Transcript No Slide Title - IEEE Standards Working Group Areas

doc.: IEEE 802.15-02/287
July 2002
Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs)
Submission Title: [UWB Channel Modeling Contribution from Intel]
Date Submitted: [1 July, 2002]
Source: [Jeff Foerster and Qinghua Li ]
Company [Intel Research and Development]
Company [Intel Corporation]
Address [JF3-212, 2111 N.E. 25th Ave., Hillsboro, OR, 97124]
Voice [503-264-6859], FAX: [503-264-3483]
E-Mail:[[email protected]]
Re: [In response to the Call for Contributions on Ultra-wideband Channel Models (IEEE P802.15-02/208r1SG3a).]
Abstract: [This contribution proposes a UWB path loss and multipath model for assisting in the evaluation of
possible UWB physical layer submissions for a high-rate extension to IEEE 802.15.3. ]
Purpose: [In this presentation, we propose a method for standardizing link budgets to use in comparing different
UWB PHY proposals for achieving the desired throughputs and ranges for the standard. In addition, we present
some multipath channel measurements that were preformed by Intel and compare these measurements with
different channel models that have been considered by the industry for indoor channels. The results suggest a
possible UWB multipath channel model that could be used to compare different UWB PHYs.]
Notice:
This document has been prepared to assist the IEEE P802.15. It is offered as a basis for discussion and is not binding
on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after
further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein.
Release:
The contributor acknowledges and accepts that this contribution becomes the property of IEEE and may be made
publicly available by P802.15.
Submission
Slide 1
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
UWB Channel Modeling
Contribution from Intel
Jeff Foerster and Qinghua Li
Intel Research and Development
Intel Corporation
Submission
Slide 2
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Overview of Talk
• Summary of channel measurements
• Link budget and path loss model
• Multipath model comparisons
– Path amplitude distribution
– Discrete time multipath models
• Proposed multipath model for comparing UWB
PHYs
Submission
Slide 3
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Channel Measurements
• Performed by: Robbie Adler, David Cheung, Evan Green,
Minnie Ho, Qinghua Li, Cliff Prettie, Leslie Rusch, Keith Tinsley
• 870 channel realizations between 1-20 m.
• Pulse based and frequency based measurements
Pulser
HPF
5th order
Bessel
2 MHz clock
Channel
Submission
Rx Ant.
UWB LNA
UWB LNA
Pulsicom or
Biconic
ZRON 8G
23 dB
ZRON 8G
23 dB
Slide 4
PA
ZVE-8G
34 dB
Tx Ant.
Biconic
TDS8000
DSO
--12.5 GHz low-noise head
--200 kHz sampling
--25 ps /sample
--4000 samples
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Channel Measurements
• Direct sounding of the time-varying channel transfer
function
• Various filters were used with the goal of providing
measurements equivalent to pulse measurement
using the oscilloscope
Network Analyzer
Tx (Port 1)
HPF
5th order
Bessel
Agilent 8720ES
20 GHz capability
Channel
Submission
Rx Ant.
UWB LNA
UWB LNA
Pulsicom or
Biconic
ZRON 8G
23 dB
ZRON 8G
23 dB
Slide 5
PA
ZVE-8G
34 dB
Tx Ant.
Biconic
Network Analyzer
Rx (Port 2)
-- 2-8 GHz range typical
-- 1600 pts.
-- 16 averages
-- 10 seconds/record
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Channel Measurements
Submission
Slide 6
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Channel Measurements
Mean Excess Delay (ns)
bin size
CB,LOS
CB,LOS+NLOS
CB,NLOS
0.17
3.95
13.50
17.25
0.33
3.86
13.42
17.17
0.67
3.70
13.25
17.00
1.60
3.32
12.82
16.55
Tau RMS (ns)
bin size
CB,LOS
CB,LOS+NLOS
CB,NLOS
0.17
9.13
13.34
15.00
0.33
9.13
13.34
15.00
0.67
9.13
13.34
15.00
1.60
9.11
13.34
15.00
Number of paths within 10 dB
bin size
CB,LOS
CB,LOS+NLOS
CB,NLOS
Submission
0.17
7
28
37
0.33
5
19
24
Slide 7
0.67
4
13
16
1.60
3
8
9
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Link Budget and Path Loss
• Goal: Provide a consistent way to compare the
link budget of different PHYs to ensure the
throughput and range requirements are
properly met.
• Issues
– How to make this flexible to accommodate different
UWB waveforms (bandwidths, etc.)?
– How to make this ‘generic’ that is somewhat
independent of implementation (antenna patters,
etc.)?
Submission
Slide 8
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Link Budget and Path Loss
• Approach: Is the Friis formula applicable
(in any sense) for UWB waveforms
[QCOM FCC comment]?
– Received power spectral density:
PT ( f )GT ( f )GR ( f )c 2
PR ( f ) 
(4d ) 2 f 2
– Received average power
PRave 
Submission
f c W / 2
 P ( f )df
R
f c W / 2
Slide 9
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Link Budget and Path Loss
• Assume fixed antenna gain with frequency
PRave


