Propagation measurements and models for wireless channels

PropagationMeasurements and Models
for WirelessCommunications Channels
To achieve ubiquitous PCS, new and novel waysof classifying wireless
environmentswill be needed that are bothwidely encompassingand
reasonablycompact.
Jargen Bach Andersen, Theodore 5. Rappaport, and Susumu Yoshida
JGRGEN BACHANDER-
SEN is u professor at Aalborh.
Universipand head of thc
Centerfor Personkommn-
nikation.
THEODORE S. RAPPA-
PORT 0 un associareprafes-
sor of electrical rngineenng at
Viwnia Tech.
SUSUMU YOSHIDA u
profawr of electricalengi-
neering at Kyoto Universip.
ireless personal communica-
tionscouldinprincipleusesev-
era1 physical media, ranging
from sound to radio tolight.
Sincewewant toovercome the
limitations of acoustical com-
munications, we shall concentrateon propagation
of electromagnetic wavcs in the frequency range
from some hundreds of MHz to a few GHz.
Although thereisconsiderable interest atthe moment
in millimeter wave communications in indoor
environments, theywillbe mentionedonlybrieflyin
this survey of propagation of signals.
Itisinteresting to observe that propagation results
influence personal communications systems in
severalways. First thereisobviouslythe distribution
ofmeanpoweroveracertainareaorvolumeofinter-
est, which is the basic requirement for reliable
communications. The energy should be sufficient
for the link in question, but not too strong,in
order not to create cochannel interfcrcnce at a
distance in another cell. Also, since the radio link
is highly variable over short distances, not only
the mean poweris significant; the statistical dis-
tribution is also important. This is especially true
when the fading distribution is dependent on thc
bandwidth of the signal. Secondly. evenif there is
sufficient power availablefor communications,
the qualityof the signalmay be such that large errors
occur anyway. This results from rapid movement
through thescatteringenvironment,or impairments
due to long echoes leading to inter-symbol-inter-
ference. A basic understanding of the channel is
important forfindingmodulation andcodingschemes
that improve thc channel, for designing equaliz-
ers or,if this is not possible, for deploying basc
station antcnnas in such away that the detrimen-
tal effects are lesslikely to occur.
In this article we will describe the type of sig-
nals that occur invarious cnvironments and the mod-
eling of the propagation parameters. Models are
essentially of two classes.The first class consists
of parametric statistical models thaton average
describcthephenomenonwithinagivenerror.They
are simple to use, but relativcly coarse.In the last
fewyearsa secondclass ofenvironment-specificmod-
e1shasbeenintroduced.Thesemodelsareofamore
deterministicnaturc,characterizingaspecificstreet,
building, etc. They are necessarily more time con-
sumingtouse,but arealsomore revealingconcerning
physical details and hopefully morc accurate.
Firstsome keyparametersandthemeasurement
ofthemwill be discussed and then the differentwire-
less environmentswill be treated. The latter topic
is divided here into outdoor environments, indoor
environments, and radio penetration from out-
door to indoor environments.
The Physics of Propagation
Themechanismswhich governradio propagation
are complex and diverse, andthey can general-
ly be attributed to threebasic propagation mech-
anisms: reflection, diliraction, and scattering.
Reflection occurs when a propagating electro-
magnetic wave impinges upon an obstruction
with dimensions very large compared to the
wavelength of thc radio wave. Reflections from
the surface of the earth and from buildings pro-
duce rcflcctedwaves that may interfere construc-
tively or destructively ata receiver.
Diffraction occurs when the radio path between
thetransmitterandreceivcrisobstructedbyanimpen-
etrablebody. BasedonHuygen’sprinciple,secondary
waves are formed behind the obstructing body
even though there is no line-of-sight (LOS)
between the transmitter and receiver. Diffraction
explains how radio frcquency(RF) energy can
travel in urban and rural environments without a
LOS path. This phenomenonis also called “shad-
owing,”becausethediffractedfieldcanreacharecciv-
er even when itis shadowed by an obstruction.
Scatteringoccurswhenthcradiochannelcontains
objects with dimensions that are on the orderof
thewavelengthor1cssofthepropagatingwave.Scat-
tering, which follows the same physical principles
as diffraction, causes energy from a transmitterto be
reradiatedin many different directions.Ithasproven
to be the mostdifficultof the three propagationmech-
anisms to predictin emerging wireless personal
communication systems. For example, in urban
microccllular systems, lamp posts and street signs
42 IEEE Communications Magazine January 1995
~~ -

~ ~~
scatter energy in many directions, thereby prowding
RF coverage to locations which might not rcccive
energyviareflectionordiffnction. The three mech-
anisms are illustrated in Fig. 1.
As a mobile radio moves throughout a cover-
age area, the three propagation mechanismshave
an impact on theinstantaneous received signal in
differentways. Forexample, ifthemobile hasaclear
LOS path to the base station, diffraction andscat-
tering arenot likely to dominate the propagation.
