Chapter 7 - Optimal Receiver Design PDF

Digital Communications
Instructor: Dr. Phan Van Ca

Lecture 06: Optimal Receiver Design
Kyung Hee
Modulation
University
We want to modulate digital data using signal sets which are :
z bandwidth efficient
z energy efficient
A signal space representation is a convenient form for

viewing modulation which allows us to:
z design energy and bandwidth efficient signal constellations

z determine the form of the optimal receiver for a given
constellation
z evaluate the performance of a modulation type
Young S. Kim Wireless Communication Systems Lab. - 2 -

Kyung Hee
University
Digital Communications
Spring 2006
Instructor: Young Soo Kim

Lecture #6 : Optimal Receiver Design
Modulation
We want to modulate digital data using signal sets which are :
z bandwidth efficient
z energy efficient
A signal space representation is a convenient form for

viewing modulation which allows us to:
z design energy and bandwidth efficient signal constellations

z determine the form of the optimal receiver for a given
constellation
z evaluate the performance of a modulation type
Problem Statement
We transmit a signal s(t ) ∈{s1(t ), s2 (t ),…, sM (t )} , where s(t)

is nonzero only on t ∈[0,T] .
Let the various signals be transmitted with probability:

p1 = Pr[ s1(t )],…, p M = Pr[ s M ( t )]
The received signal is corrupted by noise:

r(t) = s(t) + n(t)
Given r(t) , the receiver forms an estimate s(t) of the signal

s(t ) with the goal of minimizing symbol error probability
Ps = Pr[ s(t ) ≠ s(t )]
Kyung Hee
Noise Model
University
The signal is corrupted by Additive White Gaussian Noise

(AWGN) n(t)
N
The noise n(t) has autocorrelation φnn ( τ) = 0 δ( τ) and
2
power spectral density Φnn ( f ) = N0 2
Any linear function of n(t) will be a Gaussian random

variable
Channel
s(t ) Σ r (t )
n( t )
Kyung Hee
Problem Statement
University
We transmit a signal s(t ) ∈{s1(t ), s2 (t ),…, sM (t )} , where s(t)

is nonzero only on t ∈[0,T] .
Let the various signals be transmitted with probability:

p1 = Pr[ s1(t )],…, p M = Pr[ s M ( t )]
The received signal is corrupted by noise:

r(t) = s(t) + n(t)
Given r(t) , the receiver forms an estimate s(t) of the signal

s(t ) with the goal of minimizing symbol error probability
Ps = Pr[ s(t ) ≠ s(t )]
Noise Model
The signal is corrupted by Additive White Gaussian Noise

(AWGN) n(t)
N
The noise n(t) has autocorrelation φnn ( τ) = 0 δ( τ) and
2
power spectral density Φnn ( f ) = N0 2
Any linear function of n(t) will be a Gaussian random

variable
Channel
s(t ) Σ r (t )
n( t )
Signal Space Representation
The transmitted signal can be represented as:

K
sm ( t ) = ∑ sm, k f k ( t ) ,
k =1
T
where sm,k = ∫ sm (t ) f k (t )dt .
0
K
The noise can be respresented as : n(t ) = n′(t ) + ∑ nk f k (t )
k =1
T
where nk = ∫ n(t ) f k (t )dt
0
K
and n ′ (t ) = n( t ) − ∑ nk f k ( t )
k =1
Kyung Hee
Signal Space Representation (continued)
University
The received signal can be represented as :

K K K
r ( t ) = ∑ sm, k f k ( t ) + ∑ nk f k ( t ) + n ′ ( t ) = ∑ rk f k ( t ) + n ′ ( t )
k =1 k =1 k =1
where rk = sm, k + nk

Kyung Hee
Signal Space Representation
University
The transmitted signal can be represented as:

K
sm ( t ) = ∑ sm, k f k ( t ) ,
k =1
T
where sm,k = ∫ sm (t ) f k (t )dt .
0
K
The noise can be respresented as : n(t ) = n′(t ) + ∑ nk f k (t )
k =1
T
where nk = ∫ n(t ) f k (t )dt
0
K
and n ′ (t ) = n( t ) − ∑ nk f k ( t )
k =1
Signal Space Representation (continued)
The received signal can be represented as :

K K K
r ( t ) = ∑ sm, k f k ( t ) + ∑ nk f k ( t ) + n ′ ( t ) = ∑ rk f k ( t ) + n ′ ( t )
k =1 k =1 k =1
where rk = sm, k + nk
The Orthogonal Noise: n ′( t )
The noise n′(t) can be disregarded by the receiver
T T ⎛ K ⎞
∫ m
s ( t ) n ′ ( t ) dt = ∫ m
s ( t ) ⎜ n( t ) − ∑ n f ( t )
k k ⎟ dt
0 0 ⎝ k =1 ⎠
T K ⎛ K ⎞
= ∫ ∑ sm,k f k (t )⎜ n( t ) − ∑ nk f k (t )⎟ dt
0 k =1 ⎝ k =1 ⎠
K T K T
= ∑ sm, k ∫ f k (t )n(t )dt − ∑ sm,k nk ∫ f k2 (t )dt
k =1 0 k =1 0
K K
= ∑ sm,k nk − ∑ sm,k nk = 0
k =1 k =1
We can reduce the decision to a finite

