IEEE C802.20-04-79
2004-11-15
Project
IEEE 802.20 Working Group on Mobile Broadband Wireless Access
<http://grouper.ieee.org/groups/802/20/>
Title
Overview of the Spatial Channel Model developed in 3GPP-3GPP2
Date
Submitted
2004-11-15
Source(s)
Achilles Kogiantis
Rm. 1A-251, 67 Whippany Rd
Whippany, NJ 07981
Re:
MBWA Call for Contributions
Abstract
This contribution provides a detailed overview of the spatial channel model that was developed jointly in the
3GPP and 3GPP2 standard bodies for link and system level simulations
For Discussion
Purpose
Page 1
Voice: (973) 386-4399
Fax: (973) 386-2651
Email: achilles@lucent.com
Notice
This document has been prepared to assist the IEEE 802.20 Working Group. It is offered as a basis for discussion and is not binding on
the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further
study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein.
Release
The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any
modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication
even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in
part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public
by IEEE 802.20.
Patent Policy
The contributor is familiar with IEEE patent policy, as outlined in Section 6.3 of the IEEE-SA Standards Board Operations Manual
<http://standards.ieee.org/guides/opman/sect6.html#6.3> and in Understanding Patent Issues During IEEE Standards Development
<http://standards.ieee.org/board/pat/guide.html>.
IEEE C802.20-04-79
Overview of the Joint 3GPP 3GPP2 Spatial
Channel Model Recommendation
Achilles Kogiantis
802.20 Session 11
San Antonio, TX
November 17, 2004
Page 2
SCM standards activity
IEEE C802.20-04-79
• A joint adhoc was created by 3GPP and 3GPP2 to specify spatial
channel models (April 2002)
– The adhoc group completed its task in June 2003.
» Recommendation to 3GPP RAN1 was adopted as permanent document TR
25.996
» Recommendation to 3GPP2 WG3 made in June 2003
– SCM Participating Companies:
Atelier Telecom
DISA
Elektrobit
Ericsson
ETRI
France Telecom
Infineon
Interdigital
IP Wireless
LG Electronics
Lucent Technologies
MERL
Page 3
Mitsubishi
Motorola
Nokia
Nortel Networks
Panasonic
Samsung
Spirent Communications
Telia
Texas Instruments
TTPCom
Qualcomm
SCM Approach
IEEE C802.20-04-79
• The SCM AHG work was intended to be an extension to the available
evaluation methodology for multiple antenna studies
• The add-on features of the SCM recommendations are:
– System Level channel modeling are the core specifications (link level
definitions are also provided for calibration)
– The wideband model addresses both 1.25MHz and 5MHz channel bandwidths
– Modeling is developed as a general framework for multiple antenna transmit
and/or receive configurations
– No specific antenna topologies are enforced. Specifications are independent
of antenna arrangements
– Specifications reflect on numerous measurement campaigns and literature
surveys (also COST 259 recommendations)
– The SCM AHG sought a balance between the representation of realistic spatial
environments and modeling complexity.
Page 4
SCM Structure
IEEE C802.20-04-79
• System Level Model
– Each drop is assigned spatial-temporal parameters from pre-determined
distributions
• Evaluation Methodology specifications
– Define the methods to be used for utilizing the SCM in system level
simulations.
– Objective is not to deviate from currently established methodologies
• Link Level Model
– Contains a specific set of fixed spatial-temporal parameters
– Used for calibration purposes only
Page 5
System Level SCM –
Channel Scenarios
IEEE C802.20-04-79
• Choose a Channel Scenario common to all drops. Each channel scenario
defines a specific set of typical physical parameter values.
– Suburban Macro
» BS-BS distance aprox. 3Km. BS antenna position: High
» Describes small angle spread (AS), small delay spread (DS) environments: rural areas
– Urban Macro
» BS-BS distance aprox. 3Km. BS antenna position: High
» Describes moderate to high AS, large DS environments: large cells with urban buildings
in the vicinity; significant scattering
– Urban Micro
» BS-BS distance aprox. 1Km
» Describes large AS, moderate DS environments. BS antenna located at rooftop. Small
urban cells with a wide range of per drop AS,DS values
– Additional modeling options:
» Line of Sight (LOS). Applicable to urban micro
» Far Scatterer Cluster. Applicable to urban macro. Models the bad urban case.
» Urban Canyon. Applicable to urban macro
Page 6
System channel model overview
1. Choose scenario
Suburban
macro
Urban
macro
IEEE C802.20-04-79
Urban
micro
2. Determine user parameters
Angle spread  AS
Lognormal shadowing LN
n,AoD Angles of departure (paths)
n ,m,AoD Angles of departure (subpaths)
Delay spread DS
Pathloss
Orientation, Speed Vector
BS MS  MS v
n Path delays
Pn Average path powers
n,AoA Angles of arrival (paths)
n ,m ,AoA Angles of arrival (subpaths)
Far scattering cluster
(urban macro)
Urban canyon
(urban macro)
Antenna gains
3. Generate channel coefficients
Page 7
Polarization
LOS (urban micro)
Options
Pathloss - Shadowing
IEEE C802.20-04-79
Channel Scenario
Suburban Macro
Urban Macro
Pathloss model (dB)
(d in meters)
31.5 + 35log10(d)
34.5 + 35log10(d)
h BS  32m
hMS  1.5m
Lognormal shadowing
st.dev (  SF )
h BS  32m
hMS  1.5m
BS-MS distance 35m
BS-MS distance 35m
8dB
8dB
• Macro channels follow the Hata COST 231
• Urban micro follows the Walfish-Ikegami pathloss model
Page 8
Urban Micro
NLOS: 34.53 + 38log10(d)
LOS: 30.18 +26log10(d)
h BS  12.5m
hMS  1.5m
Bldg-bldg: 50m
Street Width: 25m
NLOS: 10dB
LOS: 4dB
Correlation of Narrowband
Parameters
IEEE C802.20-04-79
• Site to Site Correlation = 0.5
• Correlation between Angle Spread, Delay Spread, Shadowing (LN) :
  Correlatio n between DS & AS   0.5
c11 c12
c
 21 c 22
c 31 c 32
  Correlatio n between LN & AS   0.6
  Correlatio n between LN & DS   0.6
• Define 1 ,  2 ,  3 and
index (n = 1 … N).
wn1 , wn 2 , wn3
Page 9



