You are on page 1of 6

KEEE494: 2nd Semester 2009 Week 3

Some Variants of QPSK: Offset QPSK, DQPSK and Effect of


Phase Error

1 Offset QPSK
• Transmitter

Figure 1: Transmit block diagram of Offset QPSK

• The transmit signal s(t) can be represented as


X X Ts
s(t) = A anI g(t − nTs ) cos(ωc t) − A anQ g(t − nTs − ) sin(ωc t)
n n
2

Figure 2: I and Q signal waveform of OQPSK

• Receiver

1
Figure 3: Receiver block diagram of OQPSK

• Bit error performance

– If φ = 0 (i.e., no phase error), the BER of OQPSK is the same as QPSK.


– BER of QPSK in the presence of phase error, φ 6= 0
Z Ts Z Ts
yI (Ts ) = (a0I cos(ωc t) − a0Q sin(ωc t))2 cos(ωc t − φ) dt + n(t)2 cos(ωc t) dt
0 0
= Aa0I cos(φ)Ts − Aa0Q sin(φ)Ts + NI

where NI N (0, N0 Ts ). Then, the BER assuming a0I = 1 is

Pb (E|a0I = 1) = Pb (E|a0I = 1, a0Q = 1) · Pr(a0Q = 1) + Pb (E|a0I = 1, a0Q = 1) · Pr(a0Q = −1)


µ ¶ µ ¶
1 A(cos(φ) − sin(φ)) 1 A(cos(φ) + sin(φ))
= erfc √ + erfc √
4 2σ 4 2σ

Using A = P and Ts = 2Tb , and
r r r
AT P Ts P Tb Eb
√ √ s = = =
2 N0 Ts 2N0 N0 N0
the BER can be rewritten as
Ãr ! Ãr !
1 Eb 1 Eb
Pb (E|a0I = 1) = erfc (cos(φ) − sin(φ)) + erfc (cos(φ) + sin(φ))
4 N0 4 N0

Finally, the BER under the equal probabilities of Pr(a0I ) = Pr(a0Q ) = 0.5 is

Pb (E) = Pb (E|a0I = 1) Pr(a0I ) + (E|a0Q = −1) Pr(a0Q )


= Pb (E|a0I = 1)

– BER of BPSK with the presence of phase error, φ 6= 0


Ãr !
1 Eb
Pb (E) = erfc cos(φ)
2 N0

So we can conclude that QPSK gives worse performance than BPSK when the phase error exists.

2
Figure 4: I and delayed Q

– BER of OQPSK
Ãr !
1 1 Eb
Pb (E) = · erfc cos(φ)
2 2 N0
" Ãr ! Ãr !#
1 1 Eb 1 Eb
+ erfc (cos(φ) − sin(φ)) + erfc (cos(φ) + sin(φ))
2 4 N0 4 N0
Ãr ! Ãr ! Ãr !
1 Eb 1 Eb 1 Eb
= erfc cos(φ) + erfc (cos(φ) − sin(φ)) + erfc (cos(φ) + sin(φ))
4 N0 8 N0 8 N0

– Comparison of BER with the presence of phase error, φ 6= 0

BPSK < OQPSK < QPSK

2 Differential Phase Shift Keying


• Generation and Detection of binary DPSK

– Let {bk } be the information bit and {dl } denote the differentially encoded sequence. We now introduce
the following definitions in the generation of this sequence.
∗ If the incoming binary bit bk is 1, leave the symbol dk unchanged with respect to the previous bit.
∗ If the incoming binary bit bk is 0, change the symbol dk with respect to the previous bit.
The differentially encoded sequence {dk } thus generated is used to phase-shift a carrier with phase angles
0 and π radians representing the bit 0 and 1, respectively. The differential-phase encoding process is
illustrated in the following Table.

Table 1: Illustrating the generation of DPSK signal


{bk } 1 0 0 1 0 0 1 1
{dk−1 } 1 1 0 1 1 0 1 1
{dk } 1 1 0 1 1 0 1 1 1
Transmitted phase 0 0 π 0 0 π 0 0 0

• Block diagram of binary transmitter DPSK is one example of noncoherent orthogonal√ modulation, when it is
considered over two bit intervals. Suppose the transmitted DPSK signal equals P cos(2πfc t) for 0 ≤ t ≤ Tb ,
where Tb is the bit duration and Eb is the signal energy per bit. Let s1 (t) denote the transmitted DPSK signal
for 0 ≤ t ≤ 2Tb for the case when we have binary bit 1 at the transmitter input for the second part of this

3
Figure 5: Block diagram of binary DPSK transmitter

interval, namely, Tb ≤ t ≤ 2Tb . The transmission of bit 1 leaves the carrier phase unchanged over the interval
0 ≤ t ≤ 2Tb , and sw we define s1 (t) as
½ √
s1 (t) = √P cos(2πfc t), 0 ≤ t ≤ Tb
P cos(2πfc t), Tb ≤ t ≤ 2Tb

Let s2 (t) denote the transmitted DPSK signal for 0 ≤ t ≤ 2Tb for the case when we have binary symbol 0 at
the transmitter input for Tb ≤ t ≤ 2Tb . The transmission of 0 advances the carrier phase by 180 degrees, and so
we define s2 (t) as ½ √
s1 (t) = √P cos(2πfc t), 0 ≤ t ≤ Tb
P cos(2πfc t + π), Tb ≤ t ≤ 2Tb
We see that s1 (t) and s2 (t) are indeed orthogonal over the two-bit interval 0 ≤ t ≤ 2Tb .
• Detection block diagram is illustrated in Fig. 2.

Figure 6: Block diagram of binary DPSK receiver

• Suppose next, in differentially coherent detection of binary DPSK, the carrier phase is unknown (or the phase er-
ror given by φ). Then, in light of the receiver being equipped√with an in-phase
√ and a quadrature
√ channel,
√ we have
a signal space diagram where the received signal points are ( P cos φ, P sin φ) and (− P cos φ, − P sin φ).
The geometry of possible signals is illustrated in 2. The receiver measures the coordinates (yI0 , yQ0 ) at time
t = Tb and (yI0 , yQ0 ) at time t = 2Tb . The issue to be resolved is whether these two points map to the same
signal point or different ones. Assume that the phase error φ is the constant over the two bit interval 2Tb . Then
the multiplication of two consecutive symbols must be positive when the bit 1 is transmitted and negative when
the bit 0 is transmitted. Accordingly we may write

If yI0 yI1 + yI0 yQ1 > 0, then say 1,

If yI0 yI1 + yI0 yQ1 < 0, then say 0,


Hence, the receiver block diagram can be drawn as in Fig. 8.
• Differential M -PSK

4
Figure 7: Signal-space diagram of received DPSK signal

Figure 8: Signal-space diagram of received DPSK signal

– In general the differential encoding can be expressed by the phase change with 2π modulo operation as
shown in Figs. 9 and 10. In Figs. 9 and 10, ∆θi is the phase of the M -PSK modulated signal and then
it is differentially encode. For example for D-QPSK, the following table shows the phase change at the
transmitter

Table 2: Phase change of D-QPSK


Data: ∆θn π 0 0 − π2 π2 0
π
θn 0 π π π 2 π π

5
Figure 9: Differential encoder and decoder (real form)

Figure 10: Differential encoder and decoder (complex form)

You might also like