02279r0p802 15 sg3a channel model cont intel
TRANSCRIPT
-
8/8/2019 02279r0P802 15 SG3a Channel Model Cont Intel
1/33
24 June, 2002 IEEE P802.15-02/208r1-SG3a
IEEE P802.15Wireless Personal Area Networks
Project IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs)
Title UWB Channel Modeling Contribution from Intel
DateSubmitted
[24 June, 2002]
Source[Jeff Foerster and Qinghua Li][Intel Research and Development][JF3-2062111 N.E. 25th Ave.Hillsboro, OR 97124]
Voice: [503-264-6859]Fax: [503-264-3483]E-mail:
Re: [In response to the Call for Contributions on Ultra-wideband Channel Models
(IEEE P802.15-02/208r1-SG3a).]
Abstract [This contribution proposes a UWB path loss and multipath model for assisting inthe evaluation of possible UWB physical layer submissions for a high-rateextension to IEEE 802.15.3.]
Purpose [In this paper, we propose a method for standardizing link budgets to use incomparing different UWB PHY proposals for achieving the desired throughputsand ranges for the standard. In addition, we present some multipath channelmeasurements that were preformed by Intel and compare these measurements with
different channel models that have been considered by the industry for indoorchannels. The results suggest a possible UWB multipath channel model that couldbe 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) ororganization(s). The material in this document is subject to change in form andcontent after further study. The contributor(s) reserve(s) the right to add, amend orwithdraw 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 Page 1 Jeff Foerster and Qinghua Li, Intel
-
8/8/2019 02279r0P802 15 SG3a Channel Model Cont Intel
2/33
24 June, 2002 IEEE P802.15-02/208r1-SG3a
UWB Channel Modeling Contribution from Intel
Introduction
The purpose of this contribution is to respond to the UWB channel modeling call forcontributions (IEEE P802.15-02/208r1-SG3a) to assist the IEEE 802.15.3a study group inevaluating different UWB PHY proposals. The goal here is not to provide a universal model forall environments in which UWB devices will be operating, but rather to provide a set of toolsthat can be used to fairly evaluate the performance of different UWB PHY proposals in realisticchannels.
Path Loss and Link Budget Model
There have been many proposed path loss models in the literature (see [1]-[8] for examples), butthe purpose of this channel model is to fairly compare different physical layer proposals at thetarget operating distances. This can be done by simply adopting the free space path loss modeland asking the proposers to provide the resulting link margin that will be available to make upfor additional channel losses, implementation losses, waveform distortion, imperfect multipathenergy capture, etc. Note that the IEEE 802.15.2 indoor path loss model, for example, which isa bifurcated model with free space loss out to 8 meters and 3.3 path loss exponent greater than 8meters, only differs by 1.26 dB from free space at the target distance of 10 meters. Therefore,the simple free space path loss model will be close to the IEEE 802.15.2 path loss at the targetranges of interest. Of course, this does not include any losses that might be caused by propagation through walls, furniture, or other obstacle. The table below identifies theparameters needed by the proposer and how those parameters could be used to compute the final
link margin. This model is based on the narrowband path loss calculations (known as the Friistransmission formula), and the applicability of this model to UWB systems depends on theassumptions about the frequency response of the antennas, which is discussed next.
A justification of using the narrowband path loss model for UWB systems was actually presented in [9], and is replicated here for completeness. Assuming perfect isotropicallyradiating antennas at the transmitter and receiver, the received power as a function of frequency,can be expressed as the following:
22
2
)4(
)()()()(
fd
cfGfGfPfP RTTR
= (1)
where )(fPT is the average transmit power spectral density (
= dffGfPP TTave )()( is the
total average transmit power), c is the speed of light, and )( fGT and )( fGR are the transmit
and receive antenna frequency response, respectively. Clearly, this depends on the frequencyresponse of the antennas, which may be difficult to generalize. However, since the FCCregulations for UWB requires the transmitter to meet a certain electric field strength limit (500uV/m for UWB systems operating between 3.1 and 10.6 GHz) at a specified range (3 m), which
Submission Page 2 Jeff Foerster and Qinghua Li, Intel
-
8/8/2019 02279r0P802 15 SG3a Channel Model Cont Intel
3/33
24 June, 2002 IEEE P802.15-02/208r1-SG3a
is equivalent to a total transmit power spectral density limit, then it would desirable to have the
product )()( fGfP TT be flat within the bandwidth of interest. On the other hand, it may be
difficult to generalize that the receiving antenna will be flat across the desired frequency band.However, as a first order approximation, a flat frequency response, isotropic antenna is
considered next. So, for a perfectly flat UWB waveform occupying the band 2/Wfc to2/Wfc + with power spectral density aveP /W, and a flat frequency response of the receiving
antenna with constant gain across the whole bandwidth ( RG ), the total average received power at
the output of the receiving antenna will be given by the following [9]:
( )
( )
=
=
+
==
+
2
222
2
2
22/
2/
2/1
1
2/1
1
)4(2/
1
2/
1
)4()(
c
NB
aveRave
cc
Rave
cc
Rave
Wf
Wf
RRave
fWPP
fWfd
cGP
WfWfdW
cGPdffPP
c
c
(2)
where
22
2
)4( c
RaveNB
avefd
cGPP
= (3)
corresponds to the well known narrowband path loss model equation and the second termaccounts for the difference between the narrowband and wideband model. For the largest
fractional bandwidth allowed by the FCC (occupying 3.1-10.6 GHz), RaveP will differ fromNB
aveP
by only 1.5 dB, and this difference becomes smaller for smaller fractional bandwidths. Alsonote that the FCC rules results in W
-
8/8/2019 02279r0P802 15 SG3a Channel Model Cont Intel
4/33
24 June, 2002 IEEE P802.15-02/208r1-SG3a
ranges desired for the standard. The highlighted parameters below are up to the proposers todefine, while all other parameters will be consistent with all proposals so easy comparisons canbe made.
