ECE 4371, Fall, 2009

Zhu Han
Department of Electrical and Computer Engineering
Class 7 Sep. 15th, 2009

Overview
  

Homework Hint Signal to Noise Ratio AM with noise

Coherent decoder Non-coherent decoder

FM with noise
Analysis Threshold effect Preemphasis and deemphasis

Signal to Noise Ratio

Channel model: additive white Gaussian noise (AWGN) Receiver model: a band-pass filer followed by an ideal demodulator Receiver model

 

Idealized characteristic of band-pass filtered noise. The baseband transmission model, assuming a message signal of bandwidth W, used for calculating the channel signal-to-noise ratio.

The PSD of w(t) is denoted by N 0 2

SNR
The iltered noise in narro band noise representation : n(t ) ! nI (t ) cos( 2T f c t )  nQ (t ) sin( 2T f c t ) The iltered signal or demodulation is x(t ) ! s (t )  n(t ) The channel signal - to - noise ratio average po er o s (t ) average po er o n(t ) The output signal - to - noise ratio ( )
!

(2.79) (2.80)

average po er o the demodulated signal average po er o noise at the output ( )O igure o merit ! (2.81) ) ( )O !

Noise in linear receiver using coherent detection



Model of DSB-SC receiver using coherent detection

s ( t ) ! CA c cos( 2 T f c t ) m ( t ) P !

W

S M ( f ) df !
2 C 2 Ac P

(2.83) 2 (2.84)

(SNR)

W 0 2 C 2 Ac P (baseband) ! 2W 0 C : system dependent scaling factor

C, DSB

Noise in linear receiver using coherent detection

x ( t ) ! s (t )  n (t )

!
1 2 1  ? 2

cos( 2T f c t ) (t )  n I (t ) cos( 2T f c t )  nQ (t ) sin( 2T f c t ) (2.85) 1 n I (t ) 2

v (t ) ! x (t )cos ( 2T f c t )

(t ) 

1 (t )  n I (t ) Acos( 4T f c t )  nQ (t ) sin( 4T f c t ) 2 high requency components

c

1 Lo - pass ilter y (t ) ! 2 (2.86) indicates :

1 ( t )  n I (t ) 2

(2.86)

1. (t ) and n I (t ) are additive at the receiver output. 2. nQ (t ) is completely rejected by the coherent detector.
59

Noise in linear receiver using coherent detection

1 2 The average output signal ( CAc m (t ) ) power ! C 2 Ac P 4 2 et BT ! 2W 1 1 2 The average noise ( n I (t ) ) power ! ( ) 2W 2 2 2 2 C 2 Ac P C 2 Ac P 4! (SNR ) , DSB SC ! 2W 0 W 0 2 (SNR) !1 (SNR) C DSB -SC 1 W 0 ! 2
0

(2.87)

(2.88)

1. Coherent SSB has the same figure of merit of DSB - SC 2. No trade - off between performanc e and bandwidth. Serious problem!
60

Noise in AM receiver using envelope detection



Model of AM receiver

s (t ) ! Ac ?  k a m(t )Acos(2T f ct ) 1 ! Ac cos(2T f c t )  Ac k a m(t ) cos(2T f c t )

) c (1  k a ( ) , ! 2WN 0 t the output of the filter : x(t ) ! s (t )  n(t )
2 2

(2.89)

(2.90)

!? c 

c a

(t )  nI (t )Acos(2T f c t )  nQ (t ) sin(2T f c t ) (2.91)

Noise in AM receiver using envelope detection



(a) Phasor diagram for AM wave plus narrowband noise for the case of high carrier-to-noise ratio. (b) Phasor diagram for AM wave plus narrowband noise for the case of low carrier-to-noise ratio.

Noise in AM receiver using envelope detection

y (t ) ! envelope of x (t ) !

_A  A k m(t )  n (t )A  n ?
2 c c a I
c c

(t )

(2.92)

ssume y (t ) !
2

 k

c a

(t )  nI (t ) "" nQ (t )

c a

(t )  nI (t )

1.

2 2. k a

"" WN 0 1 )O, )O ) k P } c a 2WN 0 2 ka P } 2 1  ka P

2 2

( ( (

(2.94) (2.95)
62

Waste energy

63

Threshold Effect


Output signal-to-noise ratio of an envelope detector for varying carrier-to-noise ratio.

System Model and Noise Model

Discriminator consists of a slope network and an envelope detector.

Signal after bandpass filter



The incoming FM signal s(t) is defined by

At the bandpass filter output

Discriminator Output


Note that the envelope of x(t) is of no interest to us (limiter)

Because Ac "" r (t ) r (t ) U (t ) } J (t )  sin ? (t )  J (t ) A ] ( 2.138 ) Ac t r (t ) ! 2T k f m(X )dX  ] sin ? (t )  J (t )A(2.139) 0 Ac The discrimina tor output is ( ig 2.40)

1 d U (t ) v (t ) ! 2T dt ! k f ( t )  n d (t )
message additive noise

(2.140)

where 1 d n d (t ) ! _r (t ) sin?] (t )  J (t ) Aa (2.141) 2T Ac dt

Noise After Discriminator

Assume ] (t )  J (t ) is uniformly distribute d over (0, 2T ), then nd (t ) is independen t of message signal. We may simplify nd (t ) as 1 d _r (t ) sin ? (t ) Aa (2.142) ] n d (t ) } 2T Ac dt

rom definition of r (t ) and ] (t ) , e have nQ (t ) ! r (t ) sin ? (t ) A ]

1 n d (t ) } 2T dn Q (t )
c

(2.143)
(2.144)

dt

The quadrature appears

Noise After Discriminator cont.



