antennas prop

7/31/2019 Antennas Prop

1/70

Antenn as and Propagat ion Component s

September 2004


2/70ii

Notice

The inform at ion cont ained in t his docum ent is subject to cha nge without notice.

Agilent Techn ologies ma kes n o warr an ty of an y kind with r egard t o th is mat erial,

including, but n ot limited t o, the implied war ra nt ies of mer cha nt ability an d fitn ess

for a pa rt icula r pur pose. Agilent Techn ologies sh a ll not be liable for er rors cont ain ed

her ein or for incidenta l or consequen tial da ma ges in conn ection with th e fur nishing,

perform an ce, or use of this m at erial.

Warranty

A copy of the specific warr an ty t erm s t ha t apply to th is softwar e product is available

upon r equest from your Agilent Techn ologies rep resent a tive.

Restricted Rights LegendUse, duplica t ion or disclosure by the U. S. Government is subject to res t r ict ions as se

fort h in subpar agr aph (c) (1) (ii) of th e Right s in Techn ical Da ta an d Comput er

Software clause at DFARS 252.227-7013 for DoD agencies, and subparagraphs (c) (1

an d (c) (2) of the Commercial Comput er Softwa re Rest ricted Right s clau se a t FAR

52.227-19 for other agencies.

Agilent Technologies

395 Page Mill Road

Palo Alto, CA 94304 U.S.A.

Copyright 1998-2004, Agilent Technologies. All Right s Reserved.

Acknowledgments

Mentor Gra phics is a t ra dema rk of Mentor Gra phics Corporat ion in th e U.S. and

oth er coun tr ies.

Microsoft , Windows, MS Windows, Windows NT, an d MS-DOS ar e U.S.

registered trademarks of Microsoft Corporation.

Pentium is a U.S. register ed t ra dema rk of Intel Corporat ion.

PostScript an d Acrobat ar e tr adem ar ks of Adobe System s Incorporat ed.

UNIX is a register ed tra dema rk of th e Open Group.

J ava is a U.S. tr adem ar k of Sun Microsystem s, Inc.


3/70ii

Contents1 Antennas and Propagation Components

Introduction............................................................................................................... 1-1

Multipath and Fading ................................................................................................ 1-5

Pathloss.................................................................................................................... 1-9

References ............................................................................................................... 1-10

AntArray.................................................................................................................... 1-11

AntBase.................................................................................................................... 1-14

AntMobile.................................................................................................................. 1-16

Fader ........................................................................................................................ 1-18

PropFlatEarth ........................................................................................................... 1-21

PropGSM.................................................................................................................. 1-24

PropNADCcdma....................................................................................................... 1-31

PropNADCtdma........................................................................................................ 1-35PropWCDMA............................................................................................................ 1-39

UserDefChannel....................................................................................................... 1-43

UserDefVectorChannel............................................................................................. 1-46

3GPPFDD_Channel ................................................................................................. 1-48

TDSCDMA_FwdChannel.......................................................................................... 1-51

TDSCDMA_RevChannel .......................................................................................... 1-53

WCDMAVectorChannel ............................................................................................ 1-55

WLAN_ChannelModel.............................................................................................. 1-57

Index


4/70iv


5/701-1

Chapter 1: Antennas and PropagationComponents

IntroductionAnt enn a a nd pr opagat ion m odels simu late t he effects of ra dio cha nn el on t he

tr an smitt ed signa l. These effects include signal fading an d pat hloss. Both an ten na

and propaga t ion channel models a re TSDF componen t s with input and ou tpu t t imed

signals.

Antenna models a re iden t ified by their coordina te and ga in . For mobile an tennas, the

velocit y vect or is a lso in clu ded in t h e pa r a met er s. Th e in pu t of a n ten n as a r e m u lt i t o

provide flexibility of adding t he cont ribu tion from var ious noise or fadin g cha nn els.The pr opa gat ion chann el models ar e typica lly ident ified by the t ype of fad ing, th e

specificat ion of power delay pr ofile, and whet her pa th loss is to be activat ed.

The input as well as output impedan ce of an tenn as an d cha nn el models ar e left a s

infin ite a nd zer o, respectively. For t he inclusion of an ten na impeda nce, cosimu lat ion

with Circuit E nvelope is recomm ended (for det ails, refer t o Introduction: Circuit

Cosimula tion Component s). In t his a pplicat ion, S-par am eter block r epresent ing

mea surem ent (file), fun ctions or impeda nce componen t s can be placed on t he circuit

schema tic page an d co-simu lat ed with signa l processing designs. To access a n

example tha t demonst ra tes t his: from th e Main window, choose File > Exam pleProject > An tenn as-Prop > R ad ioChan nel_prj; from the Schem at ic window, choose

File > Open Design, A N T LOAD.dsn .

The sepa ra te specificat ion of an ten na s a nd (propagat ion) cha nn el component s is to

provide a more int uit ive an d flexible use model. Dur ing th e simulat ion, th e effects o

both an tenn a a nd cha nn el models ar e merged and t he combined antenn a a nd

propagat ion chan nel component s ar e replaced with equivalent models in t he

pre-processing phase of the simulation.

Becau se of inter dependency of an ten na s a nd cha nn el models, cert ain rest rictions intopology mu st be cons idered by th e user : th ere should be at least one propagation

chann el between a p air of transm it and receive an tennas. Exa mples of common

an ten na an d propagation cha nn el conn ections a re shown in Figure 1-1 through

Figure 1-4.

Figure 1-1 dep ict s the mos t common use model where a channel (P ropGSM) is p laced

between a base st at ion (Ant Base) and mobile (Ant Mobile) an ten na s. (Note t ha t th e


6/701-2

Antennas and Propagation Components

cha nn el test port can be left un conn ected. This port is supplied only if th e user is

inter ested in th e behavior of th e cha nn el models.)

Figure 1-2 sh ows a topology wher e two cha nn els (Pr opNADCtdm a) ar e placed in

par allel between a base a nd m obile an ten na s. One of th ese cha nn els could model a

line-of-sight scenario (NoMultipa th opt ion) while th e oth er one models a fla t fadin gcha nn el. The combined effect of these t wo cha nn el types ar e simu lated in t his case.

Ant enn as (Ant Mobile a nd Ant Base) require a single carr ier frequency for mu ltiple

input port s. If th e cha nn el out put s ha ve different car rier frequen cies, an RF

convert er (Sum mer RF) can be used as sh own in t his case. (Note tha t a ny nu mber of

par allel cha nn els between two an ten na s ar e allowed.)

Figur e 1-1. GSM Chan nel


7/701-3

Figure 1-2. Two TDMA Cha nn els

Figure 1-3 is a t opology wher e a AWGN noise sour ce is added to th e inpu t of th e Rx

an tenna in addit ion to the fad ing channel (P ropNADCcdma). The user can crea te and

add a ny arbitra ry cha nn el between a pa ir of an tenn as.

Figure 1-4 shows a Tx ant enn a outpu t en ter ing two cha nn els, each conn ected t o an

Rx antenna. This topology is used for s imula t ion of antenna divers ity or in terference


8/701-4


Figur e 1-3. AWGN Noise Sour ce

Figur e 1-4. Tx Ant enna Out put

Figure 1-5 depicts t he t opologies th a t a re n ot a llowed; th ese include:

chan nel model without Tx or Rx ant enna chan nel model without Tx and Rx ant enna

cha nn el model with more th an one distinguishable Rx (or Tx) an tenn a.


9/701-5

Figur e 1-5. Disa llowed Topologies

Multipath and FadingThis section defines term s a nd r elat ions r elevan t to the m ultipath an d fading in th e

propagation model types.

Definitions:

V = veh icle speed, in m /s

Fc = propa gat ion (car rier ) frequ ency, in H z

c = propaga tion (car rier ) frequ ency, in r ad ian /sec

= Doppler frequency, in H zm = ma ximum Doppler frequency, in H zS(t) = tra nsm itted (RF) signal

s(t) = complex envelope of tr ansmit t ed signa l

R(t ) = received (RF) signa l

r(t ) = complex envelope of received s igna l

n= ran dom a mplitude ofn th signa l echo


10/701-6


n = phase r etar dat ion ofn th signa l echon = tim e-delay ofn th signa l echoGt( , ) = directive gain of tr an smitt ing an ten na as a function of elevation a ndazimu th a ngles

Gr( , ) = directive gain of receiving an ten na as a fun ction of elevat ion a ndazimu th a ngles

Rad io waves ar e received not only via dir ect pa th but oft en by scat ter ing off

nu mer ous objects. Delay, att enu at ion a nd car rier ph ase sh ift a re some of th e

alter at ions t he t ra nsm itted signa l experiences. This process can be modeled as a

linea r filter with ra ndomly time-var ying impu lse response.

In a mu ltipat h environm ent, a tra nsmitt ed RF signal

is r eceived in th e form

wher e n is t he n um ber of differen t echoes, each h aving a delay.

The received complex envelope is therefore

The (lowpass) impu lse response of th e discret e cha nn el h( ,t ), is t her eforecha ra cter ized by severa l discret e pa th seach ha ving a specific delay an d

attenuation.

Signa l fad ing occur s du e to destr uctive or cons tr uctive addit ion of a la rge nu mber of

pha sors. Ifh( ,t ) is modeled as a zero mea n Gau ssian process, the en velope |h( , t )|at any t ime is Rayleigh-dis t r ibuted. The t ransform of h(,) with respect to t ime, giveth e spectr um of time var iation S(,), gener a lly referr ed to as delay-Doppler spr eadfunction [1]. The variable r ep resen t s the Doppler fr equency sh ift due to changes inth e electr ical pa th length a s a resu lt of mobile movement.

For t wo vertically polarized tra nsm it an d receive an ten na s an d horizont al

propagat ion of plan e waves [2], the Doppler spectr um is

S t( ) s t( )ejw ct =

R t( ) G t n , n( )Gr n , n( )n t( )s t n( )ej c t n[ ] n[ ]

n

=

r t( ) G t n , n( )Gr n , n( )n t( )s t n( )ej cn n+[ ]

n=


11/701-7

where

is th e ma ximu m Doppler sh ift du e to vehicle speed. When a dir ect pa th exists th e

spectr um is Rician an d is given by

with k1, k2, k3 const an ts r elated t o proport ion of direct a nd scatt ered signal an d th e

direct wave angle of ar rival. Assum ing th e wide sense sta tiona ry u ncorr elated

scat ter ing (WSSUS) [3], the avera ge delay profiles and Doppler spectr a in form ation

is needed for the s imula t ion of radio channel. Delay profiles [4] P( ) ca n be m ea su r ed(or approximated) as

where

is the power a ssociated with each pat h.

Assu ming a u niform distr ibut ion of independent scat ter ers in t he h orizont al plan e,

each with a Doppler sh ift rela tive to the velocity of the mobile, the delay-Dopplerspread function S (, ) a n d t h e im pu lse r espon se of t h e ch a nn el ca n be con st r uct ed . Awide-band,frequ ency select ive, mult ipath fading m odel can th erefore be const ru cted

us ing a ta pped-delay-line filter. The typical ta pped-delay-line filter model for

simula tion is illust ra ted in Figure 1-6.

S ( ) 1

2m 1

m-------

2

--------------------------------------------- m=

0= m>

mV

c----f

c=

S ( )k1

2m 1

m-------

2

--------------------------------------------- k2

k3

m

( ) m

+=

P ( ) n2 n( )n=

n2


12/701-8


Figure 1-6. Tapped-Delay Line Model for a Wide-Ban d Ch an nelTo gener at e a Ra yleigh fad ing profile for ea ch pat h, independent AWGN sources (in

cascade with a filter rep resent ing th e effects of Doppler spr ead) can be used; see

Figure 1-7.

J ak es [5] proposes a more efficient a ltern at ive to Figure 1-7. In J akes model a

nu mber of low-frequen cy oscillators a re u sed to gener at e signa ls th at ar e added

togeth er. The am plitu de an d ph ases of th ese oscillators a re chosen so th at th e pdf of

th e resulta nt phase approximates t o a un iform distr ibut ion. The spectr um of the

resu lting complex fun ction appr oxima tes th e Doppler spectr um .

Figure 1-7. Genera tion of Rayleigh pdf with a Given P SD

Delay

Attenuator

DelayDelay Delay Delays(t)

Gain

Functions

(colored

Gaussian)

S

r(t)

Attenuator Attenuator Attenuator Attenuator

Gaussian #1

Gaussian #2

I

Q

Rayleigh

S ( )

S ( )


13/701-9

Pathloss

This section defines ter ms a nd r elations r elevant to the pat hloss in th e propagat ion

model types.

Definitions:LFS = pat hloss in free-spa ce environm ent , in dB

LRA = path loss in ru ra l area environm ent, in dB

LHT = path loss in h illy ter ra in environm ent , in dB

LTU = pat hloss in typical ur ban environm ent , in dB

LTS = path loss in typical subu rba n en vironmen t, in dB

fc = propagat ion (car rier ) frequ ency, in MH z

c = wavelength associated with propagation (carrier) frequency

D = ma jor an ten na dimensionHBS = base sta tion a nt enna height , in meters

HMS = mobile station an tenn a h eight, in meters

R = dista nce between tr an smit a nd r eceive antenn a, in km

a = correction factor, in dB

Free-space path loss opt ion provides the user with an opt imis t ic model. This opt ion is

given (in dB) by

LFS

= + 20log10

fc

+ 20log10

R + 32.4

Ther e ar e a wide var iety of pa th loss m odelsth e most widely used is Ha ta s. Hat a s

pat hloss m odel [6] is based on a n extensive data base der ived by Okum ur a [7] from

mea sur ement s in an d ar oun d Tokyo. Hat as pat hloss models cover ur ban , rur al, and

subur ban environments a nd include the tra nsm it and r eceive an tenn a h eights.

The typical urban H ata model is given by

LTU = 69.55 + 26.16log10fc 13.82log10HBS a (HMS)+ (44.9 6.55log10HBS)log10R

The corr ection factor for small- to medium -size cities is given by

a = (1.1log10 fc 0.7)H MS (1.56 log10 fc 0.8)

Ha ta s u rba n model is equivalent to Type=TU in pr opagat ion component s.

The typical suburban H ata model is given in ter ms of LTU with a corr ection factor:


14/701-10


LTS = LTU 2[log10(fc/28)]2 5.4

Similar ly, the rural Hata m odel is a corr ected form of LTU as

LRA = LTU 4.78(log10fc)2 + 18.33log10fc 40.94

Ha ta s ru ra l model is equivalent to Type=RA in propagat ion componen ts.

For GSM Hilly Terr ain (HT) environmen t, a djust men ts prescribed in [5] ar e used.

All th e path loss form ulas a re validassu ming th at th e receiving ant enn a is in t he

far field of th e tr an smit a nt enn a. The comm on criterion for a nt enn as wh ose physical

size is in th e order of wavelength is th at pat h length should exceed D2/c.

References

[1] J. D. Par sons, Th e Mobile R ad io Propagation Chann el, Halst ed Pr ess, 1992.

[2]R. H. Clarke, A Statistical Theory of Mobile-Radio Reception, The Bell S ys tem

Techn ical Journ al, Ju ly-Augu st 1968.

[3] Raymond Steele,Mobile R adio Com m unication s, Pent ech Pr ess, 1992.

[4] GSM 05.05 Recomm endat ion, R adio Transm ission and R eception .

[5]W. C. Jakes (Editor),Microwave M obile Com m unica tions, John Wiley & Sons,

1974.

[6] M. Ha ta , Em pirica l Form ula for P ropaga tion Loss in La nd Mobile Radio,

IEEE Trans. VT-29, pp. 317-325, August 1980.

[7] Y. Okum ur a, Field St ren gth an d its Var iability in VHF an d UH F La nd Mobile

Service,R eview of E lect rica l Com m unication Labora tory, Vol 16, pp. 825-873,

Sep-Oct . 1968.


15/70AntArray 1-11

AntArray

Description Antenna Array

Library Antennas & Propagation

Class TSDFAntArray

Derived From antenna

Parameters

Pin Inputs

Name Description Default Sym Unit Type Range

Gain Gain of antenna, in dB 0.0 g real (-, )X X-position coordinate of

antenna

0.0 km real (-, )

Y Y-position coordinate of

antenna

0.0 km real (-, )

Height base station height of

antenna array, measured

from average rooftop

15 km real (0, )

AOA angle of arrival array

representing the multipath

wavefront azimuth angles

received by the array in the

uplink operation mode

75.0 45.0 15.0

-15.0 -45.0 -75.0 real array (-, )

OperationMode operation mode of the

array: UpLink, DownLink

UpLink enum

NumberOfElements number of array elements 6 M int (1, )IntervalOfElements interval between two

antenna elements

0.075 d real (0, )

Pin Name Description Signal Type

1 inm Antenna input signals from channel multiple timed


16/701-12 AntArray


Pin Outputs

Notes/Equations

1. This component simu lates an an tenn a ar ray using gain of an tenn a an d the

an gle of ar rival of each mult ipat h echo specified by th e AOA pa ra met er.

In th e uplink m ode, set by t he Opera tionMode para met er, the model receives L

input timed signa ls from a mu ltipat h cha nn el, processes these signa ls and

out put s M timed signal associated with an M-elemen t a rr ay.

In t h e down lin k m ode, it su m s a ll r eceived in pu t t im ed sign a ls a n d t r a nsm it s M

timed signa ls corr esponding t o the M-element ar ra y.

At ea ch fir ing, one t imed token is consum ed, and one t oken is produced.

2. Figure 1-8 shows the geomet ry of the a r r ay and a received mult ipa th wavefron t

Figure 1-8.

As sh own in Figure 1-8 th e un iform ar ra y is placed along t he X-axis with a

separation d. The a rr ay r esponse vector is described by th e an gle measuredfrom th e broadside direction perpen dicular to the ar ra y in t he a zimu th plane.

The a rr ay response vector for a n a rr ay with M element s is given by


2 outm Output signals associated with array elements multiple timed


17/70AntArray 1-13

where

= 2/ is the wavenum ber

The out put of elemen t k of th e ar ra y can be expressed a s

where Xl is th e lth mu ltipath echo received by the a rr ay.

l( )

1

jd l( )sin( )exp

j2d l( )sin( )exp

...j M 1( )d l( )sin( )exp

=

y k g k l( )Xll 1=

L

=


18/701-14 AntBase


AntBase

Description Base Station Stationary Antenna Model


Class TSDFAntBase


Parameters

Pin Inputs

Pin Outputs

Notes/Equations

1. Base (or fixed) s t at ion an tennas a re linea r ly pola r ized an tennas used in mobile

comm un icat ion service at th e base st at ion of a ra dio relay link . The

specificat ion EIA/TIA-329-B, Minim um Stan da rds for Comm un icat ion

Ant enn as, Part I-Base Sta tion Anten na s describes the st an dar ds for t his class

of antenn as.


Gain Gain of antenna, in dB 0.0 g real (-, )X X-position coordinate 0.0 km real

Y Y-position coordinate 0.0 km real

Height antenna height above X-Y

plane

10 km real (0, )


1 input Antenna input signal multiple timed


2 output Antenna output signal timed


19/70AntBase 1-15

2. This component ra diation h as a domina nt vertical component of th e electr ic

field (E z).

3. The Gain un it is in dB and is defined with reference to an isot rop ic source (dBi)

To comply with comm un icat ion a nt enn a s ta nda rds, a dBd (gain with respect to

a h alf-wave dipole) sh ould be used, which is 2.15 dB over isot ropic.4. To accomm odate for specificat ion of input impeda nce versu s frequency, circuit

subn etworks can be crea ted a nd co-simulat ed with th e appr opriat e circuit

simulator.

5. The ant enna ha s a m ulti-input pin to receive mu ltiple cha nn els when at Rx

mode. All input s to Ant Base mu st ha ve th e sam e car rier frequen cy.

6. For genera l inform at ion, refer t o In tr oduction on pa ge 1-1.


20/701-16 AntMobile


AntMobile

Description Cellular Mobile Antenna


Class TSDFAntMobile


Parameters

Pin Inputs

Pin Outputs


Gain Gain of antenna, in dB 0.0 g real (-, )X X-position coordinate, in

distance units

100.0 km real

Y Y-position coordinate, in

distance units

0.0 km real

Height antenna height above X-Y

plane, in length units

2.0 km real (0, )

SpeedType velocity unit: km/hr,miles/hr

km/hr enum

Vx X component of velocity

vector

0.0 real

Vy Y component of velocity

vector

0.0 real


1 input Antenna input signal multiple timed


2 output Antenna output signal timed


21/70AntMobile 1-17

Notes/Equations

1. Mobile an ten na s ar e moun ted on vehicles an d used in th e lan d-mobile

communications services. The specification EIA/TIA-329-B-1, Minimum

Standar ds for Commun ication Antenna s, Part II-Vehicular Antenna s describe

th e sta ndar ds for t his class of an tenn as.2. This component ra diation h as a domina nt vertical component of th e electr ic

field (E z).

3. The Ga in u nit is in dB and is an isotr opic source (dBi). To comply with

s tandards for communica t ion an tennas, a dBd (ga in with respect to a ha lf-wave

dipole) sh ould be used, wh ich is 2.15 dB over isotr opic.

4. To accomm odate for specificat ion of input impeda nce versu s frequency, circuit

subn etworks can be crea ted a nd co-simulat ed with th e appr opriat e circuit

simulator.

5. The ant enna ha s a m ulti-input pin to receive mu ltiple cha nn els when at Rx

mode. All inpu ts to Ant Mobile mu st ha ve th e sa me car rier frequ ency.

6. The m obile an ten na position chan ges from initial locat ion a long a str aight line

dur ing simu lation. The new coordina tes a re

X (t) = X(t) + Vx tY (t) = Y(t) + Vy t

7. Pr opagation pat hloss is updat ed based on cha nging distance between tra nsmitan d receive antenn as.



22/701-18 Fader


Fader

Description Fading channel model


Class TSDFFader

Derived From channel

Parameters

Pin Inputs

Pin Outputs

Notes/Equations

Name Description Default Type Range

GainArray path gain in terms of

decible

0dB -10dB real array (-, )

DelayArray path delay in terms of ns 0ns 976ns int array [0, )RicianFactor ratio of specular power and

the fading power

0.0 real [0, )

Algorithm the algorithm used:

SoS_Stochastic,

SoS_Deterministic,

Filter_AWGN_Noise

SoS_Stochastic enum

for each element of the array


1 input channel input signal timed


2 output output signal timed

3 outChM fading factor multiple timed


23/70Fader 1-19

1. This model is th e fading cha nn el emu lator. The input signa l is faded by

mu ltiplying t he fading coefficients genera ted using t he selected algorith m. If

Rician Factor is 0.0, th e fad ing pr obability dens ity fun ction is Rayleigh

dist ribu ted ; oth erwise th e fading probability density fun ct ion is Rician

distributed.

2. GainArra y specifies th e gain of each fading pa th in t erm s of decibel, while the

delay of each pa th is specified by DelayArra y in ter ms of na nosecond . The

nu mber of fading pa th s t o be genera ted is equa l to th e size of GainArra y and

DelayArray. The generated multiple path fading coefficients are output from

pin 3 for t est pu rposes. Pin 2 is th e out put of th e signa l passing th rough a

mu ltipath fading cha nn el.

3. The gain of th is model ha s been normalized so tha t t he output power is the

sam e as th e input power r egardless of th e cha nn el configur at ion. RicainFa ctor

is the ra t io of direct s ignal power and fading power. Fading power is d is t r ibutedto th e mu ltiple pat h s igna l according t o the gain sett ing of each pat h. For

exam ple, if th e inpu t power is 1, Rician Factor is 2, and Ga inArra y is 0dB

-2dB, then the direct s ignal power will be 2/(1+2)=2/3 and fading power will be

1/(1+2)=1/3. Fad ing power a lloca ted to pa th 1 will be

an d power for pa th 2 will be

.

4. Thr ee algorith ms can be selected t o genera ted t he fading coefficient.

stocha stic sum-of-sinusoid meth od in which t he n um ber of oscillators a re

selected as 64[1];

J ak es determin istic sum -of-sinusoid method [2];

to pass a n AWGN n oise through a sha ping filter; th is filter is identical with

the one used in CDMA2K_ClassicSpec which is available in CDMA2K design

library.

5. Th is m odel is u sed in con ju n ct ion wit h a n ten n as. E it h er a ba se st a t ion a n ten n a

or a m obile ant enn a is conn ected with t he inpu t a nd outpu t pins. This model

reads velocity informat ion from the antenna to calcula te the Doppler frequency

shift.

13---

0.0 1010 100.0 10 10 2 10+( )

13---

2 1010 100.0 10 10 2 10+( )


24/701-20 Fader


6. If th e input time st ep is too large, int erpolat ion will be perform ed to up-sample

th e signa l so th at th e resulted t ime step will be less th an 1 nsec. Simu lation

time in th e case of a la rge inter polation ra te would increase; in oth er cases

when th e delay for a pat h is lar ger, the signals t o be buffered a nd int erpolat ed

would increase which would lead to increa sed simu lation time.

7. If Filter _AWGN_Noise met hod is selected , its bett er t o ha ve the t ime st ep less

than 1 sec so tha t t he fadin g coefficient s will be more rea sonable.

8. The fadin g coefficient s gener at ed by SoS_Stocha st ic an d F ilter_AWGN_Noise

methods are independent for d ifferent paths. However, caut ion must be given to

proved [3] th at th e Jak es SoS Deter min istic met hod doesn t h ave good

corr elation between different pat hs.


References

[1]Yahong R. Zhen g an d Chen shan Xiao, "Imp roved models for t he genera tion of

multiple un-correlated Rayleigh fading waveforms," IEEE Communications

Let ters, vol. 6, no. 6, pp. 256-258, June 2002.

[2] W. C. Jakes, Microwave Mobile Commu nicat ions: Wiley, 1974. Reprin ted by

IEE E P ress in 1994.

[3]P. Dent, G. E. Bottomley, and T. Croft, "Jakes fading model revisited,"

Elect ronics Let ter, vol. 29, no. 13, pp. 1162-1163, June 1993.


25/70PropFlatEarth 1-21

PropFlatEarth

Description Direct and Reflected Ray Propagation Model


Class TSDFPropFlatEarth


Parameters

Pin Inputs

Pin Outputs

Notes/Equations

1. Pr opFlatE ar th models th e sum of a direct a nd r eflected ra y propagation

cha nn el model based on polar izat ion a nd flat ear th properties.

2. Th is componen t is based on the two-ray LOS model where the received s igna l is

th e sum of cont ribut ions from t he direct an d reflected r ays. By summ ing the

Name Description Default Type

Polarization polarization type: Vertical,

Horizontal

Vertical enum

Permittivity earths average relative

permittivity

15 real

Conductivity earths average

conductivity

0.005 real




2 output channel output signal timed


26/701-22 PropFlatEarth


cont ribu tions from ea ch r ay, the received signa l at t he Rx end for a pa ir of

an tenn as can be expressed as

where Pt is the t ra nsmitt er power, r1 is the direct dista nce from t ra nsmitt er to

receiver, r2 is the d is t ance th rough the reflect ion on the ground, and ( ) is t hecomplex reflection coefficient as a function of incident angle an d complexperm ittivity of th e groun d er:

where =90 and q=1 or (er)-1 for ver tical or horizont al polar ization,respectively.

3. The complex relative permitt ivity is related t o medium s a verage perm ittivity

and conductivity:

er = r(average) j60

Typical values for permit t ivity and conduct ivity of var ious ear ths mediums are

given in Table 1-1.

Figure 1-9 shows the r eceived power as a fun ction of an ten na separ at ion u sing

FlatE ar th cha nn el [1].

Table 1-1. Typical Ear th s Const an ts

Type of SurfaceAverage Average (mho/meter)

Fresh water (lakes and rivers) 81 0.001

Sea water 81 5.0

Good ground 25 0.02

Average ground 15 0.005

Poor ground 4 0.001

Mountains 0.00075

Pr

Pt

4------

2 1r1-----e

jk r1 ( ) 1r2-----e

jk r2+2

=

( ) q er

2sin cos

cos q er

2sin +

---------------------------------------------------=

r


27/70PropFlatEarth 1-23

Figure 1-9. Received Power a s a Fu nction of Ant enn a

Separat ion Using Flatear th Chan nel


References

[1]H. Xia , H. Bert oni, Rad io Pr opa gat ion Cha ra cter istics for LOS Microcellular

an d Per sona l Comm un icat ions,IEEE Trans, APS, October 1993.


28/701-24 PropGSM


PropGSM

Description GSM Propagation Model


Class TSDFPropGSM


Parameters

Pin Inputs


Type GSM type: NoMultipath,

RuralArea1, RuralArea2,

HillyTerrain6Tap1,

HillyTerrain6Tap2,

HillyTerrain12Tap1,

HillyTerrain12Tap2,

UrbanArea6Tap1,

UrbanArea6Tap2,

UrbanArea12Tap1,

UrbanArea12Tap2,

EqualizationTest

NoMultipath enum

Pathloss inclusion of large-scale

pathloss: No, Yes

No enum

Seed number to randomize

channel output

1234567 int

Test test port accessing single

path random gain

functions: Tap1, Tap2,

Tap3, Tap4, Tap5, Tap6,

Tap7, Tap8, Tap9, Tap10,

Tap11, Tap12

Tap1 enum




29/70PropGSM 1-25

Pin Outputs

Notes/Equations

1. PropGSM models a direct ional mult ipa th channel based on GSM specificat ions

The m odel is a mu lti-ta p filter, each t ap h aving a delay an d power. And , each

ta p is modu lat ed by a colored noise sour ce (gain fun ction).

Including th e mu ltipat h fading and pa th loss effects, th e out put of th e cha nn el

can be described as

where

= pathloss at tenu at ion, 1i = rela t ive power of each echo, per specificat ioni = relat ive delay of each echo, plus t he dir ect pa th delay 0

s(t) = complex envelope inpu t s ignalgi = ra ndom gain function associated with each echo

For t he N oMultipat h option, M=1, 1= o

The ga in function can be described as a sum of non-overlapp ing sinu soids of

equal am plitu des an d different frequen cy and ph ases.

The pat hloss at ten ua tion factor =1 when Pat hloss=No. And, note th at whenPat hloss=No, the su m r(t) may add up t o be more th an th e input s(t), implying

th at cha nn el has a gain. For t his reason, when Path loss=No, the output is

norm alized to the linea r su m of echo powers with ea ch option.

For more deta ils, refer t o In tr oduction on pa ge 1-1.

2. The specificat ion GSM 05.05, Eu ropean Digita l Cellula r Telecomm un icat ions

System (Ph ase 1) Radio Tra nsm ission an d Reception, Annex 4 is th e basis for

th e Pr opGSM model. With th e except ion of Type= NoMult ipat h, wh ich is a n

addit ion, GSM system defines eleven different propagat ion profiles descr ibed by



3 test Test port timed

r t( ) is t i( )ejic

gi

t( )i 1=

M

=


30/701-26 PropGSM


th e Type par am eter : two for r ur al; four for h illy; four for u rba n; an d one for

equalizer test. Table 1-7 through Table 1-6 depict t he GSM delay profiles

associated with each Type.

Table 1-2. Hilly Terr ain 6-Tap Types

Tap Number

Relative Time

( ) Average Relative Power (dB)Doppler

SpectrumOption 1 Option 2 Option 1 Option 2

1 0.0 0.0 0.0 0.0 Classical

2 0.1 0.2 -1.5 -2.0 Classical

3 0.3 0.4 -4.5 -4.0 Classical

4 0.5 0.6 -7.5 -7.0 Classical

5 15.0 15.0 -8.0 -6.0 Classical

6 17.2 17.2 -17.7 -12.0 Classical

Table 1-3. Hilly Terr a in 12-Tap Types

Tap Number

Relative Time



1 0.0 0.0 -10.0 -10.0 Classical

2 0.1 0.2 -8.0 -8.0 Classical

3 0.3 0.4 -6.0 -6.0 Classical

4 0.5 0.6 -4.0 -4.0 Classical

5 0.7 0.8 0.0 0.0 Classical

6 1.0 2.0 0.0 0.0 Classical

7 1.3 2.4 -4.0 -4.0 Classical

8 15.0 15.0 -8.0 -8.0 Classical

9 15.2 15.2 -9.0 -9.0 Classical

10 15.7 15.8 -10.0 -10.0 Classical

11 17.2 17.2 -12.0 -12.0 Classical

12 20.0 20.0 -14.0 -14.0 Classical

sec

sec


31/70PropGSM 1-27

Table 1-4. Ur ban Area 6-Tap Types

Tap Number

Relative Time



1 0.0 0.0 -3.0 -3.0 Classical

2 0.2 0.2 0.0 0.0 Classical

3 0.5 0.6 -2.0 -2.0 Classical

4 1.6 1.6 -6.0 -6.0 Classical

5 2.3 2.4 -8.0 -8.0 Classical

6 5.0 5.0 -10.0 -10.0 Classical

Table 1-5. Urban Area 12-Tap Types

Tap Number

Relative Time



1 0.0 0.0 -4.0 -4.0 Classical

2 0.1 0.2 -3.0 -3.0 Classical

3 0.3 0.4 0.0 0.0 Classical

4 0.5 0.6 -2.6 -2.0 Classical

5 0.8 0.8 -3.0 -3.0 Classical

6 1.1 1.2 -5.0 -5.0 Classical

7 1.3 1.4 -7.0 -7.0 Classical

8 1.7 1.8 -5.0 -5.0 Classical

9 2.3 2.4 -6.5 -6.0 Classical

10 3.1 3.0 -8.6 -9.0 Classical

11 3.2 3.2 -11.0 -11.0 Classical

12 5.0 5.0 -10.0 -10.0 Classical

Table 1-6. Equa lizationTest Types

Tap Number Relative Time ( )Average Relative Power

(dB)

Doppler

Spectrum

1 0.0 0.0 Classical

2 3.2 0.0 Classical

3 6.4 0.0 Classical

4 9.6 0.0 Classical

sec

sec

sec


32/701-28 PropGSM


3. Type= NoMultipat h s imulat es a single line-of-sight (LOS) pat h with th e

free-space pat hloss.

4. Ea ch Type option cont ains a u nique set of par am eters: the environm ent (rur al

h illy, or u rban), number of t aps, and two op tions for 6- and 12-tap set t ings. And

a m odel (Type=Equ alizat ionTest) is ar tificially crea ted t o tes t t he equalizer.

5. As an example, the average delay profile of each Type is dep icted in Figure 1-10

As shown in th is figur e, th e ru ra l environmen t is th e least hostile (roughly a

one-path non-dispersive model), while h illy and ur ban environm ent s a re

exam ples of more disper sive cha nn els.

6. Type opt ions include a r eferen ce delay of

where R is the initial dista nce between two an ten na s an d c is the free space

speed of light. If th ere is more th an one pa th in a propagat ion m odel, th e

relat ive delay of each pat h is with respect to 0.

7. For each Type option (except Type= NoMultipat h an d Type=Rur alArea ) the

pat hs a re a ssum ed to have a Ra yleigh en velope distribut ion with a (classical)

Doppler spectr um . For Type=NoMultipat h a simple Doppler sh ift is assu med.

5 12.8 0.0 Classical

6 16.0 0.0 Classical

Table 1-7. Rur al Types

Tap Number

Relative Time



1 0.1 0.0 0.0 0.0 Rician

2 0.2 0.2 -4.0 -2.0 Classical

3 0.3 0.4 -8.0 -10.0 Classical4 0.4 0.6 -12.0 -20.0 Classical

5 0.5 -16.0 Classical

6 0.6 -20.0 Classical

Table 1-6. Equa lizationTest Types

Tap Number Relative Time ( )Average Relative Power

(dB)

Doppler

Spectrum sec

sec

0

R

c

----=


33/70PropGSM 1-29

For Type=RuralArea , the firs t tap has a Rician envelope dis t r ibut ion implying a

direct LOS in a ddition to Rayleigh fading.

Figur e 1-10. Typical Avera ge Delay Pr ofile of GSM

Pr opagation Chan nels

8 . The path loss computa t ion is dynamic, wh ich m ea ns, du rin g sim ula tion du e t o

mobile travel the distance between th e tr an smit a nd r eceive anten na s is

cha nging and pa th loss is adjus ted a ccord ingly.

9. The par am eter Seed ra ndomizes th e out put of th e propagat ion m odel. A fixed

Seed results in th e sam e out put from simulation t o simulation an d am ong

models in a mu lti-cha nn el design.

10. The t est port is t he outpu t of a (user-selected) single path without pa th loss or

delay effects. The outpu t is t he gain fu nction , which is described in th eintr oductory pa rt of th is cha pter.

11. Typical short-term fading signal (Rayleigh) envelope and RF Doppler spectrum

ar e depicted in Figures 1-11 and 1-12.

Rural Environment

0

-20

10 20

Relative Delay (microseconds)

RelativePower(

dB)

0.50

Hilly Environment

RelativePower

(dB)


0

-20

0 6 10 15 20

Equalizer Test Profile

RelativePower(dB)


0 3.2 6.4 9.6 12.816.0 20

-20

0

Urban Environment

RelativePower(dB)


0 5 10 15 20

-20

0


34/701-30 PropGSM


Figur e 1-11. Typical Fadin g Envelope of Pr opGSM

Figure 1-12. Typical Doppler Spectr um of PropGSM



35/70PropNADCcdma 1-31

PropNADCcdma

Description Propagation Channel (CDMA) Model


Class TSDFPropNADCcdma


Parameters

Pin Inputs

Pin Outputs


Type propagation type:

NoMultipath, OnePath,

TwoPath, ThreePath

NoMultipath enum

Pathloss inclusion of large scale

pathloss: No, Yes

No enum

Env Environment Type Options:

TypicalUrban,

TypicalSuburban,

RuralArea, FreeSpace

TypicalUrban enum

Seed number to randomize

channel output

1234567 int

Test test port for single path

random function: Tap1,

Tap2, Tap3

Tap1 enum







36/701-32 PropNADCcdma


Notes/Equations

1. PropNADCcdma models a mult ipath channel based on IS-97 specificat ions. The

model is a m ulti-ta p filter, each ta p h aving a delay an d power. And, each t ap is

modula ted by a colored noise sour ce.

Including th e mu ltipat h fading and pa th loss effects, th e out put of th e cha nn elcan be described as

where

= pathloss att enua tion, 1

i = rela t ive power of each echo, per specificat ioni = relat ive delay of each echo, plus t he dir ect pa th delay 0s(t) = complex envelope inpu t s ignal

gi = ra ndom gain fun ction associated with ea ch echo

For th e NoMultipath option, M=1, 1=o



The pat hloss at ten ua tion factor =1 when Pa th loss=No. And, note th at whenPat hloss=No, the su m r(t) may add up t o be more th an th e input s(t), implyingth at cha nn el has a gain. For t his reason, when Path loss=No, the output is


For more deta ils, refer t o the In tr oduct ion on pa ge 1-1.

2. This component is based on th e Nort h Amer ican Dua l Model Cellular (NADC)

IS-97 specification.

3. Except for Type=NoMultipat h, where free space path loss a nd a pur e Doppler

sh ift is assumed, other Type opt ions have a Rayle igh envelope dis t r ibut ion witha (classical) Doppler spectr um an d pat hloss defined by th e En v para met er.

4. Figure 1-13 depicts two- an d t hr ee-pat h power delay profiles u sed in

PropNADCcdma.


gi

t( )i 1=

M

=


37/70PropNADCcdma 1-33

Figur e 1-13. Avera ge Delay Pr ofiles of CDMA Pr opa gat ion Chann els


where R is the initial dista nce between two an ten na s an d c is the free space

speed of light. If th ere is more th an one pa th in a propagat ion m odel, th e

relat ive delay of each pat h is with respect to 0.

6. The two- an d th ree-pat h m odels have a Rayleigh en velope distribut ion with a(class ical) Doppler spectrum .


Seed results in th e sam e out put from simulation to simulat ion an d am ong


8. The t est port is t he outpu t of a (user-selected) single path with out pat hloss or

delay effect s. The ou tpu t is the gain function descr ibed in the in t roductory par

of th is cha pter.

9. Sufficient simula tion time is required for accur at e pdf and cpdf. For evaluat ion

of envelope st at istics, it is prefera ble to have independent sam ples (th at is,

sam ples at a low ra te, which implies TStep is lar ge). The dur at ion of th e

simula tion (Stop par am eter of the da ta collecting sink ) should be lar ge enough

to cover 100 or more wavelengt hs of mobile t ra vel.

This tran slates to simulation time th at is great er tha n

CDMA Three-path

RelativePower(dB)

Relative Delay (microseconds)0 2 10.0 20

0

-3

14.5

CDMA Two-path

Rela

tivePower(dB)


0

-3

0 2 10.0 20

oR

c----=


38/701-34 PropNADCcdma


where

is th e ma ximum doppler frequen cy.

10. Typical probability density fun ction (pdf) an d cum ula tive probability dens ity

function (cpdf) of the signal envelope are depicted in Figures 1-14 and 1-15.

Sufficient s imu lat ion t ime is requir ed for a ccur at e pdf an d cpdf.

Figur e 1-14. Typical pdf of Signa l En velope

Figur e 1-15. Typical cpdf of Signa l En velope

100

---------

V

----=


39/70PropNADCtdma 1-35

PropNADCtdma

Description NADC Propagation (TDMA) Model, Directional


Class TSDFPropNADCtdma


Parameters

Pin Inputs


Type propagation type:

NoMultipath, FlatFade,

TwoPath

NoMultipath enum

Pathloss inclusion of large scale

pathloss: No, Yes

No enum

Env Environment Type Options:

TypicalUrban,

TypicalSuburban,


TypicalUrban enum

Delay relative delay

(Type=TwoPath) with

respect to first path, in

microseconds

0.0 real

Pwr relative power

(Type=TwoPath) with

respect to first path, in dB

0.0 real

Seed integer number to

randomize channel output

1234567 int

Test test port for single path:

Tap1, Tap2

Tap1 enum




40/701-36 PropNADCtdma


Pin Outputs

Notes/Equations

1. PropNADCtdma models a unid irect ional mult ipa th channel based on IS-55 and

IS-56 propagat ion specificat ions. The model is a mult i-tap filter, each tap having

a delay an d power. And , each t ap is modula ted by a colored n oise sour ce gain

fu nction described in th e intr oductory pa rt of th is cha pter.

Including th e mu ltipat h fading and pa th loss effects, th e out put of th e cha nn el

can be described as

where

= pathloss att enua tion, 1i = rela t ive power of each echo, per specificat ion

i = relat ive delay of each echo, plus t he dir ect pa th delay 0s(t) = complex envelope inpu t s ignalgi = ra ndom gain fun ction associated with ea ch echo

For t he N oMultipat h option, M=1, 1=0 .



The pat hloss at ten ua tion factor =1 when Pa th loss=No. And, note th at whenPat hloss=No, the su m r(t) may add up t o be more th an th e input s(t), implying

th at cha nn el has a gain. For t his reason, when Path loss=No, the output is


For more deta ils, refer t o the In tr oduct ion on pa ge 1-1.

2. This component is based on t he N ort h Amer ican Dua l Model Cellular (NADC).

3. Flat fade is a ssum ed t o be the non-frequen cy-selective fading.





g i t( )i 1=

M

=


41/70PropNADCtdma 1-37

4. Except for Type=NoMultipat h wh ere free spa ce pat hloss a nd a pur e Doppler

sh ift is assu med, all oth er Type opt ions h ave a Rayleigh envelope distr ibut ion

with a (classica l) Doppler spect rum and path loss defined by the Env parameter


where R is the initial distance between t wo an tenn as a nd c is th e free spa ce

speed of light.

Delay in t he t wo-ra y model is the delay of th e second r ay with respect t o th e

first ra y, which is a ssum ed to be 0.

Pwr in t he t wo-ra y model is th e power of th e second r ay below th e power of th e

first ra y.


Seed results in th e sam e out put from simulation to simulat ion an d am ong


7. The t est port is t he outpu t of a (user-selected) single path with out pat hloss or

delay effect s. The ou tpu t is the gain function descr ibed in the in t roductory par

of th is cha pter.

8 . Typica l envelope and envelope square spect rum of PropNADCtdma is shown in

Figure 1-16. Also, pat hloss profiles (no mu ltipa th ) for d ifferen t en vironm ent s

an d different sets of tr an smit a nd r eceive anten na height s a re sh own in

Figure 1-17.

oR

c----=


42/701-38 PropNADCtdma


Figure 1-16. Typical En velope an d En velope Squa re Spectr um

of PropNADCtdma

Figure 1-17. Pat hloss P rofiles for Different En vironmen ts

Envelope

Freespace

RuralSuburban

Urban


43/70PropWCDMA 1-39

PropWCDMA

Description WCDMA Propagation Channel


Class TSDFPropWCDMA


Parameters

Pin Inputs

Name Description Default Sym Type Range

ChannelType Indicate the test

environment ( Delay,

Power, and Doppler

Spectrum of each path ).:

NoMultipath, Indoor A,

Indoor B, Pedestrian A,

Pedestrian B, Vehicular A,

Vehicular B

Vehicular B enum

Pathloss option for inclusion of

large-scale pathloss: No,

Yes

No enum

Environment IMT2000 environment for

pathloss computation:

Indoor, Pedestrian,

Vehicular, FreeSpace

FreeSpace enum

NumberOfFloors number of floors for indoor

pathloss

4 int (0, )

N 2N+1 is the number of sine

waves used in Jakes model

10 N int (0, )


randomize the channel

output (Jakes model)

1234567 int (0, )




44/701-40 PropWCDMA


Pin Outputs

Notes/Equations

1. This component models a fading cha nn el based on t he IMT2000 sta nda rd

specification. The specification includes the delay spread, doppler spread and

pa th loss for var ious en vironm ent s. In a ddition to these, a line of sight (LOS)

opt ion called NoMult ipat h is ava ilable wher e th e doppler sh ift du e to mobility

a n d t h e LOS dela y a r e m odeled. Th e r esult s a r e a va ila ble a t ch a n nel ou t pu t pin

mOut.

2. For doppler spr ead, J ak es model [2] is used which provides th e doppler

spectr um as well as th e sta tistics of th e fading cha nn el. An a dditiona l out put

pin t estOut is provided tha t conveys th e complex gain mu ltipliers of J ak es

model.

3. The delay sprea d is modeled via a ta p delay line wher e the n um ber of ta ps is

based on t he IMT2000 specificat ions for a given environm ent . There a re 6

options (A an d B for in door, pedestrian an d vehicular environm ent s) th at ar e

depicted in Table 1-8 through Table 1-10. In ea ch case the inpu t signal is

delayed and the car r ier phase due to the delay s igna l is incorpora ted . Howeverin the nar rowband case (FLAT opt ion in the indoor office tes t environment) only

th e car rier ph ase cha nge associated with t he mu ltipat h delay is included.

The modeling pr ocess is depicted in Figure 1-18, which indicat es th at th e

ou tpu t of complex mult ip lier s a re summed . Th is is the case when PropWCDMA

cha nn el is conn ected t o a simple an ten na . However, when t he chan nel is not

conn ected t o an ar ra y ant enn a, th e out put of complex multipliers a re n ot

sum med but each m ultipat h (echo) acts like an independen t wavefront

impinging on t he a rr ay elemen ts (see WCDMAVectorCh an nel).4. Un less stat ed oth erwise, th e use model is consist ent with th e guidelines

described in t he In tr oduction on pa ge 1-1.

5. .The delay profile for t he var ious cha nn el types, summ ar ized in Table 1-8

through Table 1-10, resu lt in t he frequen cy select ive fad ing.


2 testOut complex gain of the channel complex

3 mout Antenna input signal multiple timed


45/70PropWCDMA 1-41

Table 1-8. Indoor Office Test En vironmen t Tapped-Delay-Line Par am eters

Tap

Channel A Channel B

Doppler

Spectrum

Relative Delay

(nSec) Avg. Power (dB)

Relative Delay


1 0 0 0 0 FLAT

2 50 -3.0 100 -3.6 FLAT

3 110 -10.0 200 -7.2 FLAT

4 170 -18.0 300 -10.8 FLAT

5 290 -26.0 500 -18.0 FLAT

6 310 -32.0 700 -25.2 FLAT

Table 1-9. Out door t o Indoor a nd Pedestr ian Test En vironmen t

Tapped-Delay-Line Pa ra met ers

Tap

Channel A Channel B

Doppler Spectrum

Relative Delay


Relative Delay


1 0 0 0 0 CLASSIC

2 110 -9.7 200 -0.9 CLASSIC

3 190 -19.2 800 -4.9 CLASSIC

4 410 -22.8 1200 -8.0 CLASSIC

5 2300 -7.8 CLASSIC6 3700 -23.9 CLASSIC

Table 1-10. Vehicular Test E nvironm ent , High Anten na ,

Tapped-Delay-Line Pa ra met ers

Tap

Channel A Channel B

Doppler

SpectrumRelative Delay (nSec) Avg. Power (dB)

Relative Delay


1 0 0 0 -2.5 CLASSIC

2 310 -1.0 300 0 CLASSIC

3 710 -9.0 8900 -12.8 CLASSIC

4 1090 -10.0 12900 -10.0 CLASSIC

5 1730 -15.0 17100 -25.2 CLASSIC

6 2510 -20.0 20000 -16.0 CLASSIC


46/701-42 PropWCDMA


Figure 1-18. Delay an d Doppler Spread an d Car rier Ph ase Shift

References

[1]Draft New Recommendation ITU-R M.[FPLMT.REVAL], Guidelines for

Evaluation of Radio Transmission Technologies for IMT-2000/FPLMTS

(Quest ion ITU-R 39/8).

[2] W. C. Jakes, Microwave M obile Com m unication s, IEEE Pr ess, 1994.

Delay Delay Delay DelayDelay

Power

In

Out

Distribution

JakesModel


47/70UserDefChannel 1-43

UserDefChannel

Description User-Defined Channel


Class TSDFUserDefChannel


Parameters

Pin Inputs


PathNumber number of multipath echos 2 L int (0, )AmpArray amplitude weight of path

array

1.0 0.5 real array (-, )

DelayArray delays associated with

path array, in

microseconds

0.1 0.2 real array (-, )


randomize channel output

(Jakes model)

1234567 int (0, )

N 2N+1 is number of sine


10 N int (0, )



Yes

No enum

Env environment for pathloss

computation:TypicalUrban,

TypicalSuburban,


TypicalUrban enum




48/701-44 UserDefChannel


Pin Outputs

Notes/Equations

1. This component models a fading cha nn el based on u ser-specified mu ltipath

delay profile via t he DelayArr ay an d AmpArra y para met ers. The resu lts ar e

available at t he cha nn el out put port .

2. For doppler spread , Ja kes model [1] is used, which pr ovides the doppler

spectr um as well as t he st at istics of th e fading cha nn el.

3. The delay sprea d is modeled via a ta p delay line wher e the n um ber of ta ps is

ba sed on t h e s ize of Dela yAr r ay a n d Am pAr r ay. In ea ch ca se t h e in pu t sign a l isdelayed an d th e car rier ph ase du e to the delay signal is incorporat ed.

Figure 1-19 illustr at es t his m odeling pr ocess when conn ected t o a simple

antenna.

4. Un less stat ed oth erwise, th e use model is consist ent with th e guidelines

described in th e In tr oduction on pa ge 1-1. The d elay profile specified by the

user determ ines th e frequen cy selective nat ur e of th e cha nn el.





Jakes

Model

Power

In

Out

Distribution


49/70UserDefChannel 1-45

References



50/701-46 UserDefVectorChannel


UserDefVectorChannel

Description User-Defined Vector Channel


Class TSDFUserDefVectorChannel


Parameters

Pin Inputs


PathNumber number of multipath echos 2 L int (0, )AmpArray amplitude weight of path

array

1.0 0.5 real array (-, )


path array, in

microseconds

0.1 0.2 real array (-, )



output (Jakes model)

1234567 int (0, )



10 N int (0, )



Yes

No enum

Env environment for pathloss

computation:TypicalUrban,

TypicalSuburban,


TypicalUrban enum




51/70UserDefVectorChannel 1-47

Pin Outputs

Notes/Equations

1. UserDefVectorCha nn el is ident ical t o UserDefCha nn el, except t he outp ut of

th is model keeps the m ultipat hs (echos) inta ct t hu s a llowing a vector cha nn el

simula tion when an Ant Arr ay component or a ny mu ltiple timed input port is

connected.

2. Figure 1-20 illus t ra tes the user -defined vector channel model when its output is

conn ected to an a rra y antenn a.

Figur e 1-20. User-Defined Vector Cha nn el Model when

Conn ected t o an Arr ay Ant enn a


2 mout multiport channel output multiple timed


Jakes

Model

Power

In

Distribution


52/701-48 3GPPFDD_Channel


3GPPFDD_Channel

Description Fading channel model


Parameters

Pin Inputs

Name Description Default Unit Type RangeSpecVersion version of specifications:

Version_03_00,

Version_12_00,

Version_03_02

Version_12_00 enum

Velocity mobile velocity in km/hour 100km km real (0, )GainArray path gain in terms of

decibel

0dB -10dB real array (-, )

DelayArray path delay in terms of

nanosecond

0ns 976ns int array [0, )

RicianFactor ratio of specular power and

the fading power

0.0 real[0, )

Algorithm the algorithm used:

SoS_Stochastic,

SoS_Deterministic,

Filter_AWGN_Noise

SoS_Stochastic enum

ChProfile channel profile:

UserDefined, Case1,

Case2, Case3, Case4,

Case5

UserDefined enum

for each array element; number of elements in arrays GainArray and DelayArray must be the same


1 in input timed signal timed


53/703GPPFDD_Channel 1-49

Pin Outputs

Notes/Equations

1. This subnetwork model can be used to simulat e th e mu ltipat h fading cha nn el:

according to 3GPP a s defined in [1], when ChP rofile is set to Case1, Case2,

Case3, Case4, or Ca se5.

according to user-defined fading cha nn el profile, when ChP rofile is set to

UserDefined.

2. This subnet work is a hiera rchical model th at includes the Fader component ,

which is locat ed in th e Ant enn as & Propagat ion librar y; 3GPPF DD_Cha nn el

provides the int erface to Fader. The schematic for t his su bnet work is shown in

Figure 1-21.

Figur e 1-21. 3GPPFDD_Chan nel Schema tic

3. When Ch Pr ofile=UserDefined, the fading cha nn el profile (relat ive cha nn el

delay sprea d an d average power) is determined by the extern al input from th e

GainArra y an d DelayArra y para meters.

When Ch Pr ofile is set t o a pr e-defined fading cha nn el profile defined for 3GP P,

th e relat ive cha nn el delay spread a nd a verage power ar e given in Table 1-11.


2 out output signal timed

3 outChM fading factor multiple timed


54/701-50 3GPPFDD_Channel


4. RicianFactor is the ra t io of d ir ect s igna l power and fad ing power ; r efer to Fader

documentation for details.

References

[1]3GPP Techn ical Specificat ion TS 25.101 V3.10.0, UE Rad io tr an sm ission a nd

reception (FDD) March 2003, Release 1999.

http://www.3gpp.org/ftp/Specs/2002-03/R1999/25_series/25101-3a0.zip

Table 1-11. Cha nn el Pr ofile Configur at ions

Case 1,

3km/h

Case 2,

3 km/h

Case 3,

120 km/h

Case 4,

3 km/h

Case 5,

50 km/h

Relative

Delay(nsec)

Average

Power(dB)

Relative

Delay(nsec)

Average

Power(dB)

Relative

Delay(nsec)

Average

Power(dB)

Relative

Delay(nsec)

Average

Power(dB)

Relative

Delay(nsec)

Average

Power(dB)

0 0 0 0 0 0 0 0 0 0

976 -10 976 0 260 -3 976 0 976 -10

20000 0 521 -6

781 -9


55/70TDSCDMA_FwdChannel 1-51

TDSCDMA_FwdChannel

Description Multipath fading channel for forward link


Parameters

Pin Inputs

Pin Outputs

Notes/Equations

1. This subn etwork is used t o simula te propagat ion conditions for m ultipat h

fading en vironmen ts.

The schema tic for th is subn etwork is shown in Figure 1-22.

Ea ch firin g, 1 Out put token is produced when 1 Inpu t is consu med.


Case propagation conditions for

multipath fading

environments: case_1,

case_2, case_3

case_1 enum


1 Input input data timed


2 Output data after fading channel timed


56/701-52 TDSCDMA_FwdChannel


Figure 1-22. TDSCDMA_FwdCha nn el Schem at ic

Reference

[1] 3GPP TS 25.142, 3rd Generation Partn ership Project; Technical S pecification

Group Radio Access Network; Base station conformance testing (TDD) (Release

4), version 4.5.0, Jun ., 2002.


57/70TDSCDMA_RevChannel 1-53

TDSCDMA_RevChannel

Description Multipath fading channel for reverse link


Parameters

Pin Inputs

Pin Outputs

Notes/Equations

1. This subn etwork is used t o simula te propagat ion conditions for m ultipat h

fading en vironmen ts.

The schema tic for th is subn etwork is shown in Figure 1-23.

Ea ch firin g, 1 Out put token is produced when 1 Inpu t is consu med.


Case propagation conditions for

multipath fading

environments: case_1,

case_2, case_3

case_1 enum


1 Input input data timed


2 Output data after fading channel timed


58/701-54 TDSCDMA_RevChannel


Figur e 1-23. Schem at ic of TDSCDMA_RevChann el

References

[1] 3GPP TS 25.142, 3rd Generation Partn ership Project; Technical S pecification

Group Radio Access Network; Base station conformance testing (TDD) (Release

4), vers ion 4.5.0, J un e 2002.


59/70WCDMAVectorChannel 1-55

WCDMAVectorChannel

Description WCDMA Vector Channel


Parameters

Pin Inputs


ChannelType type of Channel:

NoMultipath, Indoor A,

Indoor B, Pedestrian A,

Pedestrian B, Vehicular A,

Vehicular B

Vehicular B enum

Echos number of multipath echos 6 L int (0, )Pathloss option for inclusion of


Yes

No enum

Environment IMT2000 environment for

pathloss computation:

Indoor, Pedestrian,Vehicular, FreeSpace

FreeSpace enum

NumberOfFloors number of floors for indoor

option

4 int (0, )



10 N int (0, )



output(Jakes model)

1234567 int (0, )


1 input channel input complex


60/701-56 WCDMAVectorChannel


Pin Outputs

Notes/Equations

1. WCDMAVectorChannel is a subcircuit design consist ing of PropWCDMA and

Bus componen t s. The s ize of Bus must equa l the number of mult ipath (echos) in

th e cha nn el; oth erwise, WCDMAVectorCh an nel a cts ident ical t o PropWCDMA

The conn ectin g Bus a llows each mult ipat h (echo) to be int act wh en it is

conn ected t o an Ant Arr ay component or a ny mu ltiple timed inpu t port .

2. Figure 1-24 illustr at es th e WCDMA vector cha nn el model out put when

conn ected to an a rra y antenn a.

Figur e 1-24. WCDMA Vector Ch an nel Model when

Conn ected t o an Arr ay Ant enn a


2 output channel output multiple complex


Jakes

Model

Power

In

Distribution


61/70WLAN_ChannelModel 1-57

WLAN_ChannelModel

Description channel model


Class TSDFWLAN_ChannelModel


Parameters


Algorithm fading algorithm: Jakes,

NoiseFilter

NoiseFilter enum

ModelType fading model: NoMultipath,

A, B, C, D, E, UserDefined

A enum

PathNumber number of multipath

echoes, effective only

when ModelType is set as

UserDefined

4 int [1, 150]

PowerArray average relative power ofpath array, in dB, effective

only when ModelType is

set as UserDefined

0.0 -14 -18 -20 real array (-, )


path array, in nsec,

effective only when

ModelType is set as

UserDefined

0.0 56 106 185 real array [0, )

FadingType fading type of first path,

effective only when

ModelType is set as

UserDefined: Rayleigh,

Ricean

Rayleigh enum

RiceanFactor Ricean factor, effective

only when ModelType is

set as UserDefined and

FadingType is set as

Ricean

10 int [1, )


randomize channel

output(Jakes model)

1234567 int (-, )


62/701-58 WLAN_ChannelModel


Pin Inputs

Pin Outputs

Notes/Equations

1. This component is used to simula te a m ulti-pat h fading cha nn el based on a

ta pped-delay line model. Ea ch firing, one t oken is consu med a t th e input pin,an d one t oken is produced at th e out put pin.

2. The mu ltipath delay profile determ ines th e frequen cy selective na tu re of th e

cha nn el. Delay profile is specified by the user via t he ModelType, DelayArr ay

an d PowerArray para meters.

3. The fading type of th e first pat h can be Rayleigh or Ricean ; if FadingType =

Ricean , RiceanFactor determines the power ra t io of the direct s ignal to a ll other

indirect signa ls.

4. If ModelType = A, B, C, D or E , th e Pa thN um ber, DelayArra y, FadingType, andRicean Factor pa ra met ers a re a ut oma tically set a ccording t o [2].

If ModelType = UserDefined, the u ser det erm ines cha nn el cha ra cter istics by

setting th ese para meters.

If ModelType = NoMult ipat h, only Doppler frequ ency shift is incorpora ted .

N number of oscillators in

Jakes model

80 int (PathNumber,

)Pathloss option for inclusion of


Yes

No enum

Env environment type options:

TypicalUrban,

TypicalSuburban,


TypicalUrban enum








5. Th e dela y sp rea d is m odeled via a t a pped dela y lin e wh er e t h e n u mber of t a ps is

based on t he s ize of DelayArr ay a nd PowerArr ay. In ea ch case th e inpu t signa l

is delayed an d th e car rier ph ase du e to the delay signal is incorporat ed.

Figure 1-25 illustr at es t his m odeling pr ocess when conn ected t o a simple

antenna.


6. If DelayArr ay values a re n ot th e time as Tstep, th e interpolation is m ade t ogain t his point of signal a nd a delay of 64 tokens is in tr oduced.

7. For ea ch ta p, Ja kes m odel or noise filter m odel provides Doppler spectr um as

well a s th e fad ing cha nn el specifica tions.

Ja kes model uses N0 low-frequ ency oscillators to genera te a fad ing

waveform.

where

.


J/NF

Model

Power

In

Out

Distribution

T t

( ) n( )cos i

n( )sin+

[ ] nt

n+

( )cos

n 1=

N0

=

n m 2n N( )cos=

N 2 2 N0 1+( )=




The maximum Doppler frequency offset m = 2V/, denotes wave-length ocarrier. n is set a s n = 2/N0. Original ph ase shifts n ar e ran domized andun iform ly dist ribu ted over [0, 2].N0 = 80 is u sed in t he component .

Deta ils about J ak es model can be foun d in r eferen ce [1].

Noise filter m odel feeds white Gaussia n n oise to a digital filter ma tched to

th e respective fading spectru m, i.e., th e filter frequency response mu st be a

good approximat ion to the square root of the (normalized) tap power spect ra

density. An in ter pola tor with ra te conver sion factor follows. Cha nging t ap

Doppler sp rea ds is done by simply cha nging the in ter polat ion factor.

This implement at ion is illust ra ted in Figure 1-26. The AWGN, filter, and

in terpola tor ou tpu t s a re complex. The filt er and AWGN opera te a t the s igna

sam pling ra te. An 8t h-order IIR filter is used for class ic an d flat spectr um s

whose norm alized fad ing frequ encies ar e 0.05686. A simple linea rint erpolat or often su ffices for changing Doppler sp rea ds.

Figure 1-26. Noise Filter Model Implementa tion

8. If speed is zero, the chan nel is not a time-var iant . However, multipa th still

exists so a st at ic cha nn el is a pplied an d each ta p is a complex const an t

value.The complex constant value is randomly genera ted and can be changed by

cha nging t he Seed pa ra meter.

9. Five model types ha ve been designed ba sed on r eferen ce [2]:

Model A corr esponds to a typical office environm ent .

Model B corr esponds to a t ypical lar ge open-space environmen t with NLOS

conditions or an office environm ent with a large delay sprea d.

AWGN Filter Interpolator

Delay(0)

C0

s(t)

AWGN Filter Interpolator

Delay(n)

Cn

.

.

.

0

n



Models C an d E corr espond to typical lar ge open-space indoor an d outdoor

environments, respectively, with a large delay spread.

Model D corr esponds t o LOS cond itions in a la rge open-spa ce indoor or

out door environm ent .

Cha ra cter istics of th ese models a re listed in Table 1-12through Table 1-16.

Table 1-12. Model A: Typical Office Environm ent with

NLOS Conditions an d 50ns Average rm s Delay Spread

Tap Number Delay(ns) Average Relative Power Ricean K

Doppler

Spectrum

1 0 0.0 0 Class

2 10 -0.9 0 Class

3 20 -1.7 0 Class

4 30 -2.6 0 Class

5 40 -3.5 0 Class

6 50 -4.3 0 Class

7 60 -5.2 0 Class

8 70 -6.1 0 Class

9 80 -6.9 0 Class

10 90 -7.8 0 Class

11 110 -4.7 0 Class

12 140 -7.3 0 Class

13 170 -9.9 0 Class

14 200 -12.5 0 Class

15 240 -13.7 0 Class

16 290 -18.0 0 Class

17 340 -22.4 0 Class

18 390 -26.7 0 Class

Table 1-13. Model B: Typical Large Open-Space an d Office En vironm ent swith NLOS Conditions an d 100ns Average rm s Delay Spread


Doppler

Spectrum

1 0 -2.6 0 Class

2 10 -3.0 0 Class

3 20 -3.5 0 Class




4 30 -3.9 0 Class

5 50 0.0 0 Class

6 80 -1.3 0 Class

7 110 -2.6 0 Class

8 140 -3.9 0 Class

9 180 -3.4 0 Class

10 230 -5.6 0 Class

11 280 -7.7 0 Class

12 330 -9.9 0 Class

13 380 -12.1 0 Class

14 430 -14.3 0 Class

15 490 -15.4 0 Class

16 560 -18.4 0 Class

17 640 -20.7 0 Class

18 730 -24.6 0 Class

Table 1-14. Model C: Typical La rge Open Space En vironmen t with



Doppler

Spectrum

1 0 -3.3 0 Class

2 10 -3.6 0 Class

3 20 -3.9 0 Class

4 30 -4.2 0 Class

5 50 0.0 0 Class

6 80 -0.9 0 Class

7 110 -1.7 0 Class

8 140 -2.6 0 Class

9 180 -1.5 0 Class

10 230 -3.0 0 Class

11 280 -4.4 0 Class

12 330 -5.9 0 Class

13 400 -5.3 0 Class

Table 1-13. Model B: Typical Large Open-Space an d Office En vironm ent s

with NLOS Conditions a nd 100ns Average rms Delay Spread (cont inued)


Doppler

Spectrum



14 490 -7.9 0 Class

15 600 -9.4 0 Class

16 730 -13.2 0 Class

17 880 -16.3 0 Class

18 1050 -21.2 0 Class

Table 1-15. Model D: Typical La rge Open Space En vironmen t with

LOS Conditions an d 150ns Average rm s Delay Spread;

a 10 dB spike at zero delay has been a dded result ing in a n r ms delay sprea d of

appr oxima tely 140ns


Doppler

Spectrum

1 0 0.0 10 Class+spike

2 10 -10.0 0 Class

3 20 -10.3 0 Class

4 30 -10.6 0 Class

5 50 -6.4 0 Class

6 80 -7.2 0 Class

7 110 -8.1 0 Class

8 140 -9.0 0 Class

9 180 -7.9 0 Class

10 230 -9.4 0 Class

11 280 -10.8 0 Class

12 330 -12.3 0 Class

13 400 -11.7 0 Class

14 490 -14.3 0 Class

15 600 -15.8 0 Class

16 730 -19.6 0 Class

17 880 -22.7 0 Class

18 1050 -27.6 0 Class

Table 1-14. Model C: Typical La rge Open Space En vironmen t with

NLOS Conditions a nd 150ns Average rms Delay Spread (cont inued)


Doppler

Spectrum




References


[2] Cha nn el Models for H IPERLAN/2 in Differen t In door Scena rios, ETSI EP

BRAN 3E R1085B 30 March 1998.

Table 1-16. Model E: Typical La rge Open Spa ce En vironmen t with



Doppler

Spectrum

1 0 -4.9 0 Class

2 10 -5.1 0 Class

3 20 -5.2 0 Class

4 40 -0.8 0 Class

5 70 -1.3 0 Class

6 100 -1.9 0 Class

7 140 -0.3 0 Class

8 190 -1.2 0 Class

9 240 -2.1 0 Class

10 320 0.0 0 Class

11 430 -1.9 0 Class

12 560 -2.8 0 Class

13 710 -5.4 0 Class

14 880 -7.3 0 Class

15 1070 -10.6 0 Class

16 1280 -13.4 0 Class

17 1510 -17.4 0 Class

18 1760 -20.9 0 Class


69/70Index-1

Index

Numerics3GPPFDD_Channel, 1-48

AAntArray, 1-11AntBase, 1-14AntMobile, 1-16

FFader, 1-18

PPropFlatEarth, 1-21PropGSM, 1-24PropNADCcdma, 1-31PropNADCtdma, 1-35PropWCDMA, 1-39

TTDSCDMA_FwdChannel, 1-51TDSCDMA_RevChannel, 1-53

UUserDefChannel, 1-43UserDefVectorChannel, 1-46

WWCDMAVectorChannel , 1-55WLAN_ChannelModel, 1-57


70/70

antennas prop

Documents