Table 1: Medium Rate Link Budget
Parameter Value Value
Throughput > 110 Mbps > 200 Mbps
Average Tx power ( TP ) dBm dBm
Tx antenna gain ( TG ) 0 dBi 0 dBi
cf : center frequency of waveform Hz Hz
Path loss at 1 meter ( )/4(log20 101 cfL c= )8
103=c m/s
dB dB
Path loss at dm ( )(log20 102 dL = ) 20 dB at d=10meters
12 dB at d=4meters
Rx antenna gain ( RG ) 0 dBi 0 dBi
Rx power ( 21 LLGGPP RTTR ++= (dB)) dBm dBm
Noise Bandwidth at antenna port (W) Hz Hz
Noise power ( )(log*10174 10 WN += ) dBm dBm
Rx Noise Figure ( FN ) 7 dB 7 dB
Rx Noise Power ( FN NNP += ) dBm dBm
Processing gain ( GP : please explain how this is
derived)
dB dB
Minimum C/N (S) dB dB
Link Margin ( SPPPM NGR += ) dB dB
Proposed Min. Rx Sensitivity Level dBm dBm
Although the proposers may need to alter the above table for their specific UWB PHY proposal,it gives them an initial framework of the kind of justification and detail that should be part of theproposal. The final desired output of the link budget should be a final Link Margin that will beneeded to account for additional channel losses, implementation losses, waveform distortion,imperfect multipath energy capture, amplitude fading, etc.
Multipath Model
Although there are both frequency domain and time domain models that may be appropriate forUWB systems, we chose to focus our work on evaluating discrete time models. This model isbased upon the following channel impulse response model:
)()(1
0
l
L
l
l tth =
=
(5)
where l is the amplitude fading factor on path l(could be complex), l is the random delay
Submission Page 4 Jeff Foerster and Qinghua Li, Intel
-
8/8/2019 02279r0P802 15 SG3a Channel Model Cont Intel
5/33
24 June, 2002 IEEE P802.15-02/208r1-SG3a
of path l,L is the number of multipath components, and )(t is the Derac delta function. Thereare several parameters that need to be defined to complete this particular model, and each will beaddressed in the following subsections.
RMS Delay SpreadTypical values for the multipath delay spread of indoor channels have been reported to bebetween 15 nsec in a residence to over 100 nsec in an office to a 150 nsec in a commercialbuilding [10]. Other measurements at 10 meter distances suggest RMS delay spreads of 19-47nsec [17]. In addition, the multipath delay spread has been found to increase as the separationdistance between the receiving and transmitting antenna is increased. The following table showssome of the published RMS delay spread numbers that have been suggested for the indoorchannel (both industry adopted models and published academic papers).
Ref. Application Delay spread Comments
[10] WPAN (ITU P.1238) RMS values:
70 nsec for Residential100 nsec for office150 nsec for commercial
WSSUS model with tap-
delay line and Gaussiandistributed taps
[11] 802.11 LAN for evaluatingHRb proposals
25 nsec100 nsec250 nsec
WSSUS model with tap-delay line and Gaussiandistributed taps with zeromean (Rayleigh fading)
[12] IEEE 802.15.3 High ratePAN
25 nsec minimum WSSUS model with tap-delay line and Gaussiandistributed taps with zeromean (Rayleigh fading)
[13] Indoor at distances up to30 meters (results here for10 meters)
< 20 nsec for LOS< 70 nsec for NLOSfor 2.4 GHz
Delay spreads for 2.4GHz tends to be higherthan 11.5 GHz. Delayspread increase withdistance separation.
[14] Indoor at ~ 1.5 GHz Ave. rms delay spreads:Brick: 26-30 nsConcrete: 28-29 nsOffice: 25 & 50 nsLOS factory: 96 nsOBS factory: 105 ns
Max rms delay spreads:Brick: < 70 nsConcrete: < 70 nsOffice: 50 & 218 nsLOS factory: 300 nsOBS factory: 300 ns
[15] UWB propagation indoor