UMIACS/LTS SeminarMarch 17, 2004
Overview of Ultra Wide Band (UWB)Impulse Radio
Dr. Jay PadgettSenior Research Scientist
Telcordia Technologies/ Applied ResearchWireless Systems and Networks
2 What is Ultra Wide Band?� FCC definition of UWB signal: bandwidth is at least 20% of
center frequency, or a bandwidth of at least 500 MHz� Bandwidths can exceed 1 GHz� Example: Pulsed UWB signal (pulse width: 0.5 ns).
f (GHz)0 1 2 3 4 5
Φ(f
) / Φ
max
(dB
)
-50
-40
-30
-20
-10
0
τ = 0.5 ns
Signal is a sequence of short pulsesPulse energy spectrum
� The large UWB passband spans the operating frequencies of many existing licensed services� Interference concerns: UWB to narrowband systems (main
concern of FCC, NTIA), narrowband-to-UWB interference (UWB design issue)
3 Brief FCC / UWB History� FCC motivation: encourage development of potentially useful
new technology without causing harmful interference to existing services and devices� Sequence of FCC Notices and Orders
– Notice of Inquiry in September 1998– Notice of Proposed Rule Making, June 2000– First Report and Order, April 2002: made legal unlicensed operation
of UWB devices subject to certain restrictions– Reconsideration Order and Further NPRM, March 2003: responsive
to Petitions for Reconsideration filed in response to the first R&O
� FCC categorizes its approach to legalizing UWB “conservative” with respect to preventing harmful interference to other devicesand services. FCC has promised further evolution to UWB Rules� Allowed UWB emission levels are less than or equal to the level
allowed for unintentional radiators such as computers and other electronic devices (-41.3 dBm/MHz).
4 FCC General UWB Emission Limits
Frequency in GHz
1 2 3 4 5 6 7 8 9 10 11 12
UW
B e
mis
sio
n li
mit
in d
Bm
/MH
z
-80
-70
-60
-50
-40
IndoorHandheld
UWB Passband (3.1 to 10.6 GHz)1.99 GHz
1.61 GHz
−75.3 dBm
−51.3 dBm−53.3 dBm
−61.3 dBm−63.3 dBm
−41.3 dBm
960MHz
5 Applications: DARPA’s NETEX Program
� Phase I: Gain understanding of effects of UWB system operation on existing narrowband spectrum users– Testing (NAVAIR/Pax River): 16 legacy system, 39 modes, 1600
tests completed.– Modeling/Simulation (Telcordia) details below– UWB propagation and channel modeling (Virginia Tech)
� Phase II: (DARPA/ATO BAA 03-20, Industry Day April 7 ‘03)– Push UWB physical layer design to the point where it is capable of
reliably supporting advanced LPD, ranging, location and networking protocols
– Develop algorithms, protocols, and distributed control for robust, scalable ad hoc UWB networking
– Demonstrate 20-node hand held radios, 50-node video network, and 50-node radar sensor integrated into high data rate short range network.
Objective: Develop military UWB sensors and communications systems for operation in extreme environments
http://www.darpa.mil/ato/solicit/NETEX/documents.htm
6 UWB: An Undesired Result
7 UWB Interference Modeling
( )fG
( ) ( )∑∞
−∞=
−=k
kTtgtx ( )ty
victim receiver or measurement instrument (e.g., spectrum analyzer) impulse response (baseband equivalent)UWB impulse train receiver
response
( )thb
?t
( )tg
f
( )fH
f
tWhat is the effective interference power due to the UWB signal, as seen by the non-UWB receiver?