The average output signal power = kf2P

Recall

d F .T j 2Tf dt

n (t)

S NQ ( f )

1 d 2T Ac dt

nd(t)

SNd ( f )

S Nd ( f ) !

2 2 c

S NQ ( f )

(2.145)

Noise After Discriminator cont.



Assume that n (t) has ideal low-pass characteristic with bandwidth BT 2

S Nd ( f ) ! If BT N0 f
2 c

f e

BT 2

(2.146)

" W 2 At the receiver output S 0(f)!

0 f Ac2 2

f eW

(2.147)

SNR of FM
verage po er of n0 (t ) ! N0
2 c

W
W

f 2 df (2.148)

2 N 0W 3 ! 3 c2 w
( )O ,
!

1
c 2

noise quieting effect

(2.149)

2 c

k2P f
3

2 N 0W Ac2 The average power of s (t ) is , 2 the average noise power in message bandwidth is W (SNR ) C , FM

Ac2 ! 2W
!

(2.150)
0

(SNR ) (SNR ) C

3k 2 P f W
2

(2.151) Bandwidth effect

FM

71

Single Tone FM SNR

(f cos 2T f ct  sin( 2T f t ) c f t (f We may write , 2T k f m(X )dX ! sin( 2T f m t ) 0 fm

s (t ) !

d (f both side m(t ) ! cos( 2T f m t ) dt kf ( (f ) 2 The average po er of (t ) (across 1; load) is P ! 2k 2 f 2 2 2 2 3 A ( (f ) 3A F From (2.149), (SNR ) , FM ! c ! c 4 0W 3 4 0W

( S )O 1 , ! compare to (from xample 2.4) ( ) AM 3 3 2 1 When F " , FM has better performanc e. 2 3 2 F" ! 0.471 3 Define F ! 0.5 as the transitio n between narrowband FM and wideband FM.

(SNR ) (SNR )C

!
FM

3 (f 2 3 2 ( ) ! F 2 W 2

(2.152)

FM Threshold Effect

When there is no signal, i.e., carrier is unmodulate d. The composite signal at the frequency discrimina tor input x (t ) ! ?Ac  nI (t ) Acos( 2T f c t )  n (t ) sin( 2T fct ) (2.153) Ac  nI (t )

U (t ) ! tan

1

n (t )

Occasional ly, P1 may sweep around the origin , U (t) increases or decreases 2T U ' (t ) The discrimina tor output is equal to 2T P
1

r(t)

n (t) 0

x(t)

U
nI(t)

Ac(t)

P2

Example


Illustrating impulse ike components in Ud (t) ! dU(t)/dt produced by changes of 2T in U (t); (a) and (b) are graphs of U (t) and Ud (t), respectively.

Threshold Effect


Dependence of output signalto-noise ratio on input carrierto-noise ratio for FM receiver. In curve I, the average output noise power is calculated assuming an unmodulated carrier. In curve II, the average output noise power is calculated assuming a sinusoidally modulated carrier. Both curves I and II are calculated from theory.

FM Threshold Reduction (tracking filter)

 

FM threshold extension FM demodulator with negative feedback

FM Preemphsis and Deemphasis

Figure 2.48 (a) Power spectral density of noise at FM receiver output. (b) Power spectral density of a typical message signal.

EE 541/451 Fall 2007

Improvement Factor
1 H de ( f ) ! , W e f e W H pe ( f ) The PSD at the discriminator output is f2 S d(f)! 02 , Ac H de ( f ) S d ( f ) !
2

(2.156)

f
0

BT e 2 f2
2

(2.146)
2

Ac Average output noise 0 ! 2 power with de - emphsis A c The improvement factor I is I! 2W 3 3

w -w

H de ( f )

, f e
2

BT 2
2

(2.157)

W

f H pe ( f ) df (2.158)

f 2 Hpe ( f ) df
80

(2.162)

FM Preemphsis and Deemphasis

Comparison of the noise performance of various CW modulation systems. Curve I: Full AM, Q = 1. Curve II: DSB-SC, SSB. DSBCurve III: FM, F = 2. Curve IV: FM, F = 5. (Curves III and IV include 1313-dB pre-emphasis, depredeemphasis improvement..)

Bandwidth Tradeoff
In making the comparison, it is informative to keep in mind the transmission bandwidth requirement of the modulation systems in question. Therefore, we define normalized transmission bandwidth as
n

Table 2.4 Values of Bn for various CW modulation schemes FM AM, DSB-SC Bn  SSB  F!  F!