 

 
 
as iid Gaussian, N(0,1) with BS
n  c11 c12 c13  wn1 
    c
 w  
c
c
23   n 2 
 n   21 22
  n  c 31 c 32 c 33  wn 3 
Intra-site correlations
c13  

c 23    
c 33   
0 0

0 0
0 0
0   1 

0  2 
  3 
Site to site correlation
1
2
Generation of Narrowband
Parameters
IEEE C802.20-04-79
• For a given channel scenario draw random lognormal realizations for
AS, DS, LN (  AS , n ,  DS , n ,  LN , n ) in an MS drop and for each BS n :
AS,n  10 ^  ASn  AS 
DS,n  10 ^  DS n  DS 
LN ,n  10 ^  SF n /10
• The distributions are a priori determined.
• Each macro channel scenario is characterized by a unique set of mean,
variance values:
Page 10




2

 E  log10 (AS ,n )   2AS


AS  E  log10(AS )
AS
DS  E  log10 (DS )
2

DS  E  log10 (DS ,n )   2DS


Macro Narrowband Statistics
IEEE C802.20-04-79
• Angle Spead, Delay Spead lognormal distributions (model and simulated)
1
0.9
0.9
0.8
0.8
0.7
0.7
0.6
Pr(RMS AS<=Abscissa)
Pr(RMS DS<=abscissa)
CDF of Composite Angle Spread at Node B
1
Urban Macro
Urban Macro Ideal
Suburban Macro
Suburban Macro Ideal
0.5
0.4
0.3
0.2
0.6
0.5
0.4
Suburban Macro 5o
0.3
Suburban Macro Ideal 5o
Urban Macro 8o
0.2
Urban Macro Ideal 8o
0.1
0
Urban Macro 15o
0.1
0
Page 11
0.2
0.4
0.6
0.8
1
1.2
RMS Delay Spread seconds
1.4
1.6
1.8
2
-6
x 10
0
Urban Macro Ideal 15o
0
2
4
6
8
10
12
14
RMS Angle Spread, in degrees
16
18
20
Wideband Characterization
IEEE C802.20-04-79
After narrowband parameters are drawn per drop, the per-path details are
specified:
• Generate Power Delay Profile (PDP) (path delays, path powers)
• PDP is generated using N=6 paths (for all scenarios)
• Generate per-path Angle of Arrival and Angle spread
• Each path is assigned fixed angle spread with Laplacian Angular Power Spectrum at BS
• Each path is assigned fixed angle spread with Laplacian Angular Power Spectrum at MS
• Angle of Departure of each path at BS is Gaussian distributed
• Angle of Arrival at MS is a Gaussian random variable but also a function of the path power
Page 12
Table of Channel Parameters
Channel Scenario
Number of paths (N)
Mean AS at BS
rDS (delays/DS)
rAS (AoD/PAS)
Mean AS at BS as a lognormal
RV when simulating with 6 paths
Suburban
Macro
6
E(AS)=50
1.4
1.2
AS= 0.69
AS= 0.13
Per path AS at BS (Fixed)
2 deg
6
E(AS)=80, 150
1.7
1.3
80 AS= 0.810
AS= 0.34
0
15 AS= 1.18
AS= 0.210
2 deg
Mean AS at MS
Per path AS at MS (fixed)
Mean total RMS Delay Spread
Distribution for path delays
Narrowband composite delay
spread as a lognormal RV when
simulating
with 6 paths

 10 ^   x    , x ~ (0,1)
E(AS,MS)=680
350
E(DS)=0.17 s
E(AS,MS)=680
350
E(DS)=0.65 s
DS = - 6.80