8 Victim Receiver Model and Interference Filtering
f0 f−f0
Receiver Passband
UWB PulseEnergy Spectrum
RF IF1
~ LO1
f0 fIF1IF2
~ LO2
fIF2Detector/DemodulatorCircuitry
HIF(f )
UWB Pulse Sequence IF Impulse Response (envelope shown)
9 Time Hopping PSD Example: Partial-Frame 2-ns Dithering
f0 (filter center frequency), GHz
0 1 2 3 4 5
filte
r out
put p
ower
, dB
m
-60
-50
-40
-30
-20
-10
M = 10 time hopping binsεc = 2 ns bin separationPuwb = 10 dBmR = 10 MHz pulse rateBIF = 1 MHz IF bandwidth
10 Closeup – with Simulation Results
f0 (filter center frequency), GHz
1.4 1.5 1.6
filte
r out
put p
ower
, dB
m
-60
-50
-40
-30
-20
-10
ModelSimulation
M = 10εc = 2 nsPuwb = 10 dBmR = 10 MHzBIF = 1 MHz
11 Simulation – Zero Span – 1.46 GHz
time in microseconds
0 20 40 60 80 100
filte
r out
put p
ower
in d
Bm
-60
-50
-40
-30
-20
-10
M = 10, εc = 2 ns, Puwb = 10 dBmR = 10 MHz, BIF = 1 MHzf0 = 1.46 GHz
f0 (filter center frequency), GHz
1.4 1.5 1.6
filte
r out
put p
ower
, dB
m
-60
-50
-40
-30
-20
-10
ModelSimulation
M = 10εc = 2 nsPuwb = 10 dBmR = 10 MHzBIF = 1 MHz
12 Simulation – Zero Span – 1.47 GHz
time in microseconds
0 20 40 60 80 100
filte
r out
put p
ower
in d
Bm
-60
-50
-40
-30
-20
-10
M = 10, εc = 2 ns, Puwb = 10 dBmR = 10 MHz, BIF = 1 MHzf0 = 1.47 GHzf0 (filter center frequency), GHz
1.4 1.5 1.6
filte
r out
put p
ower
, dB
m
-60
-50
-40
-30
-20
-10
ModelSimulation
M = 10εc = 2 nsPuwb = 10 dBmR = 10 MHzBIF = 1 MHz
13 Simulation – Zero Span – 1.48 GHz
time in microseconds
0 20 40 60 80 100
filte
r out
put p
ower
in d
Bm
-60
-50
-40
-30
-20
-10
M = 10, εc = 2 ns, Puwb = 10 dBmR = 10 MHz, BIF = 1 MHzf0 = 1.48 GHzf0 (filter center frequency), GHz
1.4 1.5 1.6
filte
r out
put p
ower
, dB
m
-60
-50
-40
-30
-20
-10
ModelSimulation
M = 10εc = 2 nsPuwb = 10 dBmR = 10 MHzBIF = 1 MHz
14 Simulation – Zero Span – 1.5 GHz
time in microseconds
0 20 40 60 80 100
filte
r out
put p
ower
in d
Bm
-60
-50
-40
-30
-20
-10
M = 10, εc = 2 ns, Puwb = 10 dBmR = 10 MHz, BIF = 1 MHzf0 = 1.5 GHz
f0 (filter center frequency), GHz
1.4 1.5 1.6
filte
r out
put p
ower
, dB
m
-60
-50
-40
-30
-20
-10
ModelSimulation
M = 10εc = 2 nsPuwb = 10 dBmR = 10 MHzBIF = 1 MHz
15 Example: PSD of a Swept-PRF Signal
t
r(t)
t
r(t)pulse rate
rd
2T
2T
−
r
( )22T
tT
atrtr ≤<−+=
Tra d2=
( ) ( )∑+−=
−=M
MnnTttv
1
δ
( )a
TnarrTn
42 22 +++−=
( ) ∑+−=
−=M
Mn
fTj nefV1
2π
16 Swept PRF PSD and Approximation
frequency (MHz)
0 200 400 600 800 1000
mill
iwat
ts
0
1
2
3
4
exactfirst-order analytic approximation
Swept PRF100 MHz average PRF100 kHz sweep rate4 MHz PRF span (98 to 102 MHz)1 watt total average power1.05 GHz pulse bandwidth (rectangular spectrum)
17 Closeup of Swept PRF Spectrum
frequency (MHz)
580 590 600 610 620
mill
iwat
ts
0.0
0.2
0.4
0.6
0.8
1.0
exactfirst-order analytic approximation