1
NB
 Pave 
2 
1  W / 2 f c  
NB
Pave
Pave G R c 2

(4d ) 2 f c2
– At most 1.5 dB from Friis narrowband model
• Assume increasing antenna gain with
frequency (fixed effective aperature)
GR ( f )  4AR f 2 / c 2
PRave
Submission
Pave 4AR  Pave c 2  4AR f c2 
NB







P
ave G R ( f c )
2
2
2 
2


(4d )
 (4d ) f c  c

Slide 10
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Link Budget and Path Loss
• Comparisons with measurement loss at 1 meter
– AT&T results: 47 dB LOS (4.375-5.625 GHz, 5 GHz
center freq.)
– Intel results: ~47 dB LOS (2-8 GHz, 5 GHz center freq.)
– 46.4 dB predicted by Friis at 5 GHz center freq.
• Results suggest Friis formula can be used to
predict 1 meter path loss.
• Path loss model: Free space + ‘link margin’
– Proposers to provide ‘link margin’
– What should minimum ‘link margin’ be?
Submission
Slide 11
Intel Research and Development
July 2002
Link Budget and Path Loss
doc.: IEEE 802.15-02/287
Parameter
Throughput
Average Tx power (PT )
Tx antenna gain (GT )
f c : center frequency of waveform
Value
> 110 Mbps
dBm
0 dBi
Hz
Value
> 200 Mbps
dBm
0 dBi
Hz
Path loss at 1 meter ( L1  20 log 10 (4f c / c) )
c  3  10 8 m/s
Path loss at d m ( L2  20 log 10 (d ) )
dB
dB
Rx antenna gain (G R )
Rx power ( PR  PT  GT  G R  L1  L2 (dB))
Noise Bandwidth at antenna port (W)
Noise power ( N  174  10 * log 10 (W ) )
N
Rx Noise Figure ( F )
Rx Noise Power ( PN  N  N F )
20 dB at d=10 12 dB at d=4
meters
meters
0 dBi
0 dBi
dBm
dBm
Hz
Hz
dBm
dBm
7 dB
dBm
Processing gain (PG : please explain how this is dB
derived)
Minimum C/N (S)
dB
dB
Link Margin ( M  PR  PG  PN  S )
dBm
Proposed Min. Rx Sensitivity Level
Submission
Slide 12
7 dB
dBm
dB
dB
dB
dBm
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Multipath Model Comparisons
• We compared the following models to the
measurements:
– Path amplitude distributions
• Rayleigh
• Log-normal
– IEEE 802.11 model with Rayleigh fading
– Saleh-Valenzuela (S-V) Model
– -K Model
Submission
Slide 13
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Amplitude Distribution
• Found the ML estimate of the model
parameters to match the measurement data
at particular delays
• Used the Kolmogorov-Smirnov test with 1%
significance level (one figure of merit)
LOS, 1-20 m
NLOS, 1-20 m
LOS, 1-5 m
NLOS, 1-5 m
LOS, 5-20 m
Submission
Pass Rate of
Pass Rate of
Lognormal (%)
Rayleigh (%)
55
0
25
0
30
0
30
0
50
25
Slide 14
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Amplitude Distribution
• Comparison of distributions (1-20m)
LOS
Submission
NLOS
Slide 15
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Amplitude Distribution
• Standard deviation of log-normal variable
– Mean of 4.8 dB (relatively constant with excess delay)
– Similar to other indoor channel models (Hashemi93)
Submission
Slide 16
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Amplitude Distribution
• Observations
– Rayleigh distribution of each multipath arrival does not
appear to match measurement data as well as Lognormal…did not test other distributions
– Std of log-normal random variable similar to other
published results
– Seems to agree with intuition for UWB signals
• Smaller number of paths arriving in small ‘bin’ time (167 psec),
so central limit theorem no longer valid to justify complex
Gaussian amplitudes and Rayleigh envelopes
• Paths experiencing several reflections will be multiplied by
several reflection coefficients, so log of path amplitude will be
sum of independent random variables and ~ Gaussian via
central limit theorem
Submission
Slide 17
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Multipath Model Comparison
• Selected three important channel
characteristics to use for comparisons
between model and measurements
– Mean excess delay (tm (ns) )
– RMS delay spread (trms (ns) )
– Mean number of significant multipath arrivals
within 10 dB of peak arrival (Mean NP10dB )
• For S-V and -K Models, used brute force
search of model parameters to match
channel characteristics
Submission
Slide 18
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Multipath Model Comparison
• IEEE 802.11 model
– TDL with 167 psec sample time spacing + Rayleigh fading
– Simulations of model resulted in > 2x number of paths for
fixed RMS delay spread compared to measurements
Submission
Simulated
trms (ns)
Mean
NP10dB
Measured
trms (ns)
Mean
NP10dB
5
10
12.94