Likewise, if the mobile is at streetlevel in a large
metropolitan areawithout aLOSto thebase statlon.
diffraction and scattering aremost likely to domi-
nate thepropagati0n.k the mobilemovesoversmall
distances. thc instantaneous received narrowband
signal strength will fluctuate rapidly giving rise to
small-scale fading.The reasonfor this isthat thefield
isasumofmanycontributionscomingfromdifferent
directionsand since the phases are random, the sum
bchaves as a noise signal,Le.. Rayleigh fading. In
small-scale fading, thereceived signal power may
vary by as much as threeor four orders o f magni-
tude (30 or 30 dB) when the receiverIS moved by
only a fraction of a wavelength. As the mobile
moves awayfromthe transmitter over largerdistances,
the local average received signalwill gradually
decrease. Typically, the local average signalis
computed over receiver movementsof 5 to 40
wavelengths 11-31, Figure 2 demonstrates the
effects of small-scale fading andlarge scale signal
variation foran indoor radiocommunication system.
Notice in the figure that thesignal fades rapidly as
thereceivermoves,butthelocalave.eragesignalchanges
much mure slowly with distance.
In mobile radio systems, communications
engineers are generally concernedwith two main
radio channelissues: link budget and time disper-
sion. The link budget is determined by the amount
of received power that may expected at a particu-
lar distance orlocation from a transmitter, and it
determines fundamental quantities such as trans-
mitterpowerrequirements, coverage areas,andbat-
tery life. Time dispersion arises due to multipath
propagationwherebyreplicasof the transmittcdsig-
nal reach thereceiver with different propagation
delays due to the propagation mechanisms
described above. The time-dispersive nature of
the channel determines the maximum data rate
that may be transmitted without requiring equal-
ization and also determines the accuracyof navi-
gational servicessuch as vehicle location.
Propagation parameters
Path loss -Link budget calculations requirean
estimate of the power levelso that a signal-to-noise
ratio (SNR) or, similarly, a carrier-to-intcrference
(QI) ratio may be computed. Because mobile
radio systems tendto be interferencelimited (due
tootheruserssharing thesamechanne1)rather than
noise limited,the thermal and man-madenoise effects
are often insignificant compared to thesignal levels
of cochannel users. Thus, understanding the propa-
gation mechanismsin wireless systems becomes
important for notonly predictingawerage to a partic-
ular mobile user,but alsoforpredicting theinterfering
signalsthatuserwillexperiencefromotherRFsources.
Weusepathloss(PL) heretodenotethelocalaver-
age received signal power relative to the transmit
power. Thisisausefulquantity,since receivedpower
is usuallymeasured as a local spatial averagerather
A
W Figure 1. Sketch of three lmpottantpropagation mechanisms: rejlection (R).
scattering (S),diffraction (Dl.
I
W Figure 2. Tvpicalreceived simal levelsfor anindoor radio communication
sy'tem. ,'otic> that small .scak fading produceslevel changes of 20 db or more,
where the local average signallevelchanges much moreslowly with distance.
thananinstantaneousvalue.Inrealisticmob~leradio
channels, free space does notapply. A general PL
model that has been demonstrated through mea-
surements [l,4) uses a parameter, n, to denote
the powerlaw relationship between distance and
received power. As a function of distance, d, PL
(in decibels) is expressed as
PL(d) = PL(d0) + 10n log(d/do)+Xo (1)
where n = 2 for free space, andis generally high-
er for wireless channels.
ThetermPL(do)simplygivesPLataknownclose
in reference distance do which is in the farfield of
the transmitting antenna (typically 1 km for large
urban mobile systems, 100 m for microcell sys-
tems, and 1 m for indoorsystems) andX , denotes
a zero mean Gaussian random variable (with
units of dB) that reflects the variationin average
receivedpower that naturallyoccurswhenaPL model
of this type is used.SincethePL modelonlyaccounts
for the distancewhich separates the transmitter
IEEE CommunicationsMagazme January 1995 43

140
130
120
1
0 110
f loo
2
m
I
-
90
88
All measurement locations n=A
.. .
70
0.1 1 2 3 4
10
T- R separation tkm)
n=3
n=2
n=l
HFigure 3. Scatterplot ofpath loss vs. distancefor measurements in five Ger-
man cities. Thepath loss isreferenced to afree space reference measurement at
do = I O 0m. Note that CY has a value of 11.8dB dure to the large spreud ofPL
values relativeto the straight-linefit.
and receiver, and notany of the physical features
of the propagation environment,itisnatural forsev-
eral measurements to have the same T-R separa-
tion, but to have widelyvarying PL values. Thisis
due to the fact that shadowingmay occur at some
locations and not others, etc. The precisionof a
PL model is thus measuredby the standard devi-
ation 6 ofthe random variable&, with a smaller
value of CY reflecting a more accuratePL predic-
tion model.
Figure 3 demonstrates measured valuesof PL
asafunctionofdistanceinfivecitieswithinGermany
and shows the above model superimposedon the
measured values. Thistype of plot is called a scat-
ter plot, andit is useful for quickly assessing the
tendencies and variationsof path loss throughout
a radio system. To achieve lower values of6,
more site-specific information about the environ-
ment is required or smaller coverage distances
are used when applyingthe PL model above[5].
Mulfipafhdelayspread-Timedispersionvaries
widely in a mobile radio channel dueto the fact
that reflections and scattering occur at seemingly
randomlocations,andtheresultingmultipathchan-
ne1 response appears random, aswell. Because
time dispersion is dependent upon the geometric
relationships between transmitter, receiver, and the
surroundingphysicalenvironment,communications
engineers often are concerned with statistical
models of time dispersion parameters such as the
average rmsor worst-case values. To demon-
strate how rapidly the multipath characteristicsmay
change, consider Fig.4,which illustrates oneof
the worst casesof multipath time dispersion ever
reported in aU.S.cellular radio system. Figure4
indicates that multipath energy18dSdown from the
first arriving signal incurred a100-psexcess prop-
agation delay. This corresponds to anexcess trav-
el distanceof over 30 km! When the receiverwas
moved by just a few meters, the observed multipath
characteristics became much more tame,with an
observed excessdelay of no more than 1 0 ~ s .In typ-
ical urban cellular systems, the worst case excess
delays (at99percent probability levels)for echoes
which are within 10dB of the maximum signal
are less than 25 ~s [6].
