
a x1(t) + b x2(t) -> S -> a y1 (t) + b y2(t)
Eg. y(t) = t x(t)
Non-linear
Eg. y(t) = m x(t) + c
11th May
Convolution of two functions:
∞
x[n] ∗ y[n] = ∑ x[k]y[n − k]
k=−∞
Imagine a DT system that is linear
Sifting property
∑ x[k]δ[n − k] = x[n]
−∞
∞ ∞
x[n] = ∑ −∞ x[k]δ[k] → ∑ −∞ x[k]h[n − k] = y[n]
−∞
x[n]h[n − k]
CT LTI system
Wow!
We can give an unit impulse, see what it does, then convolve x[n] to figure out y[n]!
δ Δ (t − k 0 ) → h kΔ (t)
∞
x(t) = ∑ −∞ x(kΔ)δ Δ (t − k 0 )Δ
Using Linearity:
∞ ∞
∑ ∞ x(kΔ)δ Δ (t − k) → ∑ ∞ x(kΔ)h kΔ (t)
17th May
Properties:
Commutative property
∞ ∞
y[n] = ∑ −∞ x[k]x[n − k] = ∑ −∞ x[n − k]x[k]
- Distributive property
Proof: y[n] = ∑ ∞
−∞
x[k](h 1 [n − k] + h 2 [n − k]) == ∑
∞
−∞
x[k]h 1 [n − k] + ∑
∞
−∞
x[k]h 2 [n − k] = y 1 [n] + y 2 [n]
-Associative property
Using commutativity,
If we're passing a signal x(t) through many LTI systems, order doesn't matter.
h[n] = Aδ[n]
∞
y[n] = ∑ x[k]Aδ[n − k] = Ax[n]
−∞
-Invertibility
-Causality
Reason-> Unit impulse function is 0 for n < 0, so as it is causal, unit impulse response is also 0 for n < 0
-Stability
∞ ∞
y[n] = ∑ x[k]h[n − k] < B ∑ |h[k]|
−∞ ∞
∑ |h[k| < ∞
−∞
18th May
Unit Step Response:
= h(t) u(t)
=∑
∞
h[k]u[n − k]
−∞
=∑
n
−∞
h[k]
∞ t
s(t) = ∫ u(t − τ )h(τ )dτ = ∫ h(τ )dτ
−∞ −∞
ds(t)
h =
dt
19th May
FOURIER LESGOO
Lesgoo
CT system
x(t) = e
st
st
DT system
x[n] = z
n
∞ n −k n ∞ −k
y[n] = ∑ h[k]z z = z ∑ h[k]z
−∞ −∞
z
n
is the Eigenfunction of system
∞ sk t ∞ sk
x(t) = sum −∞ a k e → sum −∞ a k H (s k )e
ϕk = e
jkw o t
−∞
ak e
jw o t
a −k = a k
∗
∗
a k = a −k