20

31
53
65
73
91
109
12.94
33
Slide 19
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Multipath Model Comparison
• Saleh-Valenzuela (S-V) Model
– Based on clustering of arrivals with double exponential
multipath intensity profile (MIP)
– Assumes Rayleigh amplitude distribution
Path Magnitude
tm (ns)
13.53
trms (ns)
13.41
Simulated Mean NP10dB 34
13
Time
T1
cluster 0
Submission
t
13
tm (ns)
13.59
trms (ns)
12.94
Measured Mean NP10dB 33
Slide 20
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Multipath Model Comparison
• -K Model
– Single exponential MIP model with two-state Marchov
process for random arrival probability
– Log-normal fading with 4.8 dB std.
NLOS+LOS
LOS only
tm (ns)
13.13
tm (ns)
4.11
trms (ns)
12.82
trms (ns)
4.21
Simulated Mean NP10dB 7
Simulated Mean NP10dB 34
tm (ns)
13.59
tm (ns)
4.01
trms (ns)
12.94
trms (ns)
8.88
Measured Mean NP10dB 33
Submission
Measured Mean NP10dB 7
Slide 21
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Multipath Model Comparison
• Observations
– S-V Models appears to have the ability to model
both LOS and NLOS channels with proper model
parameters (flexibility of double exponential MIP
model)
– -K Model can accurately model NLOS channels,
but difficult to match LOS characteristics
– IEEE 802.11 model is not appropriate for UWB
channels
Submission
Slide 22
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Observations of channel characteristics from measurements
– Channel appears to have clusters
Submission
Slide 23
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Propose a model based on the S-V model with
– Log-normal amplitudes (not Rayleigh)
• Key model parameters
–  = cluster arrival rate
–  = ray arrival rate, i.e., the arrival rate of path within each
cluster
–  = cluster decay factor
–  = ray decay factor
–  = standard deviation of lognormal fading term (dB)
Submission
Slide 24
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Discrete time model of channel impulse response
h(t ) 
L
K

l 0 k 0
k ,l
 (t  Tl  t k ,l )
 k ,l  p k ,l  k ,l
pk ,l is equiprobable +/-1
20 log 10(  k ,l )  Normal ( k ,l ,  2 )
• Should amplitude be real or complex?
– What does complex mean for UWB pulses?
• Just corresponds to fine time delay
– Suggest using real coefficients
• Assumes Rx is synchronized (same as phase coherent)
– Paths will be independent with 167 psec min. path spacings
(corresponding to inverse of measurement BW)
Submission
Slide 25
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Model parameters for different channel conditions
– For channels not captured or reflected by measurement data
LOS*
4
Channel Characteristics
Mean excess delay (nsec) t( m )
RMS delay (nsec) (t rms )
NLOS*
17
NLOS#
22
NLOS# LOS#
3
27
15
20
25
5
9
NP10dB
35
40
45
4
7
1/22
1/0.94
7.6
0.94
4.8
1/60
1/0.5
16
1.6
4.8
Model Parameters
1/15
1/14
1/11
 (1/nsec)
1/0.32
1/0.33
1/0.35
 (1/nsec)
30
22
16