A collection of power delay profiles, such asis
showninFig.4,canbedoneinmanyways.Typically,
a channel sounderis used to transmit a wideband
signal whichis received at many locations within
a desired coverage area. This channel soundermay
operate in the timeor frequency domain. The time
domain responsewillappear asin Fig.4,and a swept
frequency response will reveal that the channelis
frequency-selective, causing some portionsof the
spectrumtohave muchgreater signallevelsthan oth-
erswithinthemeasurementband.Ofcourse,thetwo
measurement techniques are equivalent and can
be related by the Fourier transform.
Power delay profiles may be averagedover time
oroverspaceasthereceivermovesabout.Theaver-
aging interval is key to determining the applica-
bility of the time dispersion statistics, derived
from power delay profiles[6-8].
Some important time statisticsmay be derived
from power delay profiles, and can be used to
quantify time dispersionin mobile channels. They
are: maximum excess delay atX dB down from
maximum (MED), mean excess delay(T), and
rms delay spreada,. The MED is simplycomputed
by inspecting a power delay profile and noting
the valueof excess time delay at which the profile
monotonically dips below the X dB level. For
example, in Fig. 4, the MED at 10dB down would
be lops. It should be noted that these parame-
ters are alsoof interest when they are derived
from instantaneous impulse responses.
The mean excess delay (7)is the first central
moment of the power delay profile and indicates the
averageexcessdelayoffered bythe channel. Therms
measure of the spreadof power about the value
of 7,aT,is the most commonly used parameter to
describe multipath channels. Chuang[9] and
more recent simulation studies have confirmed
that a good rule-of-thumbis as follows: if a digital
signal has a symbol duration whichis more than
ten times the rms delay spreada,,then an equal-
izer is not required for bit error rates better than
For such low values of spread, the shapeof
thedelayprofileisnot important.On the other hand,
if values of approach or exceed 1/10 the dura-
tionofa symbol,irreducibleerrorsdue the frequency
selectivity of the channelwill occur, and the
shape of the delay profilewill be a factorin deter-
mining performance.
Outdoor Propagation
Frequencyreuseisobtainedusingacellularstruc-
tureofcoveragezonesfrombasestations.1ntext-
books, one usually sees a honeycomb structure of
hexagonallyshapedcells. Unfortunately, arealenvi-
ronment changes the propagation conditions such
that a cell, defined as the areain which pathlossis
at or below a givenvalue, has amuch more irregular
shape. Furthermore, at isolatedplaces the cellswill
hdvepocketsinsideeachother,perhapsduetoheight
44 IEEE CommunicationsMagazine January 1995

variations from natural terrain or from man-made
structures. Often interference condition prevent
the optimal use of the cellular system because
reality does not correspond to the simple mathe-
matical models. Increasing the power to cover dead
spotswouldcreateotherproblems.Thenetworkplan-
ner hassome possibilities though,thl ouyhplacingof
the base station antenna and tosonic extent choos-
ingitsradiationpattcrn,toshapethec~ilextent.Tilt-
ingthe antennain avertical plane isanormal mcasure
and by choosing the height above o r helow
rooftops, maybe even down tu street level, v m e -
thing can be done to control thehire ot the cell.
Cells are typically classified roughly according
tosizeasmacrocells andmicrocells. TableIpresents
some distinctive cell features.
Macrocells
The average path loss ISwhat remains after aver-
aging the pathloss over the fast fading due to
multipath.The macrocellswere thebasis for thc first
generationsystcmsmeant for mobileusers,and they
generallyhavebase stationsathighpointslikebroad-
casting systems, witha coverage of several kilo-
meters. Thekey problem is t o make some order
of the seemingly noise-like path lossexistingover an
undulatingterrainwithavarietyoflandcover,mked
land and sea, etc. Thefast Rayleigh fading corre-
spondstoatruenoisedistribution,buttheslowshad-
owingtypevariationisremarkablywelldescribedby
a log-normal distribution, i,e., the average path
loss in dB has a normal distribution.A good rea-
son for this could be that the average pathloss is
the result of forward scattering overa great many
obstacles, each contributinga random multiplica-
tive factor. When converted to dB this givesa
sum of random numbers,which leads toa normal
distribution in the centrallimit.
In the carlydays, planning wasmainly based
on empirical formulas with an experimental back-
ground. Okumurd[101did thefirst comprehensive
measurements in 1968for Japanese environments,
and the derived curves have later been trans-
formed into parametric formulasby Hata [4].It
was noted that a good model for the pathloss (or
the field strength)was a simplepower law where
theexponentncouldvaryasafunctionoffrequency,
antenna heights etc. Expressed simply.on a log-log-
scale, thisis a linear dependency(Eq.1). The rela-
tionship is strictly empirical, butit has proven ex-
tremely robust overthe years notonly inJapanese
surroundings, butinother surroundings aswell.The
reason for this general validity, also for indoor
environments, is not completely clear;Walfisch and
Bertoni[ll]getgoodagreementwithdiffractionover
a multiplicity of absorbing screens.There aresimple
physical situations where the modelis exact. One
is ofcourse freespace corresponding ton =2, anoth-
er is asymptotic propagation overa flat, finitely
conducting surface(n=4),and finally, there is for-
ward scattering over a numberof absorbing screens
of equal height and distance(n = 4). The real world
in outdoor macrocells leads ton’s between2 and
4,withvaluescloserto4inurbansituations. Weshall
meet the same modelin the indoor case aswell.
