Scribd
Upload
32 views

COMM 2 Notes

1) QPSK detection in noise involves using a quadrature receiver to separate the signal into an in-phase and quadrature-phase component. Each component represents a separate BPSK system. After low-pass filtering and integration, the receiver makes decisions based on the sign of the in-phase and quadrature outputs. 2) M-PAM detection in noise involves using multiple thresholds at the receiver based on the symbol levels. The probability of symbol error depends on the distance between adjacent symbols and can be expressed in terms of the energy per bit to noise spectral density ratio. 3) Increasing the number of levels in M-PAM increases the bit rate but also decreases performance, requiring more energy per bit

Uploaded by

Georges Chouchani
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
32 views

COMM 2 Notes

1) QPSK detection in noise involves using a quadrature receiver to separate the signal into an in-phase and quadrature-phase component. Each component represents a separate BPSK system. After low-pass filtering and integration, the receiver makes decisions based on the sign of the in-phase and quadrature outputs. 2) M-PAM detection in noise involves using multiple thresholds at the receiver based on the symbol levels. The probability of symbol error depends on the distance between adjacent symbols and can be expressed in terms of the energy per bit to noise spectral density ratio. 3) Increasing the number of levels in M-PAM increases the bit rate but also decreases performance, requiring more energy per bit

Uploaded by

Georges Chouchani
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

5.

Detection of QP SK , P AM and QAM in


noise

1) DETECTION OF QPSK IN NOISE

T: symbol duration

(Gray coding:
Neighbor symbols
differ by 1 bit only)

2 t −T/2
with : m I (t), m Q (t) = ± rect ( )
2 T
1
5. Detection of QP SK , P AM and QAM in
noise

2
Detection of QP SK in noise

Detection: Quadrature receiver for QPSK

YI

YQ

3
Detection of QP SK in noise

Detection: 1st bit


YI

2nd bit

2 BPSK systems YQ

T
LPF Filtering (integrator): Similarly: YQ = ∫ [ ]dt
0
2 2
=> YI = ± A T + NI YQ = ± A T + NQ
4 4
T T
1 N T 1 N 0T
N I = ∫ n I (t)dt ;σ = 0
2
N Q = − ∫ n Q (t)dt ; σ2 = 4
0
2 4 0
2 4
Detection of QP SK in noise

2 N T
⇒ YI and YQ are identically distributed: YI , YQ : N(± A c T; 0 )
4 4

⇒ The bit error rate in the system will be:

2 A c 2 .T
Pe =Q(d/2σ); d= A c .T => Pe = Q( )
2 4σ

In terms of Eb/N0:
Eb=Esymbol/2=(Ac2T/2)/2=Ac2T/4 => A c = 2 E b / T

N 0T N 0T
σ2 = => σ =
4 2
Replacing in Pe:

2E b
=> PeQPSK = Q( )
N0
Consequently, in terms of energy per bit, the QPSK performance is exactly the same as BPSK, even
though we are transmitting twice as many bits through the same channel. 5
Detection of M -P AM in noise

2) Detection of M-PAM

A multi-level baseband PAM signal is represented by:

For example, with four-level PAM, the scalar ck could be selected from the
set {-3,-1,1,3}.

h(t): rectangular pulse

Detection: similar to binary-PAM but with multi thresholds.

6
Detection of M -P AM in noise

Multi-level PAM detector.

Ex: 4-PAM
Receiver output samples (assuming T=1)

s1 s2 s3 s4

-3A -A +A +3A

=> 3 thresholds: µ1, µ2, µ3


Prob. of symbol error Pe?
distance between adjacent symbols = 2A;
Outer symbols: Pe1,4= Q(d/2σ) = Q(A/σ) => Pe=1/4[2Q(A/σ)+2x2Q(A/σ)]=1.5Q(A/σ)
Inner symbols: Pe2,3=2Q(A/σ) 7
Detection of M -P AM in noise

In general, it can be shown that: (1)


(left as an exercise)
Digital reference model (In terms of Eb/N0):

1 M / 2−1 A 2T M / 2−1
Es = ∑ (2k + 1) A T =
M k =− M / 2
2 2
M
[2 ∑ (2k + 1) 2 ]
k =0
M / 2−1
A 2T
= [2 + 2 ∑ (2k + 1) 2 ]
M k =1

(M 2 − 1)A 2 T
= ..... =
3

M- PAM => Number of bits /symbol is: β=log2(M) => Eb=Es/β (2)

Noise variance is the same as binary PAM case: σ2=N0T/2 (3)

(1), (2), (3) => Prob. of symbol error in PAM:


8
Detection of M -P AM in noise

Advantage of M-PAM: increase the transmitted bit rate by increasing the number
of bits per symbol β=log2(M)

Disadvantage: performance penalty in Pe

Let:

M=2 (binary PAM) => => same Pe result as in bipolar PAM, BPSK,
QPSK

M=4 (4-PAM) => => significantly more energy per bit must
be transmitted to obtain the same error
rate.

You might also like