Kyung Hee
University dimensional space!
We transmit a K dimensional signal vector:
s = [ s1, s2 ,…, sK ] ∈{s1,…, s M }
We receive a vector r = [r1,…, rK ] = s + n which is the sum of

the signal vector and noise vector n = [n1,…, nK ]
Given r , we wish to form an estimate s of the transmitted

signal vector which minimizes Ps = Pr[s ≠ s]
Channel
Receiver
s Σ r s
n
Kyung Hee
The Orthogonal Noise: n ′( t )
University
The noise n′(t) can be disregarded by the receiver
T T ⎛ K ⎞
∫ m
s ( t ) n ′ ( t ) dt = ∫ m
s ( t ) ⎜ n( t ) − ∑ n f ( t )
k k ⎟ dt
0 0 ⎝ k =1 ⎠
T K ⎛ K ⎞
= ∫ ∑ sm,k f k (t )⎜ n( t ) − ∑ nk f k (t )⎟ dt
0 k =1 ⎝ k =1 ⎠
K T K T
= ∑ sm, k ∫ f k (t )n(t )dt − ∑ sm,k nk ∫ f k2 (t )dt
k =1 0 k =1 0
K K
= ∑ sm,k nk − ∑ sm,k nk = 0
k =1 k =1
We can reduce the decision to a finite

dimensional space!
We transmit a K dimensional signal vector:
s = [ s1, s2 ,…, sK ] ∈{s1,…, s M }
We receive a vector r = [r1,…, rK ] = s + n which is the sum of

the signal vector and noise vector n = [n1,…, nK ]
Given r , we wish to form an estimate s of the transmitted

signal vector which minimizes Ps = Pr[s ≠ s]
Channel
Receiver
s Σ r s
n
MAP (Maximum a posteriori probability)
Decision Rule
Suppose that signal vectors {s1 , … , s M } are transmitted
with probabilities { p1,… , pm } respectively, and the signal
vector r is received
We minimize symbol error probability by choosing the

signal sm which satisfies : Pr (sm r ) ≥ Pr (si r ), ∀m ≠ i
p( r sm ) Pr (sm ) p(r si ) Pr (si )

Equivalently : ≥ , ∀m ≠ i
p( r ) p( r )
or p( r sm ) Pr (sm ) ≥ p(r si ) Pr (si ), ∀m ≠ i
Kyung Hee
Maximum Likelihood (ML) Decision Rule
University
If p1 = = pm or the a priori probabilities are unknown,

then the MAP rule simplifies to the ML Rule

signal s m which satisfies p(r sm) ≥ p(r si ), ∀m ≠ i

MAP (Maximum a posteriori probability)
Kyung Hee
University Decision Rule
Suppose that signal vectors {s1 , … , s M } are transmitted
with probabilities { p1,… , pm } respectively, and the signal
vector r is received

signal sm which satisfies : Pr (sm r ) ≥ Pr (si r ), ∀m ≠ i
p( r sm ) Pr (sm ) p(r si ) Pr (si )

Equivalently : ≥ , ∀m ≠ i
p( r ) p( r )
or p( r sm ) Pr (sm ) ≥ p(r si ) Pr (si ), ∀m ≠ i
Maximum Likelihood (ML) Decision Rule
If p1 = = pm or the a priori probabilities are unknown,

then the MAP rule simplifies to the ML Rule

signal s m which satisfies p(r sm) ≥ p(r si ), ∀m ≠ i
Evaluation of Probabilities
In order to apply either the MAP or ML rules, we need to

evaluate: p(rsm)
Since r = sm + n where sm is constant, it is equivalent to

evaluate : p(n) = p(n1,…, nk )
n(t) is a Gaussian random process

T
z Therefore nk = ∫ n(t ) f k (t )dt is a Gaussian random variable
0
z Therefore p(n1,…, nK ) will be a Gaussian p.d.f.
Kyung Hee
The Noise p.d.f
University
⎡T T ⎤
E [ni ⋅ nk ] = E ⎢ ∫ n( t ) fi ( t )dt ⋅ ∫ n( s) f k ( s)ds⎥
⎣0 0 ⎦
⎡T T ⎤ TT
= E ⎢ ∫ ∫ n(t )n( s) fi ( t ) f k ( s)dsdt ⎥ = ∫ ∫ E [n( t )n( s)] f i ( t ) f k ( s)dsdt
⎣0 0 ⎦ 00
TT TT N
= ∫ ∫ φnn (t − s) f i (t ) f k ( s)dsdt = ∫ ∫ 0 δ( t − s) f ( t ) f ( s)dsdt
i k
00 00 2
0 f ( t ) f ( t )dt = ⎧ N 0 2 , i = k
TN
=∫ i k ⎨
0 2 ⎩ 0, i≠k

