Digital Signal Processing Reference
In-Depth Information
hefty price for most portable applications. The other alternative is that for the
same word length, an increasing number of bits is sacrificed to deal with the
overhead of the narrowband spur. In any practical receiver implementation,
this can be obtained by reducing the gain of the vga. The consequence is that
the noise floor in the unaffected frequency bands will rise, since the rms level
of the ofdm signal is now represented by a lower effective number of bits
(enob). Actually, the interferer reduces the sensitivity of the receiver in all
other subbands of the ofdm signal. The following example provides a rough
estimation of the consequences for the data rate in an 802 . 11a/g link.
Effects of interference on the sensitivity of 802.11a/g
Each subband in the 802 . 11g system supports an M-ary qam modulated sub-
carrier. The 802 . 11g standard [Wla07] specifies that the ofdm subcarriers can
be modulated either using bpsk, qpsk, 16-qam or 64-qam (Figure 1.8). The
total power allocated to a single subcarrier, however, remains constant over
all modulation depths. This implies that for increasing modulation depths, the
minimum squared euclidean distance (msed) between the constellation points
is reduced. Taking bpsk as the reference case (msed =
1), the following Eu-
clidean distances for the higher-order modulation depths are found (1.1):
Euclidean distances in 802 . 11a/g
:
msed =
=
bpsk
1
0 dB (reference case)
1 / 2
qpsk
:
msed =
=−
3dB
1 / 10
16-qam
:
msed =
=−
10dB
1 / 42
64-qam
:
msed =
=−
16dB
(1.1)
BPSK
QPSK
16-QAM
MSED = 1/
10
MSED = 1/√2
0010
0110
1110
1010
01
11
MSED = 1
0
1
0011
0111
1111
1011
reference case
0001
0101
1101
1001
00
10
0000
0100
1100
1000
Figure 1.8.
bpsk, qpsk and 16-qam encoding maps for 802 . 11a/g. In each constel-
lation, the msed is normalized so that the average transmitted power is
the same for all modulation depths.
Search WWH ::




Custom Search