Solution: From (6.2) we have that, for 8-PSK, γs = (log2 8) · 1015/10 = 94.87. Substituting this into (6.

16) yields
√
Ps ≈ 2Q( 189.74 sin(π/8)) = 1.355 · 10−7 .
Now, using (6.3), we get Pb = Ps /3 = 4.52 · 10−8 . For 16-PSK we have γs = (log2 16) · 1015/10 = 126.49.
Substituting this into (6.16) yields
√
Ps ≈ 2Q( 252.98 sin(π/16)) = 1.916 · 10−3 ,
and by using (6.3) we get Pb = Ps /4 = 4.79 · 10−4 . Note that Pb is much larger for 16-PSK than for 8-PSK given
the same γb . This result is expected because 16-PSK packs more bits per symbol into a given constellation and so,
for a fixed energy per bit, the minimum distance between constellation points will be smaller.

The error probability derivation for MPSK assumes that the carrier phase is perfectly known to the receiver. Under
phase estimation error, the distribution of p(θ) used to obtain Ps must incorporate the distribution of the phase
rotation associated with carrier phase offset. This distribution is typically a function of the carrier phase estimation
technique and the SNR. The impact of phase estimation error on coherent modulation is studied in [1, Apx. C; 2,
Chap. 4.3.2; 3; 4]. These works indicate that, as expected, significant phase offset leads to an irreducible bit error
probability. Moreover, nonbinary signaling is more sensitive than BPSK to phase offset because of the resulting
cross-coupling between in-phase and quadrature signal components. The impact of phase estimation error can be
especially severe in fast fading, where the channel phase changes rapidly owing to constructive and destructive
multipath interference. Even with differential modulation, phase changes over and between symbol times can
produce irreducible errors [5]. Timing errors can also degrade performance; analysis of timing errors in MPSK
performance can be found in [2, Chap. 4.3.3; 6].

6.1.4 Error Probability for MPAM and MQAM

The constellation for MPAM is Ai = (2i − 1 − M )d, i = 1, 2, . . ., M . Each of the M − 2 inner constellation
points of this constellation have two nearest neighbors at distance 2d. The probability of making an error when
sending one of these inner constellation points is just the probability that the noise exceeds d in either direction:
Ps (si) = p(|n| > d), i = 2, . . ., M − 1. For the outer constellation points there is only one nearest neighbor, so
an error occurs if the noise exceeds d in one direction only: Ps (si ) = p(n > d) = .5p(|n| > d), i = 1, M . The
probability of error is thus

M
1 X
Ps = Ps (si)
M
i=1
s  s 
M −2 2d 2 2 2d 2
= 2Q  + Q 
M N0 M N0
s 
2(M − 1)  2d  2
= Q . (6.18)
M N0

From (5.55), the average energy per symbol for MPAM is

M M
1 X 2 1 X 1
Es = Ai = (2i − 1 − M )2 d2 = (M 2 − 1)d2 . (6.19)
M M 3
i=1 i=1

189
Thus we can write Ps in terms of the average energy E s as
r !
2(M − 1) 6γ s
Ps = Q . (6.20)
M M2 − 1

Consider now MQAM modulation with a square signal constellation of size M = L2 . This system can be
viewed as two MPAM systems with signal constellations of size L transmitted over the in-phase and quadrature
signal components, each with half the energy of the original MQAM system. The constellation points in the in-
phase and quadrature branches take values Ai = (2i − 1 − L)d, i = 1, 2, . . ., L. The symbol
√ error probability
for each branch of the MQAM system is thus given by (6.20) with M replaced by L = M and γ s equal to the
average energy per symbol in the MQAM constellation:
√ r !
2( M − 1) 3γ s
Ps, branch = √ Q . (6.21)
M M −1
Note that γ s is multiplied by a factor of 3 in (6.21) instead of the factor of 6 in (6.20), since the MQAM constellation
splits its total average energy γ s between its in-phase and quadrature branches. The probability of symbol error for
the MQAM system is then
√ r !!2
2( M − 1) 3γ s
Ps = 1 − 1− √ Q . (6.22)
M M −1
The nearest neighbor approximation to probability of symbol error depends on whether the constellation point is
an inner or outer point. Inner points have four nearest neighbors, while outer points have either two or three nearest
neighbors; in both cases the distance between nearest neighbors is 2d. If we take a conservative approach and set
the number of nearest neighbors to be four, we obtain the nearest neighbor approximation
r !
3γ s
Ps ≈ 4Q . (6.23)
M −1
For nonrectangular constellations, it is relatively straightforward to show that the probability of symbol error
is upper bounded as
" r !#2 r !
3γ s 3γ s
Ps ≤ 1 − 1 − 2Q ≤ 4Q , (6.24)
M −1 M −1
which is the same as (6.23) for square constellations. The nearest neighbor approximation for nonrectangular
constellations is
 
dmin
Ps ≈ Mdmin Q √ , (6.25)
2N0
where Mdmin is the largest number of nearest neighbors for any constellation point in the constellation and dmin is
the minimum distance in the constellation.

Example 6.3: For 16-QAM with γb = 15 dB (γs = log2 M · γb ), compare the exact probability of symbol error
(6.22) with (a) the nearest neighbor approximation (6.23) and (b) the symbol error probability for 16-PSK with the
same γb (which was obtained in Example 6.2).

190