Kyung Hee
Evaluation of Probabilities
University
In order to apply either the MAP or ML rules, we need to

evaluate: p(rsm)
Since r = sm + n where sm is constant, it is equivalent to

evaluate : p(n) = p(n1,…, nk )
n(t) is a Gaussian random process

T
z Therefore nk = ∫ n(t ) f k (t )dt is a Gaussian random variable
0
z Therefore p(n1,…, nK ) will be a Gaussian p.d.f.
The Noise p.d.f
⎡T T ⎤
E [ni ⋅ nk ] = E ⎢ ∫ n( t ) fi ( t )dt ⋅ ∫ n( s) f k ( s)ds⎥
⎣0 0 ⎦
⎡T T ⎤ TT
= E ⎢ ∫ ∫ n(t )n( s) fi ( t ) f k ( s)dsdt ⎥ = ∫ ∫ E [n( t )n( s)] f i ( t ) f k ( s)dsdt
⎣0 0 ⎦ 00
TT TT N
= ∫ ∫ φnn (t − s) f i (t ) f k ( s)dsdt = ∫ ∫ 0 δ( t − s) f ( t ) f ( s)dsdt
i k
00 00 2
0 f ( t ) f ( t )dt = ⎧ N 0 2 , i = k
TN
=∫ i k ⎨
0 2 ⎩ 0, i≠k
The Noise p.d.f (continued)
Since E[ni ⋅ nk ] = 0,∀i ≠ k , individual noise components are

uncorrelated (and therefore independent)

.
[ ]
Since E nk2 = N0 2, each noise component has a variance of N0 2
p(n1,…, nK ) = p(n1) p(nK )
( )
K 1
=∏ exp − nk 2 N0
k =1 πN0
−K 2 ⎛ K 2 ⎞
= ( πN0) exp⎜ − ∑nk N0⎟
⎝ k =1 ⎠
Kyung Hee
Conditional pdf of Received Signal
University
Transmitted signal values in each dimension represent the

mean values for each signal
⎛ K ⎞
p(r sm ) = ( πN 0 ) −K 2
(
exp⎜ − ∑ rk − sm, k
⎝ k =1
2
N0⎟
⎠
)

Kyung Hee
The Noise p.d.f (continued)
University
Since E[ni ⋅ nk ] = 0,∀i ≠ k , individual noise components are

uncorrelated (and therefore independent)

.
[ ]
Since E nk2 = N0 2, each noise component has a variance of N0 2
p(n1,…, nK ) = p(n1) p(nK )
( )
K 1
=∏ exp − nk 2 N0
k =1 πN0
−K 2 ⎛ K 2 ⎞
= ( πN0) exp⎜ − ∑nk N0⎟
⎝ k =1 ⎠
Conditional pdf of Received Signal
Transmitted signal values in each dimension represent the

mean values for each signal
⎛ K ⎞
p(r sm ) = ( πN 0 ) −K 2
(
⎝ k =1
2
N0⎟
⎠
)
Structure of Optimum Receiver
MAP rule : s = arg max pm ⋅ p( r sm )

{s1,…,s M }
⎛ K ⎞
s = arg max m ( 0 )
p ⋅ πN −K 2
exp
⎝
−
⎜ ∑ k m, k
r − s
2
N (
0⎟
⎠
)
{s ,…,s }
1 M k =1
⎡ ⎛ K ⎞⎤
s = arg max ln ⎢ pm ⋅ ( πN 0 )
−K 2
⎝ k =1
(2
N0⎟ ⎥
⎠⎦
)
{s1,…,s M } ⎣
1 K
K
s = arg max ln[ pm ] − ln[ πN 0 ] − ∑ rk − sm, k (
2
)
{s ,…,s }
1 M
2 N 0 k =1
Kyung Hee
Structure of Optimum Receiver (continued)
University
K
s = arg max ln[ pm ] − ln[ πN 0 ]
{s1,…,s M } 2
1 ⎛ K 2 K K
2⎞
− ⎜ ∑ k r − 2 r s +
∑ k m,k ∑ m,k ⎟
s
N 0 ⎝ k =1 k =1 k =1 ⎠
Eliminating terms which are identical for all choices:

2 K 1 K
s = arg max ln[ pm ] + ∑ rk sm,k − ∑ sm, k 2
{s ,…,s }
1 M
N 0 k =1 N 0 k =1

Kyung Hee
Structure of Optimum Receiver
University
MAP rule : s = arg max pm ⋅ p( r sm )

{s1,…,s M }
⎛ K ⎞
s = arg max m ( 0 )
p ⋅ πN −K 2
exp
⎝
−
⎜ ∑ k m, k
r − s
2
N (
0⎟
⎠
)
{s ,…,s }
1 M k =1
⎡ ⎛ K ⎞⎤
s = arg max ln ⎢ pm ⋅ ( πN 0 )
−K 2
⎝ k =1
(2
N0⎟ ⎥
⎠⎦
)
{s1,…,s M } ⎣
1 K
K
s = arg max ln[ pm ] − ln[ πN 0 ] − ∑ rk − sm, k (
2
)
{s ,…,s }
1 M
2 N 0 k =1
Kyung Hee
Structure of Optimum Receiver (continued)
University
K
s = arg max ln [pm ]− ln[ πN 0 ]
2
{s1,…,s M }
1 ⎛ K 2 K K
2⎞
− ⎜ ∑ k r − 2 r s +
∑ k m,k ∑ m,k ⎟
s
N 0 ⎝ k =1 k =1 k =1 ⎠
Eliminating terms which are identical for all choices:

2 K 1 K
s = arg max ln[ pm ] + ∑ rk sm,k − ∑ sm, k 2
{s ,…,s }
1 M
N 0 k =1 N 0 k =1

Final Form of MAP Receiver
Multiplying through by the constant N0 2 :
N0 K 1 K
s = arg max ln[ pm ] + ∑ rk sm,k − ∑ sm,k 2
{s1,…,s M } 2 k =1 2 k =1
Kyung Hee
Interpreting This Result
University
N0
ln[ pm ] weights the a priori probabilities
2
z If the noise is large, pm counts a lot
z If the noise is small, our received signal will be an accurate
estimate and pm counts less
K T
∑ rk sm, k = ∫ sm (t )r (t )dt represents the correlation with
k =1 0
the received signal
1 K 1T 2 E
∑ sm,k = ∫ sm (t )dt = m
2
represents signal energy
2 k =1 20 2

Kyung Hee
Final Form of MAP Receiver
University
Multiplying through by the constant N0 2 :
N0 K 1 K
s = arg max ln[ pm ] + ∑ rk sm,k − ∑ sm,k 2
{s1,…,s M } 2 k =1 2 k =1
Interpreting This Result

N0
ln[ pm ] weights the a priori probabilities
2
z If the noise is large, pm counts a lot
z If the noise is small, our received signal will be an accurate
estimate and pm counts less
K T
∑ rk sm, k = ∫ sm (t )r (t )dt represents the correlation with
k =1 0
the received signal
1 K 1T 2 E
∑ sm,k = ∫ sm (t )dt = m
2
represents signal energy
2 k =1 20 2
An Implementation of the Optimal Receiver -
Correlation Receiver
r (t ) × T
∫0 dt Σ Σ
s1(t ) − E1 2 N0
ln( p1 )
2 Choose
Largest
r (t ) × T
∫0 dt Σ Σ
sM (t ) − EM 2 N0
ln( pM )
2
Kyung Hee
Simplifications for Special Cases
University
ML case: All signals are equally likely ( p1 = = pM ). A

priori probabilities can be ignored.
All signals have equal energy ( E1 = = EM ). Energy

terms can be ignored.
We can reduce the number of correlations by directly

implementing:
N0 K 1 K
s = arg max ln[ pm ] + ∑ rk sm, k − ∑ sm, k 2
{s1,…,s M } 2 k =1 2 k =1

An Implementation of the Optimal Receiver -
Kyung Hee
University Correlation Receiver
r (t ) × T
∫0 dt Σ Σ
s1(t ) − E1 2 N0
ln( p1 )
2 Choose
Largest
r (t ) × T
∫0 dt Σ Σ
sM (t ) − EM 2 N0
ln( pM )
2
Simplifications for Special Cases
ML case: All signals are equally likely ( p1 = = pM ). A

priori probabilities can be ignored.
All signals have equal energy ( E1 = = EM ). Energy

terms can be ignored.
We can reduce the number of correlations by directly

implementing:
N0 K 1 K
s = arg max ln[ pm ] + ∑ rk sm, k − ∑ sm, k 2
{s1,…,s M } 2 k =1 2 k =1
Reduced Complexity Implementation:
Correlation Stage
r (t ) × T
∫0 dt r1
f1(t )
r = [ r1 rK ]
r (t ) × T
∫0 dt rK
f K (t )
Wireless Communication Systems Lab.
Reduced Complexity Implementation -

Kyung Hee
University Processing Stage
K
∑ s1,k rk Σ Σ
k =1
r ×
− E1 2 N0
ln( p1) Choose
2
⎡ s11, sM,1 ⎤ Largest
⎢ ⎥
⎢ ⎥
⎢⎣s1, K K
sM, K ⎥⎦ ∑ Σ Σ
sM,k rk
k =1
− EM 2 N0 ln p
( M)
2

Reduced Complexity Implementation:
Kyung Hee
University Correlation Stage
r (t ) × T
∫0 dt r1
f1(t )
r = [ r1 rK ]
r (t ) × T
∫0 dt rK
f K (t )
Reduced Complexity Implementation -

Processing Stage
K
∑ s1,k rk Σ Σ
k =1
r ×
− E1 2 N0
ln( p1) Choose
2
⎡ s11, sM,1 ⎤ Largest
⎢ ⎥
⎢ ⎥
⎢⎣s1, K K
sM, K ⎥⎦ ∑ Σ Σ
sM,k rk
k =1
− EM 2 N0 ln p
( M)
2
Matched Filter Implementation
Assume fk (t) is time-limited to t ∈[0, T ] , and let hk (t ) = f k (T − t )
T T
Then rk = ∫ r ( t ) f k ( t )dt = ∫ r ( t ) f k (T − ( T − t ) )dt
0 0
T
= ∫ r ( t ) hk ( T − t )dt = r ( t ) ⊗ hk ( t ) t = T
0
where r(t ) ⊗ hk (t ) t =T denotes the convolution of the signals r(t)
and hk (t) evaluated at time T
We can implement each correlation by passing r(t) through

a filter with impulse response hk (t)
Matched Filter Implementation of

Kyung Hee
University Correlation
t=T
r (t ) h1(t ) r1
r = [ r1 rK ]
t=T
r (t ) hK (t ) rK