Swept PRF100 MHz average PRF100 kHz sweep rate4 MHz PRF span (98 to 102 MHz)1 watt total average power1.05 GHz pulse bandwidth (rectangular spectrum)
18 Individual Tones of the Swept PRF PSD
frequency (MHz)
585 586 587 588 589 590 591 592 593 594 595
mill
iwat
ts
0.0
0.2
0.4
0.6
0.8
1.0
exactfirst-order analytic approximation
Swept PRF100 MHz average PRF100 kHz sweep rate4 MHz PRF span (98 to 102 MHz)1 watt total average power1.05 GHz pulse bandwidth (rectangular spectrum)
19 UWB Effects on Receiver Performance� is the ratio of the pulse rate to the IF bandwidth� There are 3 primary interference cases:
– , which gives pulsed interference to the detector– with constant pulse rate, which can give a tone
within the passband– with dithering or modulation, which can give a
noise-like signal� The first two cases were explored for fixed-frequency
digital communications receivers and for an analog FM receiver. The third case, in the limit, gives results similar to those of Gaussian noise, which are well-known.�With Nu > 1, there can also be a combination of a tone and
noise (as seen previously).� Results are calculated based on the average UWB
interference power within the receiver passband.
uN
1≤uN1>uN
1>uN
20 Modeling the Test Procedure for an FM Receiver
Carrier-to-interference ratio (CIR), dB
-10 0 10 20 30
Bas
eban
d SI
NA
D (d
B)
0
10
20
30
40
Nu = 0.1
Nu = 1.0
Nu = 0.01
Gaussian noise
∆f = 2.4 kHzfm = 1 kHzBIF = 30 kHzBbb = 3 kHz
Nu = PRF ÷ IF bandwidth
21 Analysis vs. dB-Average Test Results (FM)
Test Waveform (TW) Number
1 2 3 4 5 6
CIR
for
10 d
B S
INA
D r
ecei
ver
upse
t, dB
-5
0
5
10
15
20
dB average from measurementsanalysis
Gaussian noise
22 UWB into a Noncoherent FSK Receiver
-20 -10 0 10 20 3010
-4
10-3
10-2
10-1
100
SIR (dB)
BE
RNu=0.01
Nu=0.02
Nu=0.05
Nu=0.10
Nu=1.00
−
2exp
21
BER AWGN Equivalent
SIR
-20 -10 0 10 20 3010
-4
10-3
10-2
10-1
100
SIR (dB)
BE
RNu=0.01
Nu=0.02
Nu=0.05
Nu=0.10
Nu=1.00
-20 -10 0 10 20 3010
-4
10-3
10-2
10-1
100
-20 -10 0 10 20 3010
-4
10-3
10-2
10-1
100
SIR (dB)
BE
RNu=0.01
Nu=0.02
Nu=0.05
Nu=0.10
Nu=1.00
−
2exp
21
BER AWGN Equivalent
SIR
23 Coherent PSK – Comparison of Rates
-5 0 5 10 15 10
-6
10 -5
10 -4
10 -3
10 -2
10 -1
10 0
SINR (dB)
BE
R
I/N = -30 dB I/N = 0 dB I/N = 30 dB
25.0 :Red =uN
1.0 :Green =uN05.0 :Blue =uN
u
u
NN
log10 SINR BER, lowAt 2BER SINR, lowAt
−≅≅
24 UWB/Legacy NB Coexistence Example
Rb
UWB data rate (kb/s)
1 10 100 1000
d uwb
Max
imum
UW
B p
ath
dist
ance
(met
ers)
0
200
400
600
800
1000
I/N = 0 dBI/N = 5 dBI/N = 10 dBI/N = 15 dBI/N = 20 dB
f0 = 300 MHzBuwb = 200 MHzdnb = 20 metersFnb = 10 dBFuwb = 1 dB(Eb/N0)uwb = 10 dBA(f0) = 0 dBGTX,uwb = 2 dBGRX,uwb = 2 dBGTX,uwb->nb = 2 dBGRX,nb = 2 dBLTX = LRX = 1 dBDithered pulses
25 Aggregate Interference – Closed form CDF
( ) ( ) ( ) ( ) ( )∑∞
=
>−
−ΓΓ
−=<=1
0,1sin1
!1
1Prk
k
Z zkzk
kzZzF νπ
