Scribd
Upload
69 views

Lect21QAM Error Probability

This document discusses error probability in QAM communication systems. It defines 16-QAM signals and decision regions. It calculates the probability of correct reception for different signal points and derives an expression for symbol error probability Pe as a function of average symbol energy to noise ratio. It then relates symbol error probability to bit error probability Pb, showing Pb is approximated as a function of average bit energy to noise ratio.

Uploaded by

Fatima Ahsan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
69 views

Lect21QAM Error Probability

This document discusses error probability in QAM communication systems. It defines 16-QAM signals and decision regions. It calculates the probability of correct reception for different signal points and derives an expression for symbol error probability Pe as a function of average symbol energy to noise ratio. It then relates symbol error probability to bit error probability Pb, showing Pb is approximated as a function of average bit energy to noise ratio.

Uploaded by

Fatima Ahsan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

'

EE4601
Communication Systems
Lecture 21
QAM Error Probability

&

c
2007,
Georgia Institute of Technology (lect21 1)

'

QAM Signals

With QAM signals


sm (t) =

v
u
u 2E0
t
acm cos (2fc t)

asm sin (2fc t)

acm , asm {1, 3}

The appropriate basis functions for the signal space are


f1 (t) =

v
u
u2
t

cos (2fc t)

v
u
u2
t

f2 (t) =

sin (2fc t)

Then

E0 acm f1 (t) + E0 asm f2 (t)


sm (t) =

E0 (acm , asm )
sm =
We randomly choose one of the 16 signals to transmit over an AWGN channel
and receive r = sm +n, where n = (n1 , n2 ), and the ni are i.i.d. Gaussian random
variables with variance 2 = No /2.
Our task is to find the probability of symbol error with minimum distance (or
maximum likelihood) decisions.
&

c
2007,
Georgia Institute of Technology (lect21 2)

'

QAM Signals

To calculate the probability of symbol error, we first must define appropriate


decision regions by placing decision boundaries between the signal points. For
16-QAM this is shown
below.

Note that a = E0 in the figure.


s0

s1

s2

s3

3a

decision
boundaries
s4

s5

-3a

-a

s6

s7

3a

-a
s8

s9

s 10

decision
regions

s 11

-3a
s 12

s 13

&

c
2007,
Georgia Institute of Technology (lect21 3)

s 14

s 15

'

QAM Signals

For problems of this type, and especially for one or two-dimensional signal spaces
(this problem is 2-D), it is often easier to calculate the probability of correct reception.
For this problem there are 3 cases to consider, since we can observe graphically
that
PC|s5 = PC|s6 = PC|s9 = PC|s10
PC|s0 = PC|s3 = PC|s12 = PC|s15
PC|s1 = PC|s2 = PC|s4 = PC|s7 = PC|s8 = PC|s11 = PC|s13 = PC|s14
All these quantities can be expressed in terms of the parameter

E
No
0

Q Q
2 =

2
&

c
2007,
Georgia Institute of Technology (lect21 4)

'

QAM Signals

All these quantities can be expressed in terms of the parameter

E
No
0

Q Q
2 =

2
We have
PC|s5 = (1 2Q)2 = 1 4Q + 4Q2

PC|s0 = (1 Q)2 = 1 2Q + Q2
PC|s1 = (1 Q) (1 2Q) = 1 3Q + 2Q2
Then
1
1
1
PC|s5 + PC|s0 + PC|s1
4
4
2
9 2
= 1 3Q + Q
4

PC =

Finally, the probability of error is Pe = 1 PC = 3Q 49 Q2

&

c
2007,
Georgia Institute of Technology (lect21 5)

'

QAM Signals

Next, we need to find the average symbol energy. Remember that the energy in
a symbol is equal to squared length of the signal vector.
In this case,
1
1
1
Eav = (E0 + E0 ) + (9E0 + 9E0 ) + (E0 + 9E0 ) = 10E0
4
4
2
Hence, E0 = Eav /10, and

Q = Q

v
u

v
u

u 2E0
uE
E0
= Q t av
= Q t

No
5No

Finally,
Pe =
where
&

u
u Eav

3Q t

5No

v
u

u Eav
9

Q2 t
4
5No

Eav
= average symbol energy-to-noise ratio
No

c
2007,
Georgia Institute of Technology (lect21 6)

'

QAM Signals

What about the bit error probability? That depends on the mapping of bits to
symbols.
With Gray coding, a symbol error will usually result in one bit error. Certainly
at most 4 bits errors will occur. Hence,
Pe <
Pb < Pe
4
Also, there are 4 bits per modulated symbol so that the average bit energy-tonoise ratio is
Eb av = Eav /4
So we can write
v
u

v
u

u 4 Eb av
9 2 u
3
4 Eb av

Q t
Pb Q t
4
5 No
16
5 No
>

&

c
2007,
Georgia Institute of Technology (lect21 7)

You might also like