Kyung Hee
Matched Filter Implementation
University
Assume fk (t) is time-limited to t ∈[0, T ] , and let hk (t ) = f k (T − t )
T T
Then rk = ∫ r ( t ) f k ( t )dt = ∫ r ( t ) f k (T − ( T − t ) )dt
0 0
T
= ∫ r ( t ) hk ( T − t )dt = r ( t ) ⊗ hk ( t ) t = T
0
where r(t ) ⊗ hk (t ) t =T denotes the convolution of the signals r(t)
and hk (t) evaluated at time T
We can implement each correlation by passing r(t) through

a filter with impulse response hk (t)

Correlation
t=T
r (t ) h1(t ) r1
r = [ r1 rK ]
t=T
r (t ) hK (t ) rK
Example of Optimal Receiver Design
Consider the signal set:

s1(t ) s 2 (t )
+1 +1
t t
-1 1 2 -1 1 2
s 3 (t ) s 4 (t )
+1 +1
t t
1 2 -1 1 2
-1

Kyung Hee
University (continued)
Suppose we use the basis functions:
f 1(t ) f 2(t )
+1 +1
t t
-1 1 2 -1 1 2
s1( t ) = 1 ⋅ f1 ( t ) + 1 ⋅ f 2 ( t ) s2 ( t ) = 1 ⋅ f1 ( t ) − 1 ⋅ f 2 ( t )
s3 ( t ) = −1 ⋅ f1 ( t ) + 1 ⋅ f 2 ( t ) s4 ( t ) = −1 ⋅ f1 ( t ) − 1 ⋅ f 2 ( t )
T=2
E1 = E2 = E3 = E4 = 2

Kyung Hee
University
Consider the signal set:

s1(t ) s 2 (t )
+1 +1
t t
-1 1 2 -1 1 2
s 3 (t ) s 4 (t )
+1 +1
t t
1 2 -1 1 2
-1

(continued)
Suppose we use the basis functions:
f 1(t ) f 2(t )
+1 +1
t t
-1 1 2 -1 1 2
s1( t ) = 1 ⋅ f1 ( t ) + 1 ⋅ f 2 ( t ) s2 ( t ) = 1 ⋅ f1 ( t ) − 1 ⋅ f 2 ( t )
s3 ( t ) = −1 ⋅ f1 ( t ) + 1 ⋅ f 2 ( t ) s4 ( t ) = −1 ⋅ f1 ( t ) − 1 ⋅ f 2 ( t )
T=2
E1 = E2 = E3 = E4 = 2
1st Implementation of Correlation Receiver
r (t ) × 2
∫0 dt Σ
s1(t ) N0
ln( p1 ) Choose
2
Largest
r (t ) × 2
∫0 dt Σ
s4 (t ) N0
ln( p4 )
2
Reduced Complexity Correlation Receiver -

Kyung Hee
University Correlation Stage
r (t ) × 2
∫0 dt r1
f1(t )
r = [ r1 r2 ]
r (t ) × 2
∫0 dt r2
f2 (t )

Kyung Hee
1st Implementation of Correlation Receiver
University
r (t ) × 2
∫0 dt Σ
s1(t ) N0
ln( p1 ) Choose
2
Largest
r (t ) × 2
∫0 dt Σ
s4 (t ) N0
ln( p4 )
2

Correlation Stage
r (t ) × 2
∫0 dt r1
f1(t )
r = [ r1 r2 ]
r (t ) × 2
∫0 dt r2
f2 (t )
Processing Stage
1 ⋅ r1 + 1 ⋅ r2 Σ
N0 ln( p1 ) 2
1 ⋅ r1 − 1 ⋅ r2 Σ Choose
N0 ln( p2 ) 2
Largest
−1 ⋅ r1 + 1 ⋅ r2 Σ
N0 ln( p3 ) 2
−1 ⋅ r1 − 1 ⋅ r2 Σ
N0 ln( p4 ) 2

Kyung Hee
University Correlations
hk ( t ) = f k ( 2 − t )
h1(t ) h 2 (t )
+1 +1
t t
1 2 1 2
t=2
r (t ) h1(t ) r1
t=2 r = [ r1 r2 ]
r (t ) h2 (t ) r2

Kyung Hee
University Processing Stage
1 ⋅ r1 + 1 ⋅ r2 Σ
N0 ln( p1 ) 2
1 ⋅ r1 − 1 ⋅ r2 Σ Choose
N0 ln( p2 ) 2
Largest
−1 ⋅ r1 + 1 ⋅ r2 Σ
N0 ln( p3 ) 2
−1 ⋅ r1 − 1 ⋅ r2 Σ
N0 ln( p4 ) 2