The modelshave been refined with a number
of correction factors dependingon the land
cover, different degreesofbuilt-up areas,forestsand
rural areas, andin some cases supplementedwith
one or two knife-edge diffractions. Neither Geo-
~~
-85
-90
-95
-1oa
-105
-1la
-115
csf47
I I I , , /
Displa threshold = -111.5 dBm per 40 ns
RMS &lay spread = 22.85 ps
Worst caw San Franciscomeasurement
Mansell St.h
Excess delay time, Cs)
Figure 4. One ufthe wor.sr cases of multipath time disperslon observedin a
2
cellular rudlo s).srern. Meawement mademSan Francisco, California.
.Table 1. Typica/parameters for macrocells and
microcells.
metrical Theory of Diffraction (GTD) nor Uniform
Theory of Diffraction (UTD) arereally successful
in macrocellssince we are dealingwith forwardmul-
tiplescatteringin the transitionzoneswhere nogood
theories exist. The resulting prediction power
normally leads toa standard deviation of errors
between 6 and 10dB,which is not really impressive.
Even the availability of modern computers and the
use of topographical data bases havc not resulted
in a real break-through in better predictions.
Thetime domainmayalsobe modeledusinggeo-
graphical data bases, andwme success has been
achieved in mountainous regions and large cities
[12].Ingeneral,theraytracingroutineslackthecon-
trhutions from random scatterers.
Microcells
Microce//u/arSystems -Microcells are attracting
much attention simply because they can accom-
modate more subscribers per unitservice area than
macrocells. Also, they permit access by low-power
portables. Propagation in microcells differssig-
nificantly from that in macrocells As suggested in
Table 1,smaller cell coverage with lowerantenna
height andlower transmissionpower resultsinmilder
propagationcharacteristicswhencomparedtomacro-
cellular systems. Thesmaller multipathdelay spread
andshallow fading implythefeasibilityofbroadband
signal transmission without excessive counter-
measure techniques against multipath fading.
Microcells are most frequently plannedin an
IEEE CommunicationsMagazine * January 1995 45

0I A
o"0awn
log (distance)
(C)
W Figure 5. Ideal urban street layour:a )typicalpath loss cuwes along an LOS
street b)and along a NLOS vtreetafter turning a corner.
urban area where theheaviest tclctraffic is expect-
ed. If a transmitter antennaheight is lower than
the surrounding building helght. then most ot the
signal power propagates along the street. The
coverageareaextendsalongthestreet. thusthename
"street microcell" is often used. In suburban
areas, microcells might be usedtorealize alast quar-
ter mile connection to a wired network. In this
case, cell coverage might be circular hut strongly
affected by the buildings and obstacles.
StreetMicrocells-Anumberofstudieshavebeen
done regarding street microccll propagation where
mostofthesignalpowerpropagatesalongthestr~et.
Let us assume an ideal rectangular street grid
with building blocks of relatively uniform height
as portrayed in Fig. 5a. In this case. typical path
loss curves along the line-of-sight (LOS) path are
shown to be characterized by two slopesand a single
breakpoint. The path loss exponent is around two,
as in free space propagation, from the transmitter
up to the breakpoint.Beyond the breakpoint theloss
exponent isaround four, implyingasteeper decrease
of the signal strengthas illustrated in Fig. 5b.
This breakpointisapproximatclygiven byZrdz,Jt,,ih
where hb is the base antenna height,h, is the
mobile antenna height. However,if the receiver turns
the corner from theLOS street intoa non-line-
of-sight(NLOS) street, thenthe receiver experiences
adramaticdecreaseofthesignalstrengthbyaround
ZOdBasshowninFig.Sc,thoughthisvariesdcpend-
ing on the streetwidth and the distance between
the transmitter and the corner[131.
On a NLOS street, the signal envelope tends
to follow a Rayleigh distribution. On a LOS path.
the signal envelope follows not a Rayleigh but a
Nakagami-Rice distribution, Multipath delay
spread is usually small due to the existenceof a
dominating, direct signal component.
Modeling and Prediction- Various propaga-
tion models for the street microcells based ona
rwoptic theory have been proposed. Prediction
ofmicrocell coveragebasedon theraymodel Israther
accurate comparedtothe case of macrocells. Afour-
ray model consisting of direct ray, ground-reflected
ray. and the two rays reflected by the building
walls along the street is often as+umed..4 six-ray
model that takes doubly reflected rays by the
building wall$ or larger-number-ol-ray models
are reported t o give more accurate prediction at
the cost of Increasedcomputation time [14]. Corner
diffractionis another problem attracting much
attention. Various modelshave been inbestigated
StZlrtingfrom a simple knife-edge or wedge diffrac-
tlon t() GTD or UTD with a multiplicity ut rays.
An analytical path loss model describing mlcro-
cellular propagation including reflectedrays and
corner diffractionhas been prehented in [151.