10
10
8.5

4.8
4.8
4.8
 (dB)
* Based on Intel measurements.
# Example of other possible channel characteristics to test.
Submission
Slide 26
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Procedure for generating channel realizations
1. Determine cluster arrival times ( T ) using the exponential inter-arrival time distribution
l
(stop when Tl  10 ). Round Tl to the next higher multiple of 167 psec to yield a
discrete model with minimum resolvable path spacings of 167 psec (corresponding to a 6
GHz waveform).
Determine ray arrival times ( t k ,l ) using the exponential inter-arrival time distribution
2. (stop when t
k ,l
 10 ). Round t k ,l to the next higher multiple of 167 psec to yield a
discrete model with minimum resolvable path spacings of 167 psec (corresponding to a 6
GHz waveform). Note that, it may be possible for a cluster arrivals and ray arrivals to
occur at the same time. When this occurs, we recommend just summing up the arrivals at
each time interval.
p Tl Tl 1    exp   Tl  Tl 1  ,



l 0

p t k ,l t ( k 1),l   exp   t k ,l  t ( k 1),l  ,
Submission
Slide 27
k 0
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Procedure for generating channel realizations
3. Determine pk ,l , with are equally likely +/- 1.
4. Determine  k ,l .
 k ,l 
 10Tl /   10t k ,l / 
ln( 10)

 2 ln( 10)
20
5. Determine  k ,l (log-normal random variable).
6. Determine  k ,l  pk ,l  k ,l .
2
L K
7. Compute the total energy of the channel ( E    k ,l ).
l 0 k 0
Submission
Slide 28
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Procedure for generating channel realizations
8. Normalize the total energy of the channel to unity:
h(t ) 
1
L
K


E
l 0 k 0
Submission
k ,l
 (t  Tl  t k ,l )
Slide 29
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Possible model options
– Changing minimum path resolution (changing BW)
• For T > 167 psec, re-sample impulse response at the
lower rate (LPF followed by re-sampler will account for
overlap of paths)
• For T < 167 psec (down to 133 psec =1 / (7.5 GHz)):
– Replace arrival times in impulse response procedure with times
that are multiples of desired T (will not account for possible
parameter changes as BW increases).
– Keep the 167 psec based channel sample rate, and interpolate to
lower sample times (may not account for possible larger number of
resolvable paths)
Submission
Slide 30
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Proposed Multipath Model
• Possible model options
– Normalization of multipath energy
• Rather than normalize each impulse response
realization…could normalize the ‘average’ over all
realizations generated by the model
– Use complex taps rather than real
• Propose to use independent, uniform phase with same
amplitude envelope generated by model
Submission
Slide 31
Intel Research and Development
doc.: IEEE 802.15-02/287
July 2002
Summary
Parameter
Path loss
Multipath model
Model
Free space propagation using example link budget model to yield
final link margin and proposed minimum sensitivity level.
Time domain multipath model, given by
L
K
h(t )    k ,l  (t  Tl  t k ,l )
l 0 k 0
Channel
Characteristics
Multipath Model
Parameters
Submission
with a double-exponential decay intensity profile based on the S-V
model.
Table 11 provides 5 different example channel characteristics that
represent both LOS and NLOS channels. These characteristics are
based on different mean excess delay, RMS delay spread, and mean
number of paths.
The proposed model requires defining the following parameters:
 = cluster arrival rate;
 = ray arrival rate, i.e., the arrival rate of path within each cluster;
 = cluster decay factor;
 = ray decay factor;
 = standard deviation of lognormal fading term (dB).
Table 11 provides values for different channel characteristics.
Slide 32
Intel Research and Development