Correlations
hk ( t ) = f k ( 2 − t )
h1(t ) h 2 (t )
+1 +1
t t
1 2 1 2
t=2
r (t ) h1(t ) r1
t=2 r = [ r1 r2 ]
r (t ) h2 (t ) r2
Summary of Optimal Receiver Design
Optimal coherent receiver for AWGN has three parts:

z Correlates the received signal with each possible transmitted
signal signal
z Normalizes the correlation to account for energy
z Weights the a priori probabilities according to noise power
This receiver is completely general for any signal set
Simplifications are possible under many circumstances
Kyung Hee
Decision Regions
University
Optimal Decision Rule:

N0 K 1 K
s = arg max ln[ pm ] + ∑ rk sm, k − ∑ sm,k 2
{s1,…,s M } 2 k =1 2 k =1
Let Ri ⊂ ℜ K be the region in which

N0 K 1 K
ln[ pi ] + ∑ rk si ,k − ∑ si , k 2
2 k =1 2 k =1
[ ]
N K 1 K
≥ 0 ln p j + ∑ rk s j ,k − ∑ s j ,k 2 , ∀i ≠ j
2 k =1 2 k =1
Then Ri is the ith “Decision Region”

Kyung Hee
Summary of Optimal Receiver Design
University
Optimal coherent receiver for AWGN has three parts:

z Correlates the received signal with each possible transmitted
signal signal
z Normalizes the correlation to account for energy
z Weights the a priori probabilities according to noise power
This receiver is completely general for any signal set
Simplifications are possible under many circumstances
Decision Regions
Optimal Decision Rule:

N0 K 1 K
s = arg max ln[ pm ] + ∑ rk sm, k − ∑ sm,k 2
{s1,…,s M } 2 k =1 2 k =1
Let Ri ⊂ ℜ K be the region in which

N0 K 1 K
ln[ pi ] + ∑ rk si ,k − ∑ si , k 2
2 k =1 2 k =1
[ ]
N K 1 K
≥ 0 ln p j + ∑ rk s j ,k − ∑ s j ,k 2 , ∀i ≠ j
2 k =1 2 k =1
Then Ri is the ith “Decision Region”

A Matlab Function for
Visualizing Decision Regions
The Matlab Script File “sigspace.m” (on course web page)
can be used to visualize two dimensional signal spaces and
decision regions
The function is called with the following syntax:

sigspace( [ x1 y1 p1; x2 y2 p2; ; x M y M pM ] , Eb N0 )
z xi and yi are the coordinates of the ith signal point

z pi is the probability of the ith signal (omitting gives ML)
z Eb N0 is the signal to noise ratio of digital system in dB
Kyung Hee
Average Energy Per Bit: E
University b
K
Ei = ∑ si ,k 2 is the energy of the ith signal
k =1
1 M is the average energy per symbol

Es = ∑ Ei
M i =1
log 2 M is the number of bits transmitted per symbol
Es
Eb = is the average energy per bit
log 2 M
z allows fair comparisons of the energy requirements of

different sized signal constellations

A Matlab Function for
Kyung Hee
University Visualizing Decision Regions
The Matlab Script File “sigspace.m” (on course web page)
can be used to visualize two dimensional signal spaces and
decision regions
The function is called with the following syntax:

sigspace( [ x1 y1 p1; x2 y2 p2; ; x M y M pM ] , Eb N0 )
z xi and yi are the coordinates of the ith signal point

z pi is the probability of the ith signal (omitting gives ML)
z Eb N0 is the signal to noise ratio of digital system in dB
Average Energy Per Bit: E

b
K
Ei = ∑ si ,k 2 is the energy of the ith signal
k =1
1 M is the average energy per symbol

Es = ∑ Ei
M i =1
log 2 M is the number of bits transmitted per symbol
Es
Eb = is the average energy per bit
log 2 M
z allows fair comparisons of the energy requirements of

different sized signal constellations
Signal to Noise Ratio for Digital Systems
N0 2 is the (two-sided) power spectral density of the

background noise
The ratio Eb N 0 measures the relative strength of signal

and noise at the receiver
Eb has units of Joules = Watts *sec
N0 has units of Watts/Hz = Watts*sec
The unitless quantity Eb N 0 is frequently expressed in dB
Kyung Hee
Examples of Decision Regions - QPSK
University
sigspace( [1 0; 0 1; -1 0; 0 -1], 20)
0.8
0.6
0.4
0.2
-0.2
-0.4
-0.6
-0.8
-1
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

Kyung Hee
Signal to Noise Ratio for Digital Systems
University
N0 2 is the (two-sided) power spectral density of the

background noise
The ratio Eb N 0 measures the relative strength of signal

and noise at the receiver
Eb has units of Joules = Watts *sec
N0 has units of Watts/Hz = Watts*sec
The unitless quantity Eb N 0 is frequently expressed in dB
Examples of Decision Regions - QPSK
sigspace( [1 0; 0 1; -1 0; 0 -1], 20)
0.8
0.6
0.4
0.2
-0.2
-0.4
-0.6
-0.8
-1
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
QPSK with Unequal Signal Probabilities
sigspace( [1 0 0.4; 0 1 0.1; -1 0 0.4; 0 -1 0.1], 5)
0.8
0.6
0.4
0.2
-0.2
-0.4
-0.6
-0.8
-1
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
QPSK with Unequal Signal Probabilities -

