Information Technology Reference
In-Depth Information
Table 1. Main Proposed Link Budget Parameters
where, E s is the bit energy, N 0 is the noise PSD,
ρ pv (.) is the normalized cross-correlation of the
received pulse and the template waveform, τ e is
the timing error, and ρ pp (.) is the normalized
auto-correlation of the received pulse.
The TOA ranging approach is based on the
estimation of the arrival time of the first detected
path. Typically, optimum detection involves the
correlation of the received waveform with a lo-
cally generated template waveform, which adds
to the receiver complexity (Arslan, Chen et al.
2006). However, this requirement is precluded
for non-coherent receivers, where the detection
process depends solely on the received pulses
(Chao and and Scholtz 2003; Arslan, Chen et al.
2006). One promising low-power non-coherent
receiver is the energy detection (ED) correlation
receiver, where the correlation is replaced by a
squaring device. Generally speaking, EDs con-
sume smaller amounts of power as compared to
coherent stored-reference receivers since they do
not require template generation. However, this
simplification is traded for degradation in the
estimation performance. Assuming the ED detec-
tor, and corresponding pulse-width 0.8 ns, the
corresponding B.W. is 2 GHz for the seventh-
order Gaussian pulse (Tuan-Anh, Krizhanovskii
et al. 2007), the maximum achievable ranging
accuracy is 6.9 cm for an integration window that
is equal to the pulse-width. This is because the
ED estimator exhibits a maximum achievable
accuracy equal to T s / 12 , where T s is the inte-
gration window. This precludes the choice of ED
detectors. Obviously, for such a high target rang-
ing accuracy, 1 mm ranging accuracy, required
for accurate human locomotion tracking, the MF
seems an appropriate choice, where its perfor-
mance approaches the CRLB at high SNRs.
For the suboptimal template, the minimum
attainable probability of error is:
Parameter
Value
Pulse Rate ( R p )
50 Mp/s
P t (transmitted power in dB
relative to a W)
-8.3 dBm
B.W min (bandwidth)
2 GHz
PSD(dBm/MHz)
-41.3 dBm/MHz
Receiver Noise Figure ( NF )
10 dB
N0 (Noise PSD = kTsys)
-164.4 dBm/Hz
PL 0 (path loss at reference
distance (d0))
44.6 dB
PL (d)(path loss at distance
(d))
31 dB
PL t (total path loss)
75.6 dB
G t and G r (Tx and Rx antenna
gains)
0 dBi
Implementation Loss ( L a )
3 dB
Average received power at
the Rx. ( P R )
-83.9 dBm
Average noise power ( P N )
-114.9 dBm (Eff. Bit-rate =
R b 90 kb/s)
Achieved E b /N 0
31 dB
Required E b /N 0 | req (dB)
18 + 3 = 21 dB
Link-margin ( LM )
10 dB
-93.9 dBm
Receiver sensitivity ( S r )
2
t
d
dt
( )
n
1
2
2
p t
( ) =
e
2
σ
(15)
n
n
2
πσ
Assuming a correlation receiver, the optimal
template v(t) should be matched to the received
pulse p t
( ) = ( ) , where the pulse parameters
are chosen to meet the specified FCC allowable
emission limits. When using a suboptimal win-
dowed sinusoidal template, v t
p t
n
cos ω for
a window-length T and carrier frequency ω c , the
oscillator frequency should be chosen to maximize
the output SNR (Sangyoub 2002):
( ) =
(
( ))
t
c
2
ρ
( )
(0)
τ
E
N
pv
e
SNR
=
s
(16)
ρ
0
pp
 
Search WWH ::




Custom Search