Due to the irregularityofthebulldingstructures,
excessive overlap or non-overlapping of the micro-
cell+are sometimes found. Thus, measurement-
based methods are also adopted to enhance the
accuracy of the prediction [16]. Coverage area is
strongly affected by the location of the transmit-
ter antenna. Whenit is installed in the intenec-
tion of the streetsin Manhattan. for instance, the
waves propagate along the streetsin four direc-
tions with subsequent corner diffractionsat each
encounteredcorner.Thisgeneratesacoveragearea
similar to the shapeof a diamond. Whena trans-
mitter is located ona street between intersec-
tions, this makes the cell more elongatedin the
direction along that street [17j.
Indoor Propagation
1 . . . .ndoor radio communication systems are becom-
mg mcreasmgly Important for extendingvoiceand
data communication services within the work-
place. Recent trials by major telephone compa-
nies show a huge demand and customeracceptance
of indoor wireless communications. Some wire-
IesssystemsinterconnectwiththePSTNorIocalPBX
to provide a seamless extensionof the convention-
dlofficetelephonesystem,whileothersystemsoffer
services separate from the PSTN. Indoorsystems
can be broken down into three main classes:
cordless telephone systems; in-building cellular
systems: and local area networks (LANs). Eachof
thesetypesofemergingcommunicationsystemsmust
he designedwith the indoor channelin mind.
Forindoor communicationsystems,many design
issues such as the distance between servers,
expectedportable batterylife,customerperformance
expectations, and the appropriate radiolink bud-
get aredirectly linkedto the propagation environ-
ment. The amountof RF interference that canhe
expectedfromcochannelusersisanequallyimportant
parameter, which is a direct functlon of the prop-
agation characteristics inside buildings.
Propagation inside Single-story
Buildings
Agreat manypropagationmeasurementshave been
conducted by telecommunication companies,
research laboratories, anduniversities in order to
determine reasonabledesign guidelines and prop-
agation parameters for indoor systems. These
measurements show that the particular typeof
46 IEEE Communications Magazine * Januar?. IYY5

Retail stores I914
1 Office, hard partition I1500 13.0 I 7.0 1
I Office, soft partition I1900 I 2.6 I 14.1 1
1 Metalworking 1 1300
HTable 2.Path loss evponent and standard dum-
lion nleuured in diffrrenr huildings.
building has a direct impact on thc obscrvcd
propagation characterictics. This has motivated more
recent research that predicts indoor propagation
using either statistical or deterministic models.
In order to differentiate propagation phenom-
ena, researchers often classify buildingsby the
followingcategorics: rcsidential homesinsuburban
area5, residential homesin urban areas, tradition-
a l older office buildings with fixed walls (hard
partitioni), open plan buildings with movable
wall panels (soft partitions), factory buildings,
grocery stores, retail stores, and sports arenas. Hard
partitions describe obstructions within the build-
ing which cannot be easily moved such as existing
walls and aislcs. Soft partitions describe movable
obstructions such asoffice furniture panels,which
have a height less than the cciling height. Inside a
building, propagation geometrymay be classifiedas
lineofsight (L0S)where the transmitter andreceiv-
er are visible to one another or obstructed (OBS),
where objects in the channelblock a visible prop-
agationpath. Often,physicalsimilaritiesexistbetween
diffcrcnt typcs of buildings. For example, factory
buildings and grocery stores both contain a large
amount of metal inventory and oftcn havc fcw
hard partitic~nswithinthe building. Traditionaloffice
buildings,withmanywallsmadeofplasterandmetal
lathe,often hal.esimilarpropagationcharacteristics
to large residential homes or retails stores that
contain many partitions.
Asdiscussedinthesectiononpropagationphysics,
there arca number of important propagationevents
that must he measured or modclcd before areli-
ahlzwirelesssystemean bedesigned. Measurements
by many researchershave more or lescprovided sim-
ilar results, w,hich are summari& below.
Temporal Fadingfor Fixed and Moving Ter-
minals - For fixed terminals within buildings,
measurements have shown that ambicnt motion
by people throughout the building causes Ricean
fading.with theratio of specular signal powertomul-
tipath signal having a valueof about 10dB. This
results in a typical variation of less than 15 dB for
99.9percent of the time.
Ingeneral. a portablereceiver moving inabuild-
ing experienccs Raylcigh fading for OBS propa-
gation paths and Ricean fading forLOS paths,
regardless of the type of building.The RiceanK
factor can vary from2 to 1 0 dB for portable Ler-
minals depending on the structureof the building
and the number ofmultipathcomponents thatarrive
at thereceiver.
Multipath DelaySpread - Buildings that have
fewer metal and hard partitionstypically have small
rms delay spreads, on the ordero f 30 to 60 ns.
Such buildings cansupport dald ratesin excessof
several Mb/s without the needfor equalization.
However, larger buildings with a great dealof
metal and openaisles can have rms delay spreads
as largc as300 ns. Such buildings are limited to
data ratesof a few hundred kilobits per second
without equalization.
Path Loss -Path loss is a measure of the aver-
age RF attenuation inside a building, andit is
measured by averaging the received signal over
several wavelengths at the receiver. Equation(1)
describes the pathloss situation also for indoor
situations whcre the standard deviation is easily
derived from the scatter plotof measured data.A
smaller value of cs implies that the pathloss
model provides a better prediction for actualloss
within a building. Equation (I) is well suited for
system analysisor Monte Carlo simulationwhen the
impact of a large numberof users within a building
must be considcrcd. Typical valuesof path-loss
cxponents and standard deviations for different
classes of buildings are given in the literature.