Kyung Hee
University Extreme Case
sigspace([0.5 0 0.4; 0 0.5 0.1; -0.5 0 0.4; 0 -0.5 0.1],-6)
0.8
0.6
0.4
0.2
-0.2
-0.4
-0.6
-0.8
-1
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

Kyung Hee
QPSK with Unequal Signal Probabilities
University
sigspace( [1 0 0.4; 0 1 0.1; -1 0 0.4; 0 -1 0.1], 5)
0.8
0.6
0.4
0.2
-0.2
-0.4
-0.6
-0.8
-1
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
QPSK with Unequal Signal Probabilities -

Extreme Case
sigspace([0.5 0 0.4; 0 0.5 0.1; -0.5 0 0.4; 0 -0.5 0.1],-6)
0.8
0.6
0.4
0.2
-0.2
-0.4
-0.6
-0.8
-1
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Unequal Signal Powers
sigspace( [1 1 ; 2 2; 3 3; 4 4], 10)
3.5
2.5
1.5
0.5
0
0 0.5 1 1.5 2 2.5 3 3.5 4
Kyung Hee
Signal Constellation for 16-ary QAM
University
sigspace( [1.5 -1.5; 1.5 -0.5; 1.5 0.5; 1.5 1.5; 0.5 -1.5; 0.5 -
0.5; 0.5 0.5; 0.5 1.5;-1.5 -1.5; -1.5 -0.5; -1.5 0.5; -1.5 1.5; -
0.5 -1.5; -0.5 -0.5; -0.5 0.5; -0.5 1.5],10)
2
1.5
0.5
-0.5
-1
-1.5
-2
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2

Kyung Hee
Unequal Signal Powers
University
sigspace( [1 1 ; 2 2; 3 3; 4 4], 10)
3.5
2.5
1.5
0.5
0
0 0.5 1 1.5 2 2.5 3 3.5 4
Signal Constellation for 16-ary QAM
sigspace( [1.5 -1.5; 1.5 -0.5; 1.5 0.5; 1.5 1.5; 0.5 -1.5; 0.5 -
0.5; 0.5 0.5; 0.5 1.5;-1.5 -1.5; -1.5 -0.5; -1.5 0.5; -1.5 1.5; -
0.5 -1.5; -0.5 -0.5; -0.5 0.5; -0.5 1.5],10)
2
1.5
0.5
-0.5
-1
-1.5
-2
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Notes on Decision Regions
Boundaries are perpendicular to a line drawn between two

signal points
If signal probabilities are equal, decision boundaries lie

exactly halfway in between signal points
If signal probabilities are unequal, the region of the less

probable signal will shrink.
Signal points need not lie within their decision regions for
case of low Eb N0 and unequal probabilities.

Chapter 7 - Optimal Receiver Design PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 7 - Optimal Receiver Design PDF

Uploaded by

Copyright:

Available Formats

Digital Communications

Instructor: Dr. Phan Van Ca

 We want to modulate digital data using signal sets which are :

 A signal space representation is a convenient form for

z design energy and bandwidth efficient signal constellations

Young S. Kim Wireless Communication Systems Lab. - 2 -

Instructor: Young Soo Kim

Young S. Kim Wireless Communication Systems Lab. - 1 -

 We want to modulate digital data using signal sets which are :

 A signal space representation is a convenient form for

z design energy and bandwidth efficient signal constellations

 We transmit a signal s(t ) ∈{s1(t ), s2 (t ),…, sM (t )} , where s(t)

 Let the various signals be transmitted with probability:

 The received signal is corrupted by noise:

 Given r(t) , the receiver forms an estimate s(t) of the signal

 The signal is corrupted by Additive White Gaussian Noise

 Any linear function of n(t) will be a Gaussian random

 We transmit a signal s(t ) ∈{s1(t ), s2 (t ),…, sM (t )} , where s(t)

 Let the various signals be transmitted with probability:

 The received signal is corrupted by noise:

 Given r(t) , the receiver forms an estimate s(t) of the signal

Young S. Kim Wireless Communication Systems Lab. - 3 -

 The signal is corrupted by Additive White Gaussian Noise

 Any linear function of n(t) will be a Gaussian random

 The transmitted signal can be represented as:

 The received signal can be represented as :

Young S. Kim Wireless Communication Systems Lab. - 6 -

 The transmitted signal can be represented as:

Young S. Kim Wireless Communication Systems Lab. - 5 -

Signal Space Representation (continued)

 The received signal can be represented as :

