Vous êtes sur la page 1sur 7

INEL 5309-086 - DIGITAL SIGNAL PROCESSING I

EXAM ONE: SOLUTIONS October 1, 2012


Prof. Domingo Rodrguez

Student:____________________________________I.D. No.___________
Problem One: Ideal Filters and Sinusoidal Input Signals
Let x[ n] cos(0.25 n) cos(0.5 n) cos(0.75 n) be the input to a to an ideal
band-pass filter H B ( ) with lower cut-off frequency and higher cut-off

L 104 and H 105 , respectively.


(10 points) (1a) Plot X ( ) and H B ( ) , the DTFT of x[n] and hB [ n] ,
frequency parameters given by

respectively.
1a.- The spectrum of the input signal is given in the diagram below:

The ideal band-pass filter(not the same scale) is given in the diagram below:

INEL 5309-086 - DIGITAL SIGNAL PROCESSING I


EXAM ONE: SOLUTIONS October 1, 2012
Prof. Domingo Rodrguez

Student:____________________________________I.D. No.___________
(15 points) (1b) Find the output

y[n] by taking the inverse Fourier

transform of the product of these transforms. That is, use the following
formula:


1
1
jn
X H e jn d .
y[n]
Y e d

2
2

Remark: DTFT cos b n b b

1b.- Multiplying these transforms results in the following product:

Y ( ) X ( ) H ( ) H ( ) 0.25k 0.25k
k 1

From the characteristics of the ideal band-pass filter given above, we get:

Y ( )

0.25k 0.25k

k 2

Using the expression given in the Remark above, we obtain:

y[n] cos(0.5 n)

INEL 5309-086 - DIGITAL SIGNAL PROCESSING I


EXAM ONE: SOLUTIONS October 1, 2012
Prof. Domingo Rodrguez

Student:____________________________________I.D. No.___________
(25 points) Problem Two: Linear Phase Ideal Filter

Obtain the impulse response of an ideal low-pass filter which has linear
phase and whose frequency response is:

e jnd ,
H L
0,

m
m

Solution:

e jnd ,
H L
0,

m
m

INEL 5309-086 - DIGITAL SIGNAL PROCESSING I


EXAM ONE: SOLUTIONS October 1, 2012
Prof. Domingo Rodrguez

Student:____________________________________I.D. No.___________
Problem Three: Matrix Representation of Linear Conv.
Let x[n] , n Z , represent an input to a filter T with impulse response
function h[n] , n Z . For x[ n] [n] 2 [n 1] , perform the following tasks:
(3a.- 15 Points) Compute the output y[ n] using matrix-vector notation if
the filter is an FIR filter with impulse response h[ n] 2 [ n] 2 [n 2] . Draw
the filter as a tap-delay line (non-recursive) filter.
Remark: y[ n]

k 1

k 2

k 0

k 0

x[k ]h[n k ] h[k ]x[n k ];

n Z4

Solution:
We start by using the following equation:

y[n]

k 1

x[k ]h[n k ] x[0]h[n] x[1]h[n 1];

k 0

n Z4

The output is the computed using a matrix-vector multiplication operation:

y[0] h[0] 0
2 0
2
y[1] h[1] h[0] x[0] 0 2 1 4

y[2] h[2] h[1] x[1] 2 0 2 2


y[3] 0 h[2]
0 2
4

The diagram below represents the FIR filter:

INEL 5309-086 - DIGITAL SIGNAL PROCESSING I


EXAM ONE: SOLUTIONS October 1, 2012
Prof. Domingo Rodrguez

Student:____________________________________I.D. No.___________

Observation for Part 3a: We can use direct substitution into the general
linear convolution equation in order to get the desired result for Problem
3a:

y[n]
y[n]

k
k

x[k ]h[n k ]

[k ] 2 [k 1] 2 [n k ] 2 [n k 2];

[k ] 2 [n k ] 2 [n k 2]

nZ

2 [k 1] 2 [n k ] 2 [n k 2]

The first summation expression in the last equation is only valid or non-zero
when k 0 and the second summation expression is only valid or non-zero
when k 1 . Thus, we get the following result:

y[n] (2 [n] 2 [n 2]) 2(2 [n 1] 2 [n 3]); n Z


y[n] 2 [n] 4 [n 1] 2 [n 2] 4 [n 3]; n Z

(3b.- 10 Points) Compute the output y[ n] if the filter is an IIR filter with
impulse response h[ n] ba u[ n] . Draw the filter as a first-order recursive
filter.
Solution:
To solve this problem, we proceed in the following manner:
n

y[n] T x[n] T [n] 2 [n 1] T [n] 2T [n 1] h[n] 2h[n 1]


Using the given function h[ n] , we obtain
y[n] ba nu[n] 2ba n 1u[n 1] ba nu[n] 2ab a nu[n 1 ba n u[n] a2 u[n 1]

INEL 5309-086 - DIGITAL SIGNAL PROCESSING I


EXAM ONE: SOLUTIONS October 1, 2012
Prof. Domingo Rodrguez

Student:____________________________________I.D. No.___________
Problem Four: Linearity and Time-Invariance
(25 Points) Determine if the following system, T defined by the equation

y[n] T x[n]

x[k ] , is a filter. Show all works.

The system y[ n] T x[ n]

x[k ] is called a discrete accumulator system

since its output at the present time n is equal to the sum of the present
input sample value plus all previous input samples. A discrete system T is a
discrete filter if the system is, both, linear and time-invariant. We proceed to
check for linearity and time-invariance conditions:
Linearity Condition - A discrete system y[ n] T x[ n] is linear if it satisfies
the following condition:

T ax1[n] bx2 [n] aT x1[n] bT x2 [n]

We first compute the r.h.s. of the linearity condition:


Let

T x1[n]

T x2 [n]

k n

x1[k ] ; and let

k n

x2[k ].

Then, we have

aT x1[n] bT x2[n] a

k n

x1[k ] b

k n

x2[k ] .

We then proceed to compute the l.h.s. of the linearity condition:


Let T x3[ n] T ax1[ n] bx2 [ n]
Expanding the sum, we obtain:

T ax1[n] bx2 [n]

k n

x3[k ]

k n

ax1[n] bx2[n] .

k n

k n

ax1[k ] bx2[k ] a

x1[k ] b

k n

x2[k ] ;

Comparing the r.h.s. and l.h.s. of the linearity condition, we determine


that the system T is a discrete linear system.

INEL 5309-086 - DIGITAL SIGNAL PROCESSING I


EXAM ONE: SOLUTIONS October 1, 2012
Prof. Domingo Rodrguez

Student:____________________________________I.D. No.___________
Time-Invariance Condition - A discrete system y[ n] T x[ n] is said to be
time-invariant if it satisfies the following condition:

T x[n n0 ] y[n n0 ] .
We first compute the r.h.s. of the time-invariance condition:
Let T x[ n] y[n]

k n

x[k ] ; then, y[n n0 ]

k n n0

x[k ] .

Second, we compute the l.h.s. of the time-invariance condition:


Let T x[ n n0 ] T g[ n]
change of variables: Let
k n

x[k n0 ]

m n n0

k n

g[ k ]

k n

x[k n0 ] . We proceed to make a

m k n0 ; then, k m n0 and we have

x[m]

k n n0

x[k ]

Comparing the r.h.s. and l.h.s. of the time-invariance condition, we


determine that the system T is a discrete time-invariant system. Since
the system is also linear, we conclude that the system T is a discrete filter.

Vous aimerez peut-être aussi