DS = 0.288
DS = -6.18
DS = 0.18
 AS  10 ^   AS x   AS  , x ~ (0,1)
DS
Page 13
DS
DS
IEEE C802.20-04-79
Urban Macro
Urban Micro
6
NLOS: E(AS)=190
N/A
N/A
N/A
E(AS,MS)=680
350
E(DS)=0.251 s
U(0, 1.2s)
N/A
5 deg (LOS and NLOS)
Power Distribution Parameter
rDS = (delays/DS)
rAS = (AoD/PAS)
IEEE C802.20-04-79
PAS Example: If Powers are assigned
completely random in angle,
PAS = AoD so rAS = 1.0
When there is a trend of having stronger
powers in the direction of the MS,
PAS < AoD thus rAS > 1.0
PAS =
AoD
Page 14
PAS
AoD
Values for rDS & rAS were selected based on measurements
Power Delay Profile - Macro
IEEE C802.20-04-79
Steps 4,5 - Power Delay Profile (PDP):
– PDP is not deterministic as in ITU models
– N=6 distinct paths are present at any time.
– Generate random delays for each path (exponentially distributed intervals from zero):
 n'  rDS DS ln z n
z n : U (0,1)
– Order delays and shift so as 1st path has zero delay. Quantize delays to 1/16 of chip interval.
– Generate relative powers for each path (exponential profile with shadowing randomization).
(1rDS )( ( n )  (1) )
Pn  e
rDS  DS
 10
 n
where  n : N (0, 0.3 ) Normalization:
2
Pn 
Step 6 – AOD Generation per path at BS
– Gaussian random AODs centered on the LOS direction:
2
– With r = 1.2 (suburban macro), r = 1.3 (urban macro) N (0, r AS
Step 7 – PDP to AOD assignments:

Pn'
6
P'
n 1 n
  2 AS ) .
Powers, Delays draws
Assignment to Angles
– Order AODs in increased absolute value
– Assign path delays (in increasing order)
to the ordered AODs.
Page 15
1
2 3 4 5 6
1
2 3 4 5 6
Special Definitions - Micro
IEEE C802.20-04-79
Steps 4,5 - Power Delay Profile (PDP):
– Generate random delays for each path; uniformly distributed:
 n : U [0,1.2 sec]
where n  1,...N
– Order delays and shift so as 1st path has zero delay. Quantize delays to 1/16 of chip interval.
– Generate relative powers for each path (exponential profile with independent path shadowing).
Pn  10  ( n  z n )
where z n : N (0, (0.3) 2 )
(i.e. shadowingSt.Dev  3dB)
Step 6 – AOD Generation per path at BS
– Uniformly distributed random AODs centered on the LOS direction:
 n, AOD  U (40o ,40o )
Step 7 – PDP to AOD assignments:
– No ordering of path AODs
Page 16
Path Angle Spread Generation
IEEE C802.20-04-79
Step 8 – Per path channel generation at BS
– M=20 sub-paths used for each path, all
equal power, unequal angle spacing
Laplacian PAS
Even number of sub-rays
» 2o per path angle spread for macro
» 5o per path angle spread for urban micro
– Sub-path angles precalculated and fixed
for all realizations
– Laplacian power azimuth spectrum,
random phases.
   