ννπ ν
( ) ( ) )interferer(nearest 0exp >−= − zzzFZν
10 log z
-10 0 10 20 30P
{ Z
< z
} [
perc
ent]
Dis
trib
utio
n of
nor
mal
ized
ag
greg
ate
inte
rfer
ence
pow
er
0.1
1
10
305070
90
99
Multiple interferers
Nearest single interferer
γ = 4.0γ = 3.5γ = 3.0
• Uniformly random distribution of interfering transmitters• Equal power• No exclusion zone
26 IS-95 PCS Blocking vs. UWB Density
ρ (active UWB transmitters/m2)
0.00 0.05 0.10 0.15 0.20
Pb (
perc
ent)
blo
ckin
g pr
obab
ility
aver
aged
ove
r cel
l are
a an
d U
WB
-to-
PCS
dist
ance
0
20
40
60
80
100
Non-uniform distance distributionγ = 3.5
ΦTXuwb = −50 dBm/MHz
−55 dBm/MHz
−60 dBm/MHz
−65 dBm/MHz
ρ (active UWB transmitters/m2)
0.00 0.05 0.10 0.15 0.20
Pb (
perc
ent)
blo
ckin
g pr
obab
ility
aver
aged
ove
r cel
l are
a an
d U
WB
-to-
PCS
dist
ance
0
20
40
60
80
100
Uniform distance distributionγ = 3.5
ΦTXuwb = −50 dBm/MHz
−55 dBm/MHz
−60 dBm/MHz
−65 dBm/MHz
r (meters)
0 1 2 3 4 5 6
f d(r)
prob
abili
ty d
ensi
ty fu
nctio
n fo
rdi
stan
ce to
nea
rest
UW
B tr
ansm
itter
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
ρ = 0.1 active UWB transmitters/m2
Non-uniform
Uniform
Truncated uniform (d0 = 1 meter)
Probability density functions for minimum UWB to PCS handset distance
These curves show the average percentage of handsets blocked on the downlink as a function of UWB deployment density and transmit power spectral density
27 C/I with Aggregate Interference� Desired signal to NB receiver Rayleigh-faded (terrestrial
multipath).� UWB transmitters uniformly randomly distributed spatially, no
fading of interfering signals (short distances).� May be an “exclusion zone” surrounding victim receiver within
which there will be no UWB transmitters.
Desired NB transmitter
Rayleigh-faded signal path
exclusion zone(no UWB transmitterswithin this area)
dmin
narrowband receiver
28 C/I CDF with Aggregate UWB Interference and no Exclusion Zone
-40 -30 -20 -10 0 10 20
Pr( C
nb/ I
UW
B <
Λ),
perc
ent
0.1
1
10
30
50
70
90
99
99.9
99.99
γ = 2.5γ = 3γ = 3.5γ = 4
Monte Carlo, γ = 3Nearest interferer, γ = 3
dB , nbX
Λ
22
1exp1Pr
2
>
−Γ
Λ−−=
Λ< γ
γ
γ
nbUWB
nb
XIC
( ) 2 :ionnormalizat γπρα atx
nbnb P
CX =
29 The Larger Picture: UWB – Challenges, Questions, and Opportunities Going Forward
� Design of UWB radios that can operate adaptively in the presence of strong in-band narrowband signals.� UWB networking: MAC layer design, ad hoc routing protocols,
QoS management, especially with narrowband interference.� Applications – what is UWB best suited for (military and
commercial)?– Radar/localization– Data networking?– Sensors?– Integrated applications (e.g., localization + communications).
� Can UWB be used to provide high data rate hotspots with connectivity to CMRS networks?� Can UWB be used to increase the accuracy of E-911 indoors?