 The noise n′(t) can be disregarded by the receiver

We can reduce the decision to a finite

 We receive a vector r = [r1,…, rK ] = s + n which is the sum of

 Given r , we wish to form an estimate s of the transmitted

 The noise n′(t) can be disregarded by the receiver

Young S. Kim Wireless Communication Systems Lab. - 7 -

We can reduce the decision to a finite

 We receive a vector r = [r1,…, rK ] = s + n which is the sum of

 Given r , we wish to form an estimate s of the transmitted

 We minimize symbol error probability by choosing the

p( r sm ) Pr (sm ) p(r si ) Pr (si )

or p( r sm ) Pr (sm ) ≥ p(r si ) Pr (si ), ∀m ≠ i

 If p1 = = pm or the a priori probabilities are unknown,

 We minimize symbol error probability by choosing the

Young S. Kim Wireless Communication Systems Lab. - 10 -

 We minimize symbol error probability by choosing the

p( r sm ) Pr (sm ) p(r si ) Pr (si )

or p( r sm ) Pr (sm ) ≥ p(r si ) Pr (si ), ∀m ≠ i

Young S. Kim Wireless Communication Systems Lab. - 9 -

Maximum Likelihood (ML) Decision Rule

 If p1 = = pm or the a priori probabilities are unknown,

 We minimize symbol error probability by choosing the

 In order to apply either the MAP or ML rules, we need to

 Since r = sm + n where sm is constant, it is equivalent to

 n(t) is a Gaussian random process

Young S. Kim Wireless Communication Systems Lab. - 12 -

 In order to apply either the MAP or ML rules, we need to

 Since r = sm + n where sm is constant, it is equivalent to

 n(t) is a Gaussian random process

Young S. Kim Wireless Communication Systems Lab. - 11 -

The Noise p.d.f

 Since E[ni ⋅ nk ] = 0,∀i ≠ k , individual noise components are

We want to modulate digital data using signal sets which are :

A signal space representation is a convenient form for

We want to modulate digital data using signal sets which are :

A signal space representation is a convenient form for

We transmit a signal s(t ) ∈{s1(t ), s2 (t ),…, sM (t )} , where s(t)

Let the various signals be transmitted with probability:

The received signal is corrupted by noise:

Given r(t) , the receiver forms an estimate s(t) of the signal

The signal is corrupted by Additive White Gaussian Noise

Any linear function of n(t) will be a Gaussian random

We transmit a signal s(t ) ∈{s1(t ), s2 (t ),…, sM (t )} , where s(t)

Let the various signals be transmitted with probability:

The received signal is corrupted by noise:

Given r(t) , the receiver forms an estimate s(t) of the signal

The signal is corrupted by Additive White Gaussian Noise

Any linear function of n(t) will be a Gaussian random

The transmitted signal can be represented as:

The received signal can be represented as :

The transmitted signal can be represented as:

The received signal can be represented as :

The noise n′(t) can be disregarded by the receiver

We receive a vector r = [r1,…, rK ] = s + n which is the sum of

Given r , we wish to form an estimate s of the transmitted

The noise n′(t) can be disregarded by the receiver

We receive a vector r = [r1,…, rK ] = s + n which is the sum of

Given r , we wish to form an estimate s of the transmitted

We minimize symbol error probability by choosing the

If p1 = = pm or the a priori probabilities are unknown,

We minimize symbol error probability by choosing the

We minimize symbol error probability by choosing the

If p1 = = pm or the a priori probabilities are unknown,

We minimize symbol error probability by choosing the

In order to apply either the MAP or ML rules, we need to

Since r = sm + n where sm is constant, it is equivalent to

n(t) is a Gaussian random process

In order to apply either the MAP or ML rules, we need to

Since r = sm + n where sm is constant, it is equivalent to

n(t) is a Gaussian random process

Since E[ni ⋅ nk ] = 0,∀i ≠ k , individual noise components are

Transmitted signal values in each dimension represent the

Since E[ni ⋅ nk ] = 0,∀i ≠ k , individual noise components are

Transmitted signal values in each dimension represent the

MAP rule : s = arg max pm ⋅ p( r sm )

Eliminating terms which are identical for all choices:

MAP rule : s = arg max pm ⋅ p( r sm )

Eliminating terms which are identical for all choices:

Multiplying through by the constant N0 2 :

Multiplying through by the constant N0 2 :

ML case: All signals are equally likely ( p1 = = pM ). A

All signals have equal energy ( E1 = = EM ). Energy

We can reduce the number of correlations by directly

ML case: All signals are equally likely ( p1 = = pM ). A

All signals have equal energy ( E1 = = EM ). Energy

We can reduce the number of correlations by directly

Assume fk (t) is time-limited to t ∈[0, T ] , and let hk (t ) = f k (T − t )

We can implement each correlation by passing r(t) through

Assume fk (t) is time-limited to t ∈[0, T ] , and let hk (t ) = f k (T − t )

We can implement each correlation by passing r(t) through

Consider the signal set:

Consider the signal set:

Optimal coherent receiver for AWGN has three parts:

This receiver is completely general for any signal set

Simplifications are possible under many circumstances

Optimal Decision Rule:

Let Ri ⊂ ℜ K be the region in which

Then Ri is the ith “Decision Region”

Optimal coherent receiver for AWGN has three parts:

This receiver is completely general for any signal set

Simplifications are possible under many circumstances

Optimal Decision Rule:

Let Ri ⊂ ℜ K be the region in which

Then Ri is the ith “Decision Region”

The function is called with the following syntax: