3GPP TR V ( )

Transcription

1 TR V.. (-3 Techical Report 3rd Geeratio Partership Project; Techical Specificatio Group Radio Access Network; Spatial chael odel for Multiple Iput Multiple Output (MIMO siulatios (Release The preset docuet has bee developed withi the 3 rd Geeratio Partership Project ( TM ad ay be further elaborated for the purposes of. The preset docuet has ot bee subject to ay approval process by the Orgaizatioal Parters ad shall ot be ipleeted. This Specificatio is provided for future developet work withi oly. The Orgaizatioal Parters accept o liability for ay use of this Specificatio. Specificatios ad reports for ipleetatio of the TM syste should be obtaied via the Orgaizatioal Parters' Publicatios Offices.

2 Release TR V.. (-3 Keywords UMTS, radio, atea Postal address support office address 65 Route des Lucioles - Sophia Atipolis Valboe - FRANCE Tel.: Fax: Iteret Copyright Notificatio No part ay be reproduced except as authorized by writte perissio. The copyright ad the foregoig restrictio exted to reproductio i all edia., Orgaizatioal Parters (ARIB, ATIS, CCSA, ETSI, TTA, TTC. All rights reserved. UMTS is a Trade Mark of ETSI registered for the beefit of its ebers is a Trade Mark of ETSI registered for the beefit of its Mebers ad of the Orgaizatioal Parters LTE is a Trade Mark of ETSI curretly beig registered for the beefit of its Mebers ad of the Orgaizatioal Parters GSM ad the GSM logo are registered ad owed by the GSM Associatio

3 Release 3 TR V.. (-3 Cotets Foreword... 4 Scope... 5 Refereces Defiitios, sybols ad abbreviatios Defiitios Sybols Abbreviatios Spatial chael odel for calibratio purposes Purpose Lik level chael odel paraeter suary Spatial paraeters per path BS ad MS array topologies Spatial paraeters for the BS BS atea patter Per-path BS agle spread (AS Per-path BS agle of departure Per-path BS power aziuth spectru Spatial paraeters for the MS MS atea patter Per-path MS agle spread (AS Per-path MS agle of arrival Per-path MS power aziuth spectru MS directio of travel Per-path Doppler spectru Geeratio of chael odel Calibratio ad referece values Spatial chael odel for siulatios Geeral defiitios, paraeters, ad assuptios Eviroets Geeratig user paraeters Geeratig user paraeters for urba acrocell ad suburba acrocell eviroets Geeratig user paraeters for urba icrocell eviroets Geeratig chael coefficiets Optioal syste siulatio features Polarized arrays Far scatterer clusters Lie of sight Urba cayo Correlatio betwee chael paraeters Modelig itercell iterferece Syste Level Calibratio... 3 Aex A: Calculatio of circular agle spread Aex B: Chage history... 4

4 Release 4 TR V.. (-3 Foreword This Techical Report has bee produced by the 3 rd Geeratio Partership Project (. The cotets of the preset docuet are subject to cotiuig work withi the TSG ad ay chage followig foral TSG approval. Should the TSG odify the cotets of the preset docuet, it will be re-released by the TSG with a idetifyig chage of release date ad a icrease i versio uber as follows: Versio x.y.z where: x the first digit: preseted to TSG for iforatio; preseted to TSG for approval; 3 or greater idicates TSG approved docuet uder chage cotrol. y the secod digit is icreeted for all chages of substace, i.e. techical ehaceets, correctios, updates, etc. z the third digit is icreeted whe editorial oly chages have bee icorporated i the docuet.

5 Release 5 TR V.. (-3 Scope The preset docuet details the output of the cobied - spatial chael odel (SCM ad-hoc group (AHG. The scope of the - SCM AHG is to develop ad specify paraeters ad ethods associated with the spatial chael odellig that are coo to the eeds of the ad orgaizatios. The scope icludes developet of specificatios for: Syste level evaluatio. Withi this category, a list of four focus areas are idetified, however the ephasis of the SCM AHG work is o ites a ad b. a Physical paraeters (e.g. power delay profiles, agle spreads, depedecies betwee paraeters b Syste evaluatio ethodology. c Atea arrageets, referece cases ad defiitio of iiu requireets. d Soe fraework (air iterface depedet paraeters. Lik level evaluatio. The lik level odels are defied oly for calibratio purposes. It is a coo view withi the group that the lik level siulatio assuptios will ot be used for evaluatio ad copariso of proposals. Refereces The followig docuets cotai provisios which, through referece i this text, costitute provisios of the preset docuet. Refereces are either specific (idetified by date of publicatio editio uber, versio uber, etc. or o-specific. For a specific referece, subsequet revisios do ot apply. For a o-specific referece, the latest versio applies. I the case of a referece to a docuet (icludig a GSM docuet, a o-specific referece iplicitly refers to the latest versio of that docuet i the sae Release as the preset docuet. [] H. M. Foster, S. F. Dehgha R. Steele, J. J. Stefaov, H. K. Strelouhov, Role of Site Shieldig i Predictio Models for Urba Radiowave Propagatio (Digest No. 994/3, IEE Colloquiu o Microcellular easureets ad their predictio 994 pp. /-/6. [] L. Greestei V. Erceg, Y. S. Yeh, M. V. Clark, A New Path-Gai/Delay-Spread Propagatio Model for Digital Cellular Chaels, IEEE Trasactios o Vehicular Techology, VOL. 46, NO., May 997, pp [3] L. M. Correia, Wireless Flexible Persoalized Couicatios, COST 59: Europea Cooperatio i Mobile Radio Research, Chichester: Joh Wiley & Sos,. Sec. 3. (M. Steibauer ad A. F. Molisch, "Directioal chael odels".

6 Release 6 TR V.. (-3 3 Defiitios, sybols ad abbreviatios 3. Defiitios For the purposes of the preset docuet, the followig ters ad defiitios apply. Path: Ray Path Copoet: Sub-ray 3. Sybols For the purposes of the preset docuet, the followig sybols apply: AS DS SF Agle Spread or Aziuth Spread (Note: uless otherwise stated, the calculatio of agle spread will be based o the circular ethod preseted i appedix A delay spread logoral shadow fadig rado variable SH log oral shadow fadig costat ( a, b represets a rado oral (Gaussia distributio with ea a ad variace b. 3.3 Abbreviatios For the purposes of the preset docuet, the followig abbreviatios apply: AHG AoA AoD Ad Hoc Group Agle of Arrival Agle of Departure AS Agle Spread = Aziuth Spread = AS (Note: uless otherwise stated, the calculatio of agle spread will be based o the circular ethod preseted i appedix A BS Base Statio = Node-B = BTS DoT Directio of Travel DS delay spread = DS MS Mobile Statio = UE = Terial = Subscriber Uit PAS Power Aziuth Spectru PDP Power Delay Profile PL Path Loss SCM Spacial Chael Model SF logoral shadow fadig rado variable = SF SH log oral shadow fadig costat = SH UE User Equipet = MS 4 Spatial chael odel for calibratio purposes This clause describes physical paraeters for lik level odellig for the purpose of calibratio. 4. Purpose Lik level siulatios aloe will ot be used for algorith copariso because they reflect oly oe sapshot of the chael behaviour. Furtherore, they do ot accout for syste attributes such as schedulig ad HARQ. For these reasos, lik level siulatios do ot allow ay coclusios about the typical behaviour of the syste. Oly syste

7 Node B/ Base Statio UE/Mobile Statio Relative Path Power (db Delay (s Release 7 TR V.. (-3 level siulatios ca achieve that. Therefore this docuet targets syste level siulatios for the fial algorith copariso. Lik level siulatios will ot be used to copare perforace of differet algoriths. Rather, they will be used oly for calibratio which is the copariso of perforace results fro differet ipleetatios of a give algorith. The descriptio is i the cotext of a dowlik syste where the BS trasits to a MS; however the aterial i this 4. Lik level chael odel paraeter suary The table below suarizes the physical paraeters to be used for lik level odellig. Table 4.: Suary SCM lik level paraeters for calibratio purposes Model Case I Case II Case III Case IV Correspodig Case B Case C Case D Case A Desigator* Correspodig Model A, D, E Model C Model B Model F Desigator* PDP Modified Pedestria A Vehicular A Pedestria B Sigle Path # of Paths 4+ (LOS o K = 6dB 4 (LOS off If, Speed (k/h 3 3, 3, 3, 3, 3 3, Topology Referece.5λ Referece.5λ Referece.5λ N/A PAS LOS o: Fixed AoA for RMS agle RMS agle spread N/A LOS copoet, reaiig spread of 35 of 35 degrees per power has 36 degree degrees per path path with a uifor PAS. with a Laplacia Laplacia LOS off: PAS with a distributio distributio Laplacia distributio RMS Or 36 degree agle spread of 35 degrees uifor PAS. per path DoT N/A NOTE: (degrees AoA (degrees Topology PAS.5 (LOS copoet 67.5 (all other paths 67.5 (all paths.5 (odd ubered paths, (eve ubered paths Referece: ULA with.5λ-spacig or 4λ-spacig or λ-spacig Laplacia distributio with RMS agle spread of degrees or 5 degrees, per path depedig o AoA/AoD AoD/AoA 5 for RMS agle spread per path (degrees for 5 RMS agle spread per path *Desigators correspod to chael odels previously proposed i ad ad-hoc groups. N/A N/A N/A N/A

8 Release 8 TR V.. ( Spatial paraeters per path Each resolvable path is characterized by its ow spatial chael paraeters (agle spread, agle of arrival, power aziuth spectru. All paths are assued idepedet. These assuptios apply to both the BS ad the MS specific spatial paraeters. The above assuptios are i effect oly for the Lik Level chael odel. 4.4 BS ad MS array topologies The spatial chael odel should allow ay type of atea cofiguratio to be selected, although details of a give cofiguratio ust be shared to allow others to reproduce the odel ad verify the results. Calibratig siulators at the lik level requires a coo set of assuptios icludig a specific set of atea topologies to defie a baselie case. At the MS, the referece eleet spacig is.5, where is the wavelegth of the carrier frequecy. At the BS, three values for referece eleet spacig are defied:.5, 4, ad. 4.5 Spatial paraeters for the BS 4.5. BS atea patter The 3-sector atea patter used for each sector, Reverse Lik ad Forward Lik, is plotted i Figure 4. ad is specified by 3dB A i, A where 8 8 is defied as the agle betwee the directio of iterest ad the boresight of the atea, 3 db is the 3dB beawidth i degrees, ad A is the axiu atteuatio. For a 3 sector sceario 3 db is 7 degrees, A db,ad the atea boresight poitig directio is give by Figure 4.. For a 6 sector sceario 3 db is 35 o, A =3dB, which results i the patter show i Figure 4.3, ad the boresight poitig directio defied by Figure 4.4. The boresight is defied to be the directio to which the atea shows the axiu gai. The gai for the 3-sector 7 degree atea is 4dBi. By reducig the beawidth by half to 35 degrees, the correspodig gai will be 3dB higher resultig i 7dBi. The atea patter show is targeted for diversity-orieted ipleetatios (i.e. large iter-eleet spacigs. For beaforig applicatios that require sall spacigs, alterative atea desigs ay have to be cosidered leadig to a differet atea patter.

9 Gai i db. Release 9 TR V.. (-3 3 Sector Atea Patter Aziuth i Degrees Figure 4.: Atea patter for 3-sector cells 3-Sector Sceario BS Atea Boresight i directio of arrow Figure 4.: Boresight poitig directio for 3-sector cells

10 Gai i db Release TR V.. (-3 6 Sector Atea Patter Aziuth i Degrees Figure 4.3: Atea patter for 6-sector cells 6-Sector Boudaries BS Atea Boresight i directio of arrow Figure 4.4: Boresight poitig directio for 6-sector cells 4.5. Per-path BS agle spread (AS The base statio per-path agle spread is defied as the root ea square (RMS of agles with which a arrivig path s power is received by the base statio array. The idividual path powers are defied i the teporal chael odel described i Table 4.. Two values of BS agle spread (each associated with a correspodig ea agle of departure, AoD are cosidered: - AS: degrees at AoD 5 degrees - AS: 5 degrees at AoD degrees It should be oted that attetio should be paid whe coparig the lik level perforace betwee the two agle spread values sice the BS atea gai for the two correspodig AoDs will be differet. The BS atea gai is applied to the path powers specified i Table 4..

11 Release TR V.. ( Per-path BS agle of departure The Agle of Departure (AoD is defied to be the ea agle with which a arrivig or departig path s power is received or trasitted by the BS array with respect to the boresite. The two values cosidered are: - AoD: 5 degrees (associated with the RMS Agle Spread of degrees - AoD: degrees (associated with the RMS Agle Spread of 5 degrees Per-path BS power aziuth spectru The Power Aziuth Spectru (PAS of a path arrivig at the base statio is assued to have a Laplacia distributio. For a AoD ad RMS agle-spread, the BS per path PAS value at a agle is give by: (,, N exp G( P o where both agles ad are give with respect to the boresight of the atea eleets. It is assued that all atea eleets orietatios are aliged. Also, P is the average received power ad G is the ueric base statio atea gai described i Clause 4.5. by G (.A( Fially, N o is the oralizatio costat: N o exp G( d I the above equatio represets path copoets (sub-rays of the path power arrivig at a AoD. 4.6 Spatial paraeters for the MS 4.6. MS atea patter For each ad every atea eleet at the MS, the atea patter will be assued oi directioal with a atea gai of - dbi Per-path MS agle spread (AS The MS per-path AS is defied as the root ea square (RMS of agles of a icidet path s power at the MS array. Two values of the path s agle spread are cosidered: - AS: 4 degrees (results fro a uifor over 36 degree PAS, - AS: 35 degrees for a Laplacia PAS with a certai path specific Agle of Arrival (AoA Per-path MS agle of arrival The per-path Agle of Arrival (AoA is defied as the ea of agles of a icidet path s power at the UE/Mobile Statio array with respect to the broadside as show Figure 4.5.

12 Release TR V.. (-3 AOA = AOA < AOA > Figure 4.5: Agle of arrival orietatio at the MS. Three differet per-path AoA values at the MS are suggested for the cases of a o-uifor PAS, see Table 4. for details: - AoA: degrees (associated with a RMS Agle Spread of 35 degrees - AoA: degrees (associated with a RMS Agle Spread of 35 degrees - AoA: +.5 degrees (associated with a RMS Agle Spread of 35 degrees or with a LOS copoet Per-path MS power aziuth spectru The Laplacia distributio ad the Uifor distributio are used to odel the per-path Power Aziuth Spectru (PAS at the MS. The Power Aziuth Spectru (PAS of a path arrivig at the MS is odeled as either a Laplacia distributio or a uifor over 36 degree distributio. Sice a oi directioal MS atea gai is assued, the received per-path PAS will reai either Laplacia or uifor. For a icoig AOA ad RMS agle-spread, the MS per-path Laplacia PAS value at a agle is give by: P,, N exp, ( o where both agles ad are give with respect to the boresight of the atea eleets. It is assued that all atea eleets orietatios are aliged. Also, P is the average received power ad N o is the oralizatio costat: N o exp d. I the above equatio represets path copoets (sub-rays of the path power arrivig at a icoig AoA. The distributio of these path copoets is TBD MS directio of travel The obile statio directio of travel is defied with respect to the broadside of the obile atea array as show i Figure 4.6.

13 Release 3 TR V.. (-3 DOT = DOT < DOT > Figure 4.6. Directio of travel for MS Per-path Doppler spectru The per-path Doppler spectru is defied as a fuctio of the directio of travel ad the per-path PAS ad AoA at the MS. This should correspod to the per-path fadig behavior for either the correlatio-based or ray-based ethod. 4.7 Geeratio of chael odel The propoet ca deterie the odel ipleetatio. Exaples of ipleetatios iclude correlatio or ray-based techiques. 4.8 Calibratio ad referece values For the purpose of lik level siulatios, referece values of the average correlatio are give below i Table 4.. The referece values are provided for the calibratio of the siulatio software ad to assist i the resolutio of possible errors i the siulatio ethods ipleeted. Specifically, the average coplex correlatio ad agitude of the coplex correlatio is reported betwee BS ateas ad betwee MS ateas. The spatial paraeter values used are those defied already throughout Clause 4. Table 4.: Referece correlatio values Atea Spacig AS (degrees AOA (degrees Correlatio (agitude Coplex Correlatio BS i i i i i i.49 MS / / i.34 / i / i.34 5 Spatial chael odel for siulatios The spatial chael odel for use i the syste-level siulatios is described i this clause. As i the lik level siulatios, the descriptio is i the cotext of a dowlik syste where the BS trasits to a MS; however the aterial i this clause (with the exceptio of Clause 5.7 o Ioc odellig ca be applied to the uplik as well. As opposed to lik siulatios which siply cosider a sigle BS trasittig to a sigle MS, the syste siulatios typically cosist of ultiple cells/sectors, BSs, ad MSs. Perforace etrics such as throughput ad delay are collected over D drops, where a "drop" is defied as a siulatio ru for a give uber of cells/sectors, BSs, ad MSs, over a specified uber of fraes. Durig a drop, the chael udergoes fast fadig accordig to the otio of the

14 Release 4 TR V.. (-3 MSs. Chael state iforatio is fed back fro the MSs to the BSs, ad the BSs use schedulers to deterie which user(s to trasit to. Typically, over a series of D drops, the cell layout ad locatios of the BSs are fixed, but the locatios of the MSs are radoly varied at the begiig of each drop. To siplify the siulatio oly a subset of BSs will actually be siulated while the reaiig BSs are assued to trasit with full power. The goal of this clause is to defie the ethodology ad paraeters for geeratig the spatial ad teporal chael coefficiets betwee a give base ad obile for use i syste level siulatios. For a S eleet BS array ad a U eleet MS array, the chael coefficiets for oe of N ultipath copoets (ote that these copoets are ot ecessarily tie resolvable, eaig that the tie differece betwee successive paths ay be less tha a chip period are give by a S -by- U atrix of coplex aplitudes. We deote the chael atrix for the th ultipath copoet ( =,,N as H (t. It is a fuctio of tie t because the coplex aplitudes are udergoig fast fadig govered by the oveet of the MS. The overall procedure for geeratig the chael atrices cosists of three basic steps: Specify a eviroet, either suburba acro, urba acro, or urba icro (Clause 5.. Obtai the paraeters to be used i siulatios, associated with that eviroet (Clause Geerate the chael coefficiets based o the paraeters (Clause 5.4. Clauses 5., 5.3, ad 5.4 give the details for the geeral procedure. Figure 5. below provides a roadap for geeratig the chael coefficiets. Clause 5.5 specifies optioal cases that odify the geeral procedure. Clause 5.6 describes the procedure for geeratig correlated log oral user paraeters used i Clause 5.3. Clause 5.7 describes the ethod for accoutig for itercell iterferece. Clause 5.8 presets calibratio results.. Choose sceario Suburba acro Urba acro Urba icro. Deterie user paraeters Agle spread Logoral shadowig Delay spread Pathloss Orietatio Speed Vector BS MS MSv Atea gais AS LN DS AoD,, AoD P AoA,, AoA Agles of departure (paths Agles of departure (subpaths Path delays Average path powers Agles of arrival (paths Agles of arrival (subpaths Far scatterig cluster (urba acro Urba cayo (urba acro 3. Geerate chael coefficiets Polarizatio LOS (urba icro Optios Figure 5.: Chael odel overview for siulatios 5. Geeral defiitios, paraeters, ad assuptios The received sigal at the MS cosists of N tie-delayed ultipath replicas of the trasitted sigal. These N paths are defied by powers ad delays ad are chose radoly accordig to the chael geeratio procedure. Each path cosists of M subpaths. Figure 5. shows the agular paraeters used i the odel. The followig defiitios are used:

15 Release 5 TR V.. (-3 BS BS BS atea array orietatio defied as the differece betwee the broadside of the BS array ad the absolute North (N referece directio. LOS AoD directio betwee the BS ad MS, with respect to the broadside of the BS array. AoD AoD for the th ( = N path with respect to the LOS AoD.,, AoD Offset for the th ( = M subpath of the th path with respect to AoD.,, AoD Absolute AoD for the th ( = M subpath of the th path at the BS with respect to the BS broadside. MS MS MS atea array orietatio defied as the differece betwee the broadside of the MS array ad the absolute North referece directio. Agle betwee the BS-MS LOS ad the MS broadside. AoA AoA for the th ( = N path with respect to the LOS AoA, MS.,, AoA Offset for the th ( = M subpath of the th path with respect to AoA.,, AoA Absolute AoA for the th ( = M subpath of the th path at the MS with respect to the BS broadside. v MS velocity vector. v Agle of the velocity vector with respect to the MS broadside: v =arg(v. The agles show i Figure 5. that are easured i a clockwise directio are assued to be egative i value. BS array Cluster Subpath, AoD,, AoA AoA, AoA MS N v v N AoD BS MS,, AoD MS array broadside MS array BS BS array broadside MS directio of travel Figure 5.: BS ad MS agle paraeters For syste level siulatio purposes, the fast fadig per-path will be evolved i tie, although bulk paraeters icludig agle spread, delay spread, log oral shadowig, ad MS locatio will reai fixed durig the its evaluatio durig a drop. The followig are geeral assuptios ade for all siulatios, idepedet of eviroet: a Uplik-Dowlik Reciprocity: The AoD/AoA values are idetical betwee the uplik ad dowlik. b For FDD systes, rado subpath phases betwee UL, DL are ucorrelated. (For TDD systes, the phases will be fully correlated. c Shadowig aog differet obiles is ucorrelated. I practice, this assuptio would ot hold if obiles are very close to each other, but we ake this assuptio just to siplify the odel. d The spatial chael odel should allow ay type of atea cofiguratio (e.g. whose size is saller tha the shadowig coherece distace to be selected, although details of a give cofiguratio ust be shared to allow

16 Release 6 TR V.. (-3 others to reproduce the odel ad verify the results. It is iteded that the spatial chael odel be capable of operatig o ay give atea array cofiguratio. I order to copare algoriths, referece atea cofiguratios based o uifor liear array cofiguratios with.5, 4, ad wavelegth iter-eleet spacig will be used. e The coposite AS, DS, ad SF shadow fadig, which ay be correlated paraeters depedig o the chael sceario, are applied to all the sectors or ateas of a give base. Sub-path phases are rado betwee sectors. The AS is coposed of 6 x sub-paths, ad each has a precise agle of departure which correspods to a atea gai fro each BS atea. The effect of the ateas gai ay cause soe chage to the chael odel i both AS ad DS betwee differet base ateas, but this is separate fro the chael odel. The SF is a bulk paraeter ad is coo aog all the BS ateas or sectors. f The elevatio spread is ot odeled. g To allow coparisos of differet atea scearios, the trasit power of a sigle atea case shall be the sae as the total trasit power of a ultiple atea case. h The geeratio of the chael coefficiets (Clause 5.4 assues liear arrays. The procedure ca be geeralized for other array cofiguratios, but these odificatios are left for the propoet. 5. Eviroets We cosider the followig three eviroets. a Suburba acrocell (approxiately 3K distace BS to BS b Urba acrocell (approxiately 3K distace BS to BS c Urba icrocell (less tha K distace BS to BS The characteristics of the acro cell eviroets assue that BS ateas are above rooftop height. For the urba icrocell sceario, we assue the BS atea is at rooftop height. Table 5. describes the paraeters used i each of the eviroets.

17 Release 7 TR V.. (-3 Table 5.. Eviroet paraeters Chael Sceario Suburba Macro Urba Macro Urba Micro Nuber of paths (N Nuber of sub-paths (M per-path Mea AS at BS E( AS =5 E( AS =8, 5 NLOS: E( AS =9 AS at BS as a logoral RV AS ^ AS x AS, x ~ (, AS =.69 8 AS =.3 AS =.8 AS =.34 5 AS =.8 AS =. r /..3 N/A AS AoD AS Per-path AS at BS (Fixed deg deg 5 deg (LOS ad NLOS BS per-path AoD Distributio stadard distributio (, AoD AoD r AS where AS (, AoD AoD r AS where AS N/A U(-4deg, 4deg Mea AS at MS E( AS, MS=68 E( AS, MS=68 E( AS, MS=68 Per-path AS at MS (fixed MS Per-path AoA Distributio (, AoA (Pr (, AoA (Pr (, AoA (Pr Delay spread as a logoral RV DS ^ DS x DS, x ~ (, DS = DS =.88 DS = -6.8 DS =.8 Mea total RMS Delay Spread E( DS =.7 s E( DS =.65 s E( DS =.5 s (output r /.4.7 N/A DS delays DS Distributio for path delays U(,. s Logoral shadowig stadard 8dB 8dB NLOS: db deviatio LOS: 4dB SF Pathloss odel (db, d is i eters N/A log (d log (d NLOS: log (d LOS: *log (d The followig are assuptios ade for the suburba acrocell ad urba acrocell eviroets. a The acrocell pathloss is based o the odified COST3 Hata urba propagatio odel: d PL[ db] log hbs log ( h log ( f 3.8 log ( h.7h C s c bs s where hbs is the BS atea height i eters, h s the MS atea height i eters, f c the carrier frequecy i MHz, d is the distace betwee the BS ad MS i eters, ad C is a costat factor (C = db for suburba acro ad C = 3dB for urba acro. Settig these paraeters to h bs = 3, h s=.5, ad f c =9MHz, the pathlosses for suburba ad urba acro eviroets becoe, respectively, PL log ( d ad PL log ( d. The distace d is required to be at least 35. b Atea patters at the BS are the sae as those used i the lik siulatios give i Clause c Site-to-site SF correlatio is. 5. This paraeter is used i Clause d The hexagoal cell repeats will be the assued layout. The followig are assuptios ade for the icrocell eviroet. a The icrocell NLOS pathloss is based o the COST 3 Walfish-Ikegai NLOS odel with the followig paraeters: BS atea height.5, buildig height, buildig to buildig distace 5, street width 5, MS atea height.5, orietatio 3deg for all paths, ad selectio of etropolita ceter. With these paraeters, the equatio siplifies to:

18 Release 8 TR V.. (-3 PL(dB = *log(d + ( *f c /95*log(f c. The resultig pathloss at 9 MHz is: PL(dB = *log(d, where d is i eters. The distace d is at least. A bulk log oral shadowig applyig to all sub-paths has a stadard deviatio of db. The icrocell LOS pathloss is based o the COST 3 Walfish-Ikegai street cayo odel with the sae paraeters as i the NLOS case. The pathloss is PL(dB = *log(d + *log(f c The resultig pathloss at 9 MHz is PL(dB = *log(d, where d is i eters. The distace d is at least. A bulk log oral shadowig applyig to all sub-paths has a stadard deviatio of 4dB. b Atea patters at the BS are the sae as those used i the lik siulatios give i Clause c Site-to-site correlatio is. 5. This paraeter is used i Clause d The hexagoal cell repeats will be the assued layout. Note that the SCM odel described here with N = 6 paths ay ot be suitable for systes with badwidth higher tha 5MHz. 5.3 Geeratig user paraeters For a give sceario ad set of paraeters give by a colu of Table 5. Eviroet paraeters, realizatios of each user's paraeters such as the path delays, powers, ad sub-path agles of departure ad arrival ca be derived usig the procedure described here i Clause 5.3. I particular, Clause 5.3. gives the steps for the urba acrocell ad suburba acrocell eviroets, ad Clause 5.3. gives the steps for the urba icrocell eviroets Geeratig user paraeters for urba acrocell ad suburba acrocell eviroets Step : Choose either a urba acrocell or suburba acrocell eviroet. Step : Deterie various distace ad orietatio paraeters. The placeet of the MS with respect to each BS is to be deteried accordig to the cell layout. Fro this placeet, the distace betwee the MS ad the BS (d ad the LOS directios with respect to the BS ad MS ( BS ad MS, respectively ca be deteried. Calculate the bulk path loss associated with the BS to MS distace. The MS atea array orietatios ( MS, are i.i.d., draw fro a uifor to 36 degree distributio. The MS velocity vector v has a agitude v draw accordig to a velocity distributio (to be deteried ad directio v draw fro a uifor to 36 degree distributio. Step 3: Deterie the DS, AS, ad SF. These variables, give respectively by DS, AS, ad SF, are geerated as described i Clause 5.6 below. Note that ^( DS is i uits of secods so that the arrowbad coposite delay spread DS is i uits of secods. Note also that we have dropped the BS idicies used i Clause 5.6. to siplify otatio. Step 4: Deterie rado delays for each of the N ultipath copoets. For acrocell eviroets, N = 6 as give i Table 5.. Geerate rado variables ' ',..., N accordig to z r l z =,,N ' DS DS where ( =,,N are i.i.d. rado variables with uifor distributio U(,, rds is give i Table 5., ad ' ' ' DS ( N ( N... ( is derived i Step 3 above. These variables are ordered so that ad the iiu of these is subtracted fro all. The delay for the th path is the value of earest /6th chip iterval: ' ( ' ( are quatized i tie to the

19 Release 9 TR V.. (-3 ' Tc ( ( floor.5,,..., N, 6 T 6 c ' where floor(x is the iteger part of x, ad Tc is the chip iterval (Tc = /3.84x6 sec for ad Tc =... /.88x6 sec for Note that these delays are ordered so that N 5. (See otes ad at the ed of Clause Quatizatio to /6 chip is the default value. For special purpose ipleetatios, possibly higher quatizatio values ay be used if eeded. Step 5: Deterie rado average powers for each of the N ultipath copoets. Let the uoralized powers be give by P e ( rds ( ( rds DS ( /, =,,6 where ( =,,6 are i.i.d. Gaussia rado variables with stadard deviatio RND = 3 db, which is a shadowig radoizatio effect o the per-path powers. Note that the powers are deteried usig the uquatized chael delays. Average powers are oralized so that the total average power for all six paths is equal to oe: (See ote 3 at the ed of Clause P ' P. 6 P Step 6: Deterie AoDs for each of the N ultipath copoets. First geerate i.i.d. zero-ea Gaussia rado variables: j ' j ' ~ (, AoD, =,, N, where AoD r AS AS. The value ras is give i Table 5. ad depeds o whether the urba or suburba degrees. They are ordered i icreasig absolute value so that N are assiged to the ordered variables so that ' ' ' ( (... ( N. The AoDs AoD, =,, ' AoD (, =,,N. (See ote 4 at the ed of Clause Step 7: Associate the ultipath delays with AoDs. The th delay geerated i Step 3 is associated with the th AoD AoD geerated i Step 6. Step 8: Deterie the powers, phases ad offset AoDs of the M = sub-paths for each of the N paths at the BS. All sub-path associated with the th path have idetical powers ( P / where P is fro Step 5 ad i.i.d phases draw fro a uifor to 36 degree distributio. The relative offset of the th subpath ( =,, M,, AoD is a fixed value give i Table5.. For exaple, for the urba ad suburba acrocell cases, the offsets for the first ad secod sub-paths are respectively,, AoD =.894 ad,, AoD = degrees. These offsets are chose to result i the desired per-path agle spread ( degrees for the acrocell eviroets, ad 5 degrees for the icrocell eviroet. The per-path agle spread of the th path ( = N is i cotrast to the agle spread which refers to the coposite sigal with N paths. Step 9: Deterie the AoAs for each of the ultipath copoets. The AoAs are i.i.d. Gaussia rado variables AoA ~ (, AoA, =,, N, where AoA = 4. -exp -.75 log ( P ad P is the relative power of the th path fro Step 5. (See ote 5 at the ed of Clause 5.3.

20 Release TR V.. (-3 Step : Deterie the offset AoAs at the UE of the M = sub-paths for each of the N paths at the MS. As i Step 8 for the AoD offsets, the relative offset of the th subpath ( =,, M,, AoA is a fixed value give i Table 5.. These offsets are chose to result i the desired per-path agle spread of 35 degrees. Step : Associate the BS ad MS paths ad sub-paths. The th BS path (defied by its delay, power P, ad AoD is associated with the th MS path (defied by its AoA. For the th path pair, radoly pair each AoD AoA of the M BS sub-paths (defied by its offset,, AoD with a MS sub-path (defied by its offset,, AoA. Each sub-path pair is cobied so that the phases defied by i step 8 are aitaied. To siplify the otatio we reuber the M MS sub-path offsets with their ewly associated BS sub-path. I other words, if the first ( = BS sub-path is radoly paired with the th ( = MS sub-path, we re-associate,, AoA (after pairig with,, AoA (before pairig. Step : Deterie the atea gais of the BS ad MS sub-paths as a fuctio of their respective sub-path AoDs ad AoAs. For the th path, the AoD of the th sub-path (with respect to the BS atea array broadside is,, AoD BS AoD, AoD. Siilarly, the AoA of the th sub-path for the th path (with respect to the MS atea array broadside is,, AoA MS AoA, AoA. The atea gais are depedet o these sub-path AoDs ad AoAs. For the BS ad MS, these are give respectively as G ad G. BS (, AoD MS (, AoA Step 3: Apply the path loss based o the BS to MS distace fro Step, ad the log oral shadow fadig deteried i step 3 as bulk paraeters to each of the sub-path powers of the chael odel. Notes: Note : I the developet of the Spatial Chael Model, care was take to iclude the statistical relatioships betwee r DS delays / DS Agles ad Powers, as well as Delays ad Powers. This was doe usig the proportioality factors r ad AS AoD / PAS that were based o easureets. Note : While there is soe evidece that delay spread ay deped o distace betwee the trasitter ad receiver, the effect is cosidered to be ior (copared to other depedecies: DS-AS, DS-SF.. Various iputs based o ultiple data sets idicate that the tred of DS ca be either slightly positive or egative, ad ay soeties be relatively flat with distace. For these reasos ad also for siplicity, a distace depedece o DS is ot odeled. Note 3: The equatios preseted here for the power of the th path are based o a power-delay evelope which is the average behavior of the power-delay profile. Defiig the powers to reproduce the average behavior liits the dyaic rage of the result ad does ot reproduce the expected radoess fro trial to trial. The radoizig process is used to vary the powers with respect to the average evelope to reproduce the variatios experieced i the actual chael. This paraeter is also ecessary to produce a dyaic rage coparable to easureets. Note 4: The quatity r AS describes the distributio of powers i agle ad r AS AoD / PAS, i.e. the spread of agles to the power weighted agle spread. Higher values of r AS correspod to ore power beig cocetrated i a sall AoD or a sall uber of paths that are closely spaced i agle. Note 5: Although two differet echaiss are used to select the AoD fro the Base, ad the AoA at the MS, the paths are sufficietly defied by their BS to MS coectio power P ad delay, thus there is o abiguity i associatig the paths to these paraeters at the BS or MS Geeratig user paraeters for urba icrocell eviroets Urba icrocell eviroets differ fro the acrocell eviroets i that the idividual ultipaths are idepedetly shadowed. As i the acrocell case, N = 6 paths are odeled. We list the etire procedure but oly describe the details of the steps that differ fro the correspodig step of the acrocell procedure.

21 Release TR V.. (-3 Step : Choose the urba icrocell eviroet. Step : Deterie various distace ad orietatio paraeters. Step 3: Deterie the bulk path loss ad log oral shadow fadig paraeters. Step 4: Deterie the rado delays for each of the N ultipath copoets. For the icrocell eviroet, N = 6. The delays =,, N are i.i.d. rado variables draw fro a uifor distributio fro to. s. Step 5: The iiu of these delays is subtracted fro all so that the first delay is zero. The delays are quatized i tie to the earest /6th chip iterval as described i Clause Whe the LOS odel is used, the delay of the direct copoet will be set equal to the first arrivig path with zero delay. Step 6: Deterie rado average powers for each of the N ultipath copoets. The PDP cosists of N=6 distict paths that are uiforly distributed betwee ad. s. The powers for each path are expoetially decayig i tie with the additio of a logoral radoess, which is idepedet of the path delay: P ' ( z / where is the uquatized values ad give i uits of icrosecods, ad z ( =,,N are i.i.d. zero ea Gaussia rado variables with a stadard deviatio of 3dB. Average powers are oralized so that total average power for all six paths is equal to oe: P ' P. 6 P j Whe the LOS odel is used, the oralizatio of the path powers icludes cosideratio of P the power of the direct copoet D such that the ratio of powers i the direct path to the scattered paths has a ratio of K: ' j P P K P D. K ' K 6 P, ' j j Step 7: Deterie AoDs for each of the N ultipath copoets. The AoDs (with respect to the LOS directio are i.i.d. rado variables draw fro a uifor distributio over 4 to +4 degrees: AoD ~ U ( 4, 4, =,, N, Associate the AoD of the th path with the power of the th path P AoD. Note ulike the acrocell eviroet, the AoDs do ot eed to be sorted before beig assiged to a path power. Whe the LOS odel is used, the AoD for the direct copoet is set equal to the lie-of-sight path directio. Step 8: Radoly associate the ultipath delays with AoDs. Step 9: Deterie the powers, phases, ad offset AoDs of the M = sub-paths for each of the N paths at the BS. The offsets are give i Table 5., ad the resultig per-path AS is 5 degrees istead of degrees for the acrocell case. The direct copoet, used with the LOS odel will have o per-path AS. Step : Deterie the AoAs for each of the ultipath copoets. The AoAs are i.i.d Gaussia rado variables AoA ~ (, AoA, =,, N, where AoA = 4. -exp -.65 log ( P ad P is the relative power of the th path fro Step 5. Whe the LOS odel is used, the AoA for the direct copoet is set equal to the lie-of-sight path directio. Step : Deterie the offset AoAs of the M = sub-paths for each of the N paths at the MS.

22 Release TR V.. (-3 Step : Associate the BS ad MS paths ad sub-paths. Sub-paths are radoly paired for each path, ad the sub-path phases defied at the BS ad MS are aitaied. Step 3: Deterie the atea gais of the BS ad MS sub-paths as a fuctio of their respective sub-path AoDs ad AoAs. Step 4: Apply the path loss based o the BS to MS distace ad the log oral shadow fadig deteried i Step 3 as bulk paraeters to each of the sub-path powers of the chael odel. Sub-path # ( Table 5.: Sub-path AoD ad AoA offsets Offset for a deg AS at BS (Macrocell,, AoD (degrees Offset for a 5 deg AS at BS (Microcell,, AoD (degrees Offset for a 35 deg AS at MS,, AoA (degrees, , , , , , , , , , The values i Table 5. are selected to produce a biased stadard deviatio equal to, 5, ad 35 degrees, which is equivalet to the per-path power weighted aziuth spread for equal power sub-paths. 5.4 Geeratig chael coefficiets Give the user paraeters geerated i Clause 5.3, we use the to geerate the chael coefficiets. For a S eleet liear BS array ad a U eleet liear MS array, the chael coefficiets for oe of N ultipath copoets are give by a U -by- S atrix of coplex aplitudes. We deote the chael atrix for the th ultipath copoet ( =,,N as H (t. The (u,sth copoet (s =,,S; u =,,U of H (t is give by h u, s, ( t P M SF M G G BS MS, AoD, AoA exp jk v cos exp j kd si exp jkd u s, AoA v si t, AoD, AoA where P is the power of the th path (Step 5. M SF is the logoral shadow fadig (Step 3, applied as a bulk paraeter to the paths for a give drop. is the uber of subpaths per-path.,, AoD is the the AoD for the th subpath of the th path (Step.,, AoA is the the AoA for the th subpath of the th path (Step. G BS (, AoD is the BS atea gai of each array eleet (Step. G MS (, AoA is the MS atea gai of each array eleet (Step. j is the square root of -. k is the wave uber / where is the carrier wavelegth i eters. d s is the distace i eters fro BS atea eleet s fro the referece (s = atea. For the d referece atea s =, =.

23 Release 3 TR V.. (-3 d u v is the distace i eters fro MS atea eleet u fro the referece (u = atea. For the d referece atea u =, =. is the phase of the th subpath of the th path (Step 8. is the agitude of the MS velocity vector (Step. v is the agle of the MS velocity vector (Step. The path loss ad the log oral shadowig is applied as bulk paraeters to each of the sub-path copoets of the path copoets of the chael. 5.5 Optioal syste siulatio features 5.5. Polarized arrays Practical ateas o hadheld devices require spacigs uch less tha /. Polarized ateas are likely to be the priary way to ipleet ultiple ateas. A cross-polarized odel is therefore icluded here. A ethod of describig polarized ateas is preseted, which is copatible with the 3 step procedure give i Clause 5.3. The followig steps replace step 3 with the ew steps 3-9 to accout for the additioal polarized copoets. The (S/-eleet BS arrays ad (U/-eleet MS arrays cosist of U ad S (i.e., twice i uberateas i the V, H, or off-axis polarizatio. Step 3: Geerate additioal cross-polarized subpaths. For each of the 6 paths of Step 4, geerate a additioal M subpaths at the MS ad M subpaths at the BS to represet the portio of each sigal that leaks ito the crosspolarized atea orietatio due to scatterig. Step 4:. Set subpath AoDs ad AoAs. Set the AoD ad AoA of each subpath i Step 3 equal to that of the correspodig subpath of the co-polarized atea orietatio. (Orthogoal sub-rays arrive/depart at coo agles. Step 5: Geerate phase offsets for the cross-polarized eleets. We defie to be the phase offset of the th subpath of the th path betwee the x copoet (e.g. either the horizotal h or vertical v of the BS eleet ad ( x, y, the y copoet (e.g. either the horizotal h or vertical v of the MS eleet. Set ( xx,, to be geerated i ( xy, ( yx, ( yy, Step 8 of Clause 5.3. Geerate,,,, ad, as i.i.d rado variables draw fro a uifor to 36 degree distributio. (x ad y ca alteratively represet the co-polarized ad cross-polarized orietatios. Step 6: Decopose each of the co-polarized ad cross-polarized sub-rays ito vertical ad horizotal copoets based o the co-polarized ad cross-polarized orietatios. Step 7: The coupled power P of each sub-path i the horizotal orietatio is set relative to the power P of each sub-path i the vertical orietatio accordig to a XPD ratio, defied as XPD= P/P. A sigle XPD ratio applies to all sub-paths of a give path. Each path experieces a idepedet realizatio of the XPD. For each path the realizatio of the XPD is draw fro the distributios below. For urba acrocells: P = P - A - B* (,, where A=.34*(ea relative path power i db+7. db, ad B=5.5dB is the stadard deviatio of the XPD variatio. For urba icrocells: P = P - A - B* variatio. (,, where A=8 db, ad B=8dB is the stadard deviatio of the XPD The value (, is a zero ea Gaussia rado uber with uit variace ad is held costat for all sub-paths of a give path. By syetry, the coupled power of the opposite process (horizotal to vertical is the sae. The V-to-H XPD draws are idepedet of the H-to-V draws.

24 Release 4 TR V.. (-3 Step 8: At the receive ateas, decopose each of the vertical ad horizotal copoets ito copoets that are co-polarized with the receive ateas ad su the copoets. This procedure is perfored withi the chael coefficiet expressio give below. Step 9: Apply the path loss based o the BS to MS distace fro Step, ad the log oral shadow fadig deteried i step 3 as bulk paraeters to each of the sub-path powers of the chael odel. The fadig behavior betwee the cross pol eleets will be a fuctio of the per-ray spreads ad the Doppler. The fadig betwee orthogoal polarizatios has bee observed to be idepedet ad therefore the sub-rays phases are chose radoly. The propagatio characteristics of V-to-V paths are assued to be equivalet to the propagatio characteristics of H-to-H paths. The polarizatio odel ca be illustrated by a atrix describig the propagatio of ad ixig betwee horizotal ad vertical aplitude of each sub-path. The resultig chael realizatio is: ( exp j r exp j ( ( v T ( v, v ( v, h ( v BS, AoD MS, AoA P SF M ( h ( h, v ( h, h ( h u, s, ( BS, AoD exp exp MS, AoA h t ( r j j ( M exp jkd si( exp jkd si( exp jk v cos( θ t s, AoD u, AoA, AoA v where: ( v BS (, AoD is the BS atea coplex respose for the V-pol copoet. ( h BS (, AoD is the BS atea coplex respose for the H-pol copoet. ( v MS (, AoA is the MS atea coplex respose for the V-pol copoet. ( h MS (, AoA is the MS atea coplex respose for the H-pol copoet. r (. (. r ( xy,, is the atea gai is the rado variable represetig the power ratio of waves of the th path leavig the BS i the vertical directio ad arrivig at the MS i the horizotal directio (v-h to those leavig i the vertical directio ad arrivig i the vertical directio (v-v. is the rado variable represetig the power ratio of waves of the th path leavig the BS i the horizotal directio ad arrivig at the MS i the vertical directio (h-v to those leavig i the vertical directio ad arrivig i the vertical directio (v-v. The variables ad are i.i.d. phase offset of the th subpath of the th path betwee the x copoet (either the horizotal h or vertical v of the BS eleet ad the y copoet (either the horizotal h or vertical v of the MS eleet. r r The other variables are described i Clause 5.4. Note that the descriptio above correspods to a trasissio fro the base to obile. The appropriate subscript ad superscript chages should be ade for uplik trasissios. The x atrix represets the scatterig phases ad aplitudes of a plae wave leavig the UE with a give agle ad polarizatio ad arrivig Node B with aother directio ad polarizatio. r is the average power ratio of waves leavig the UE i the vertical directio ad arrivig at Node B i the horizotal directio (v-h to those arrivig at Node B i the vertical directio (v-v. By syetry the power ratio of the opposite process (h-v over v-v is chose to be the sae. Note that: r =/XPD; for the acrocell odel, the XPD is depedet o the path idex; for the icrocell odel, the XPD is idepedet of path idex. Expressio ( assues a rado pairig of the of the sub-paths fro the MS ad BS. The rado orietatio of the MS (UE array affects the value of the agle,, AoA of each sub-path.

25 Release 5 TR V.. (-3 If for exaple, oly vertically polarized ateas are used at both NodeB ad UE the the atea resposes becoe ad expressio ( becoes idetical to (. For a ideal dipole atea at the NodeB tilted with respect to the z- axis at degrees the above vector becoes The elevatio spectru is ot odeled Far scatterer clusters si( cos( cos(, AoA The Far scatterer cluster odel is switch selectable. It represets the bad-urba case where additioal clusters are see i the eviroet. This odel is liited to use with the urba acro-cell where the first cluster will be the priary cluster ad the secod will be the far scatterig cluster (FSC. Whe the odel is active, it will have the followig characteristics: There is a reductio i the uber of paths i the priary cluster fro N = 6 to N = 4, with the far scatterig cluster the havig N =. Thus the total uber of paths will stay the sae, ow N = N + N = 4 +. This is a odificatio to the SCM chael geeratio procedure i Clause 5.3. The FSCs will oly be odeled for the servig cell, with 3 idepedet FSCs i the cell uiforly applied to the area of the cell outside the iiu radius. The followig is a geeratio procedure for the FSC odel ad is used i cojuctio with the oral chael geeratio procedure i Clause 5.3. Step. Drop the MS withi test cell. Step. Drop three FS clusters uiforly across the cell hexago with a iiu radius of R = 5. Step 3. Choose the FS cluster to use for the obile that is closest to the obile. Step 4. Geerate 6 delays,,,,., Step 5. Sort,, 3, 4 i order of icreasig delays, Step 6. Subtract shortest delay of,, 3, 4 fro each of,, 3, 4 Step 7. Sort 5, 6 i the order of icreasig delays, Step 8. Subtract shorter delay of 5, 6 fro each of 5, 6 Step 9. Assig powers to paths correspodig to all 6 delays:. P e ( rds r DS DS /, =,,6 where ( =,,6 are i.i.d. Gaussia rado variables with stadard deviatio RND = 3 db. Step. Calculate excess path delay excess ad add to 5, 6 Step. Atteuate P5 ad P6 by db/ s of excess delay with a db axiu atteuatio. The excess delay will be defied as the differece i propagatio tie betwee the BS-MS LOS distace, ad the BS-FSC-MS distace. Step. Scale each of the powers i the ai cluster ad i the FSC by a coo log oral radoizig factor of 8dB/ draw oce per cluster to represet the effect of the idepedet per cluster shadow fadig after icludig site correlatio of the obile locatio. Step 3. Noralize powers of the 6 paths to uity power. Step 4. Select AoDs at the BS for the ai cluster fro the chael geeratio procedure i Clause 5.3. Select AoDs at the BS for the FSC refereced to the directio of the FSC ad selectig values fro a Gaussia distributio with

26 Release 6 TR V.. (-3 stadard deviatio r AS *ea(as, where the ea(as is equal to 8 degrees or 5 degrees, chose to atch the agle spread odel used. Step 5. Select AoAs at the MS usig the equatio i step 9 of Clause 5.3, where the 4 paths associated with the ai cluster are refereced to the LOS directio to the BS, ad the paths associated with the FSC are refereced to the directio of the FSC. The relative powers used are the oralized powers of step 3 of this clause. Step 6. Apply the bulk path loss ad log oral to all 6 paths ad apply the atea gais accoutig for the agles of the sub-paths associated with the ai cluster ad the FSC Lie of sight The lie-of-sight (LOS odel is a optio that is switch selectable for the urba icro sceario oly. LOS odellig will ot be defied for the suburba or urba acro cases due to the low probability of occurrece. The LOS odellig is based o the Ricea K factor defied as the ratio of power i the LOS copoet to the total power i the diffused o lie-of-sight (NLOS copoet. The K factor ay be defied i db or liear uits i the equatios i this docuet. The LOS copoet is a direct copoet ad therefore has o per-path agle spreadig. The LOS odel uses the followig descriptio whe this fuctio is selected. For the NLOS case, the Ricea K factor is set to, thus the fadig is deteried by the cobiatio of sub-rays as described i Clause 5.3 of the odel. For the LOS case, the Ricea K factor is based o a siplified versio of [] : K = 3-.3*d (db where d is the distace betwee MS ad BS i eters. The probability for LOS or NLOS depeds o various eviroetal factors, icludig clutter, street cayos, ad distace. For siplicity, the probability of LOS is defied to be uity at zero distace, ad decreases liearly util a cutoff poit at d=3, where the LOS probability is zero. P( LOS (3 d / 3, d 3, d 3 The K-factor, propagatio slope, ad shadow fadig stadard deviatio will all be chose based o the results of selectig the path to be LOS or NLOS. The total cobied power of the LOS copoet ad the 6 diffuse copoets is oralized to uity power so the coheret LOS copoet will have a relative aplitude K K ad the aplitudes of all 6 diffuse paths will have a relative aplitude, where K is i liear uits. The LOS path will coicide i tie (but ot ecessarily i agle K with the first (earliest diffused path. Whe pairig sub-rays betwee trasitter ad receiver, the direct copoets are paired represetig the LOS path. The AoD ad AoA of the LOS copoet are give by BS ad MS, respectively. Followig the defiitios i Clause 5.4, the (u,sth copoet of the chael coefficiet (s =,,S; u =,,U ad path is give by: h LOS s, u, ( t hs, u, ( t K SF K K G G BS MS ( ( BS MS exp exp exp jk v cos( MS jkd s jkd u si( v si( t BS MS LOS LOS ad for : h s u, ( t hs, u, K (, t where hs, u, ( t;,..., N is as defied i Clause 5.4 ad: BS MS LOS is the the AoD for the LOS copoet at the base is the the AoA for the LOS copoet at the terial is the phase of the LOS copoet