Table 2 provides a range of typical values that
have been measured in the past.
Molkdar's survey paper on the subject of indoor
propagationcontainsasummaryofnumerousprop-
agationmeasurementsandmcdelsthat havebeendevel-
oped [18].The early worksby Devasirvatham [19]
and by Salch and Valenzuela [20]have provided a
foundation for propagationwork inindoor channels.
Propagation Between Floors
Predicting radiocoverage between floors of build-
ings has proven to bedifficult, but measurements
have shown that there are general rules-of-thumb
that apply. Quantifying this propagation condi-
tion isimportant for in-buildingwirelesssystemsfor
multifloored buildings that needto share fre-
queneieswithinthebuilding.1nordertoavoidcochan-
ne1 interference, frequencies must be reused on
different floors. The typeof building material
used between floors has been shown to impact
the RFattenuation betweenfloors. Poured concrete
overmetallathingisapopularmodernconstruction
techniquc which provides less RF attenuation
than does solid steel plankswhich were used to
separate floors in older buildings. The aspect
ratio of the building sides also makes a difference.
Buildings that have a square footprinthave higher
attenuation levelsbetweenfloors than buildings that
are rectangular. Also, windows with mctallic tint
impedc RF transmission, thus causing greater
-Indoor radio
communica-
tion systems
are becoming
increasingly
important
for extending
voice and
data com-
munication
services
within the
workplace.
IEEE CommunicationsMagazine * January I995 47

-During the
past few
years, indoor
propagation
prediction
techniques
based on ray
tracinghave
been used to
reconstruct
the many
possible
reflections
porn wall
surfaces
within a
building.
.~ ~ ~
60 60 70 80 90
Tran
[Path loss error]
59 meters
4
90
80
59 meters
I
1
Figure 6. Error contour whch compares measuredand predicted path loss using a wtnple sire specificpur-
1
lition model described itt(51. Note that error is less fhun3 dl3 over 80percenr of the coverage area.
attenuation between floors of a building.
Measurements have shownthat the lossbetween
floors does not increase linearly in dB with
increasing separation distance. Rather, the great-
esttloorattenuationfactor(F4F)indBoccurswhen
the transmitter and receiver are separated by a
single floor. The overall pathloss increases at a
smaller rate as the number of floors increase. This
phenomenon is thought to be caused by diffrac-
tion o f radio energy along the sidesof a huild-
ing, as well a scattercd energy from neighboring
buildings, which can be reccived on different
floors of the samebuilding [Zl]. Typical values of
attenuation between floors is15 dB for onefloor
of separation and an additional6 to 10dB per
tloor of separation up tofour floors of separation.
For five or morefloors of separation, pathloss will
increase by only a few dB foreach additional floor.
Computer-aided Design for In-building
Propagation Prediction
Statisticalpropagationmodelsdonotexploltknowl-
edge of the physical surroundings to the degree
that building drawings can provide. State-of-the
art propagation prediction uses computer-aided
designdrawingsof theindoor environment asameans
of representing thephysical locations of base sta-
tions, receivers, and partitions.For economic rea-
sons,transceiversmustbe placed strategicallysothat
only desired portionsof the building are provided
with radio coverage. CAD tools enable a system
designer to interactively model the performance
of indoor systems.Underthis interactive framework,
propagation models can be modifiedto provide a
wide range of scenarios and base stations can be
added, repositioned, or deletedinteractively. Fur-
thermore, communication parameters such as
received carrler-to-noise thresholds canbe adjust-
ed by the user as needed.
Computer-aided design touls.such asAutoCAD.
can display both single-floor and multifloor build-
ings. Building partitions can be identified in
AutoCAD and assigned an attribute,so that
appropriate propagation models may bc applied.
Figure h shows an example of the modules pro-
videdinatvpicalsitemodelingtool.Workin[22,23]
demonstrated that asimple and surprisingly accu-
rate way tc) predict RF path loss inside a building
is to count partitions between a line drawn
between transmitterandrecewer. Ernpiricallyderived
attenuation factors (AFs)may lhen he applied to
each partitionwhich intersects theline. Through an
interactive proccss. parameters suchas FAFand AF
can be tunedfor a particularclass ofbuildings. Mea-
surements madewithin the environment can then
be added to the CAD modeling toolas a means
of updating the modelsin order t o provide more
accurate mcasurernents at future sites.
During the pastfcw years, indoor propagation
prediction techniques based on ray tracing have
beenusedtoreconstructthemanypossiblcreflections
from wall surfaceswlthin a building. This approach
can easily be accommodatedin computer aided
design tools and shows promise for reducingthe
standard deviationo to within 4 dB over a50 or
bo dB dynamic range within buildings[23-251.
RF Penetration into Buildings
T.he signal strength received insideof a build-
mg due to an external transmitteris impor-
tant fur wireless systems that share frequencies
48 IEEE Communications Magarme January 1995

withneighboringbuildingsorwithoutdoorsystems.
LikeKFpenetration measurementsbetweenfloors,
it is difficult to determineexact models for pene-
tration asonly a limited number of experiments
havc bccn published, and they are sometimesdif-
ficult to comparc. However, some generalizations
can be made from the literature. In measure-
ments reportedto date,signalstrengthreceived inside
a building increases with height. At thc lower
floorsofabuilding,theurbanclutterinducesgreater
attenuationandreducesthelevelofpenetratic~n.At
higher floors, a LOS path may exist. thus causing
a stronger incident signal at the exterior wallof
the building.
RF penetration has bccn foundto be afunction
of Crequency as well as height within thc building.
Most measurements haveconsideredoutdoortrans-
mitters with antenna heights farless than the
maximum height of the building under test. Mea-
surcmcntsinLiverpool[26]showedthatpenetration
loss decreascs with increasing frequency. Specifi-
cally, penetration attcnuation valuesof 16.4, 11.6,
and 7.6 dB were measured on the ground floorof
a buildiug at frequenciesof 441 MHz, 896.5
MHz, and 1400 MHz, respectively. Measure-
ments in [27] showed penetrationloss of 14.2,
13.3.and 12.8 dB for 900 MHz, 1800 MHz, and
2300 MHz, respectively. Measurements madein
front of windows indicated 6dB less penetration
loss on average than did measurements made in
parts of the building without windows.
Walker [ZX] measured radio signals into 14
different buildings in Chicago from seven exter-
nal cellular transmitters. Results showed that
building penetration loss decreased at a rateof
1.9 dB per floor from the ground level upto the
15th floor and then began increasing above the
15th floor. The increase in penetrationloss at the
higher floors was attributed to shadowing effects
ol' adjacent huildings. Similarly, [26] reported
penetration loss decreased at a rateof 2 dB per
floor from the ground levelup to the 9th floor
and then increasedabovetheYthfloor. Similarresults
were also reported in [29].
Measurements have shown that the percent-
age of windows, when compared with the build-
ing face surface area, impacts the levelof RI;
penetration loss.asdoes the presenceof tinted metal
in the windows. Metallic tints can provide from3
to 30 dBof RF attenuation in a single paneof
glass.Theang1eofincidenceofthetransmittedwave
upon theface of thebuilding alsohasa strong impact
on the penetrationloss as shown in [30].
Conclusion
Despitetheenonnouseffortsandprogrcsstodate,
much work remains in the understanding and
characterization of wireless communications
channels.Atrend in3Dnumericalmodeling isseen,
from which time delay statistics as well as cover-
age and interfercnccmight be inferred.This isespe-
cially important with the extension to ever-higher
datarates. Angle-of-arrival statisticsindifferentenvi-
ronmentswill also need to be better modeledfor use
with adaptiveantennas. In general, we mayconclude
that to achieve ubiquitousPCS new and novel
ways of classifying wireless environn~entswill be
nccdcd that are both widely encompassing and
reasonably compact.
References
[ l ] D Cox. R Murray, and A Norrls, "800 MHz Atlenuatlon Measured
in and Around Suburban Houses," AT&T BellLaboratories Tech. 1..
"01. 673. no. 6, July-Aug, 1984. pp 921-954
I21W. C.Y. Lee, Mobile Communfcatlonr Deslgn Fundamentals.
(Wiley. Second Edltion, 1993).
I31T 5. Rappaport, "Indoor Radio Comrnunicatlons for Factories of
the Future," IEEE Commun. Mag., May 1989. pp. 1524
L41M Hata."Emplrlcalformulaeforpropagat~onloss~nlandmob~lerad~o
services," IEEE Trans.Veh.Tech., L7-29, 1980, pp 317-325.
I51 5.Y. Seldel and T. 5. Rappaport, "914 MHz Path Loss Predlctlon
Models for Indoor Wireless Communtrailons ~nMultlfloored
161T 5 Rappapon. et al ,"900 MHz MultlpathPropagation Measure
Bulldmgs." IEEE Trans. Ant. Prop, vol. 40. no. 2. Feb 1992.
"01.39, no. 2, May 1990. pp 132-139.
mentsforU.S.DigltalCellularRad~otelephone."iEEETrans.Veh.Tech ,
I71D. C. Cox and R. P. Leck, "Distnbutjonsof MultipaihDelay Spreadand
Average Excess Delay for 910 MHz Urban MobileRadlo Paths."
1811. Bach Andersen. "Distribution of phase derivatlvesIn moblle com-
IEEE Trans Ant Prop, "01. AP 23. March 1975, pp. 206-213
191J. C-I Chuang. "The Effects of Time Delay Spread on Portable Radlo
munlcatlons."lEEProc,RH. "01. 137, no 4,Aug1990, pp 197.201
Communicatlons Channelswtth Digital Modulation:' IEEEJSAC.
YOI 5,no 5, June 1987. pp 879 - 889
[lo] Y.Okumura et.a/., "Field strength and its variablllty in VHF and UHF
landmobllerad10m c e , " Rw of the ECL MI 16. 1968, pp. 825-873.
