Neyman-Pearson Theorem – Example DC in WGN

Consider the following signal detection problem

H0 : x[n] = w[n] n = 0, 1, . . . , N 1

H1 : x[n] = s[n] + w[n] n = 0, 1, . . . , N 1

where the signal is s[n] = A for A > 0 and w[n] is WGN with variance 2
. Now the NP
detector decides H1 if
h PN i
1 1 1 2
N exp 2 2 n=0 (x[n] A)
(2⇡ 2 ) 2
h PN 1 2 i >
1 1
N exp 2 2 n=0 x [n]
(2⇡ 2 ) 2

Taking the logarithm of both sides and simplification results in

N 1
1 X 2
A 0
x[n] > ln + =
N n=0 NA 2
25
Test statistic is Gaussian under each hypothesis
Neyman-Pearson Theorem – Example DC in WGN
PN 1 0
The NP detector compares the sample mean x̄ = 1
N n=0 x[n] to a threshold . To deter-
mine the detection performance, we first note that the test statistic T (x) = x̄ is Gaussian
under each hypothesis and its distribution is as follows
8
< N (0, 2 ) under H0
N
T (x) ⇠
: N (A, 2 ) under H1
N
✓ 0
◆ q
0 0 2
We have then PF A = P r(T (x) > ; H0 ) = Q p 2 /N
) = N Q 1
(PF A ) and

0
!
0 A
PD = P r(T (x) > ; H1 ) = Q p
2 /N

PD and PF A are related to each other according to the following equation

r !
NA 2 Signal energy-to-noise ratio
1
PD = Q Q (PF A ) 2
26
Probability of detection versus SNR for DC level in WG

 For a given  ,  increases monotonically with SNR

 This hypothesis testing problem is called as mean-shifted Gauss-
Gauss problem since the pdf of test statistics
is where µ  µ
Neyman-Pearson Theorem – Example DC in WGN
r !
1 N A2
PD = Q Q (PF A ) 2

Remember the deflection coefficient

2 (E(T ; H1 ) E(T ; H0 ))2

d =
var(T ; H0 )

N A2
In this case d2 = 2 .
Further notice that the detection performance (PD ) increases monotonic with the deflection
coefficient.

27
Neyman-Pearson Theorem – Example Change Var

Consider an IID process x[n].

2
H0 : x[n] ⇠ N (0, 0)

2
H1 : x[n] ⇠ N (0, 1 ),

with 2
1 > 0.
2

Neyman-Pearson test:
h PN i
1 1 1 2
N exp 2 12 n=0 x[n]
(2⇡ 12 ) 2
h PN i>
1 1 1 2
N exp 2 02 n=0 x [n]
(2⇡ 02 ) 2

28
Neyman-Pearson Theorem – Example Change Var

Taking the logarithm of both sides and simplification results in

✓ ◆ NX1 2
1 1 1 2 N 1
2 2 x [n] > ln + ln 2
2 1 0 n=0
2 0

we then have
N 1
1 X 2 0
x [n] >
N n=0
2
2 1
0 N ln +ln 2
with = 1 1
0
2 2
0 1

What about PD ?

29
Neyman-Pearson Theorem – Example Change Var

For N = 1 we decide for H1 if:

q
0
|x[0]| > .
⇢ q ⇢ q
0 0
PF A = P r |x[0]| > ; H0 = 2P r x[0] > ; H0 .
p !
0
PF A = 2Q p .
2
0
q ✓ ◆q
0 1 1 2
) =Q PF A 0
2
1
p !
1 2
Q 2 PF A 0
PD = 2Q p
2
1

30
The decision regions for change in varince
hypothesis test
 19/12/16

Deterministic Signals
Binary detection problem:
H0 x[n] = w[n]
H1 x[n] = s[n] + w[n]
Assumptions

• s[n] is deterministic and known.

• w[n] is white Gaussian noise with variance 2

.

Question: What is the correlation function of w[n]?

17

Deterministic Signals
The NP detector decides H1 if the likelihood ratio exceeds a threshold or Notice that presence of s[n]
implies change in mean of
p(x; H1 )
L(x) = >
p(x; H0 ) observe signal. Optimal detector

where x = [x[0], x[1], . . . , x[N 1]]T . Since will test whether there is a
" # change in the mean of the test
N 1
1 1 X 2
p(x; H1 ) = N exp 2
(x[n] s[n]) statistic.
(2⇡ 2) 2 2 n=0
" N 1
#
1 1 X 2
p(x; H0 ) = N
exp 2
x [n]
(2⇡ 2) 2 2 n=0

we have " !#
N
X1 N
X1
1 2 2
L(x) = exp 2
(x[n] s[n]) x [n] > .
2 n=0 n=0

18