I l l ] 1. Walflsch and H. Bertani, "ATheoret~calModel of UHF Propaga-
tlon InUrban Envlronments;' IEEE Trans. Antennas and Propaga-
1121T Kurner. D. J Cichon. and W. Wiesbeck. "Concepts and Results
tion. voI 36.no 12. Dec 1988. pp 1788.1796
JSAC. vol. 11, no. 7,Sept. 1993, pp 1002 -1012
far 30 Dlgntal Terraln-Based Wave Propagatlon Models." IEEE
I131Y. Nagata. et. ai.. "Measurement and modehng of 2 GHr-band
out-of-saghi radlo propagainoncharactemtics under mlcrocellular
I14lA J Rustako,Jr et.al."Radiopropagationatmtcrowavefrequenc~es
environments," Proc. PIMRC'91. Sept. 1991. pp. 341-346.
forl~ne-of-sightmicrocellularmobileandpersonalcommunicat~ons,"
IEEE Trans Veh. Techn, "01. 40, no. 1,Feb 1991, pp 203-210
I1511 Wiart."M~cro-cellularModel~ngwhenBaseStationAniennairbelow
Roof Tops," Proc. IEEE Vehicular Technology Conference, June
I161 W.T Webb, "Sizing up themicrocellfor rnoblle rad10 communica-
1994, Stockholm, pp. 200-204.
I171A. J. Goldsmith and L. 1. Greenstem. "A Measurement-Based
tlons. "Electronrcs and Commun. Eng J., pp. 133-140.June 1993
vol 1 l . m 7,Sept1993,pp1013-1023
ModelforPredlctlngCoverageAreasofUrbanMicrocells."lEEEJSAC.
1181D. Molkdar. "Review on Radlo Propagatlon Into and Within B u W
ingr,"IEFProc voI 138. no 1.Feb 1991
1191D. Oevasirvatham, '71me Delay Spread and Slgnal Level Measure-
mentsof850MH~rad~owaver1nbu~ld~ngenv~ranments."lEEFTram
Ant. Prop.. YOI. AP-34, no. 11, NO". 1986. pp. 1300 1308.
1201A.Saleh and R Valenzuela, "A statlstlcal model for Indoor multl-
[21]W.Honcharenko,H.L.BeRon~,andJ.Da~ling,"MechanlsmsGoverning
path propagabm,"IEEEJC4C. MI. SAC 5 , no 2, Feb 1987, pp 138.146
Propagatton Between Floors ~nBulld!ngs." IEEE Tram Antennas
i221A. Motley and 1. Keenan, "Rad10 Coverage In Bulldlngs." Brrttsh
andPropagation, "01. 41, no. 6. June 1993, pp. 787-790.
Telecom Tech.1.."01.8, no. 1,Jan 1990, pp 19-24.
12315. Y.Seidel and T. S Rappapon. "A Ray Tracing Tcchnlque io Pre-
dlct Path tors and Delay Spread mslde Buildings." IEEE GLOBE-
1241 R. Valenzuela. "A Ray Traclng Approach to Predictlng Indoor
COM,Dec. 1992, pp. 1825-1829.
1251C. M. P. Ho, et. ai.. "Antenna Effectson Indoor Obstructed wire^
W~relersTransmlsrlan." IEEE Veh. Tech. Conf. 1993. pp. 214-218.
l e s ~Channels and a Determmlsttc Image-Bared Wldeband Propa-
I Wirelesslnfo. Networks, "01. 1, no. 1, 1994. pp. 61-76.
gationMod~forIn-BuildingPersonalCommunicat~onSystems,"lnf'l
1261A. M. D. Turkmanl, 1. D. Parson, D. G.Lewis, "Radio Propagatton
on Land Moblle Radlo,Dec 1987.
Into Buildings at 441, 900. and 1400 MHz." Proc 4th Intl. Conf.
1271A M 0. Turkmani.A. F. Toledo. "Propagation Into and Wlthln build^
fngr at 900, 1800, and 2300 MHz." 1992 IEEEVehTech C o d
1281E H. Walker. "Penetration of Radio Stgnals Into Bulldmgs ~nCellu-
lar RadloEnvlronments:' BellSys. Tech. 1.. vol. 62.no 9. 1983.
[29]J.M.Durante,"BuildingPenetrationLossai900MHr."lEEEVeh.Tech.
Conf., 1993.
I301J Hortklshl. et a/, "1 2 GHz Band Wave Propagatlon Measure-
mentsinConcreteBuildingsforlndoorRadioCommunications:'lEEE
TransVehTech.. L7-35. no. 4, 1986.
Biographies
IBRGEN BACH ANOERSEN IF '921 il a professor at Aalborg UnlverSlty.
Aalborg. Denmark and head of Center for Personkommunikatlan He
IS actlve in International research concerning mobile communlcatlons
bemg chalrman of COST 231 Warklng Group on UHF prapagatlon He
IS also actwe ~nthe area of bloelectromagnetlrs and is presently "Ice-
president for URSI, the lnternatlonal Sclentlflc Radio Union
THEODORES RAPPAPORT ISM '91II S an asroclate professor of electrical
englneermg and founder of the Moblle& Portable Rad10 Rerearch
Group (MPRG) at Virginia Tech, Blacksburg, Virginla.
Susuvu YOSHIDA15 a professor of electrical engineering at Kyoto Unl-
versity. Kyoto, Japan, leading a research group on wlreless personal
communtcationr He IS Involved ~n multlpath propagatmn modellng
and predlrtlon, rector-antenna dsverslty, anti-multlpath modulation
schemes, and u-servlce multlpath delay spread monitoring, eic
-The
percentage of
windows,
compared
with the
buildingface
sueace area,
impacts the
level of RF
penetration
loss, as does
thepresence
oftinted
metal in the
windows.
IEEF Cummunications Magazine January 1995 49

Propagation measurements and models for wireless channels

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Propagation measurements and models for wireless channels (20)

More from Nguyen Minh Thu (20)

Recently uploaded (20)

Propagation measurements and models for wireless channels