Ray Power * # of sub-rays
Laplacian
Sigma = 2 Degrees
1
0
-10
Page 17


-8
-6
-4
-2
0
2
4
6
Angle of Departure, in degrees
8
10
Diagram of Spatial Parameters
IEEE C802.20-04-79
Cluster n
Subpath m
BS array
n ,m ,AoD
N
n ,m ,AoA
 MS
n ,m , AoA
v
MS
n ,m , AoD
MS array broadside
 BS
BS array broadside
Page 18
v
 n ,AoA
n, AoD
 BS
N
MS array
MS direction
of travel
Spatial Parameters Defined
 BS
IEEE C802.20-04-79
 n , AoD
BS antenna array orientation, defined as the difference between the broa dside of
the BS array and the absolute North (N) reference direction.
LOS AoD direction between the BS and MS, with respect to the broa dside of the
BS array.
AoD for the nth (n = 1 … N) path with respect to the LOS AoD
0 .
 n ,m ,AoD
Offset for the mth (m = 1 … M) subpath of the nth path with respect to  n , AoD .
n ,m , AoD
Absolute AoD for the mth (m = 1 … M) subpath of the nth path at the BS with
respect to the BS broa dside.
MS antenna array orientation, defined as the difference between the broa dside of
BS
 MS
MS
the MS array and the absolute North reference direction. Random: U (0,2  ]
Angle between the BS-MS LOS an d the MS broa dside.
 n , AoA
AoA for the nth (n = 1 … N) path with respect to the LOS AoA
 n ,m ,AoA
Offset for the mth (m = 1 … M) subpath of the nth path with respect to
n ,m , AoA
Absolute AoA for the mth (m = 1 … M) subpath of the nth path at the BS with
respect to the BS broa dside.
v
v
MS velocity vector. Random direction: U (0,2  ]
Angle of the velocity vector with respect to the MS broa dside:
Page 19
 0,MS .
 n , AoA .
v =arg(v).
Path AOA Distribution at Mobile
IEEE C802.20-04-79
Step 9 – Path Angle of Arrival (AOA) at the MS
– Per path AOA is a Gaussian random variable.
– Path AOA variance decreases with the path’s relative power.
 AoA = 104.12(1-exp(-0.2175*|Pr|)
– Per path AOA:
N (0, 2 AOA ) .
Measurements
provided
By Motorola
Angle of Arrival average & standard deviation versus power
150
Error Bar
Average
Standard deviation
100
Strong
Prob(dBr)
Medium
(dBr)
Weak
Angle, degrees
MS Path Angle of Arrival
y = 104.12*(1-exp(-0.2175*|x|))
50
0
-50
LOS - 180o
LOS
LOS + 180o
2304MHz, V-V polarization
-100
-25
Page 20
-20
-15
-10
-5
Average bin power relative to total, dBr
0
Path Angle Spread at Mobile
IEEE C802.20-04-79
Step 10 – Per path channel generation at MS:
– 20 sub-paths used for each path, all equal
power, unequal angle spacing
» 35o per path angle spread for all scenarios
35 Degree Laplacian
– Sub-path angles precalculated and fixed for all
realizations
– Laplacian power azimuth spectrum, random
phases.
2
1.8
1.6
Steps 11,12 – Pairing of BS-MS sub-paths,
antenna gains.
– Random pairing of sub-paths
– Assign antenna gains for BS and MS
Ray Occurrence
1.4
1.2
1
0.8
0.6
0.4
0.2
0
-100
Page 21
-80
-60
-40
-20
0
20
Azimuth, in degrees
40
60
80
100
SCM Generation
IEEE C802.20-04-79
The per path channel realization between a Tx-Rx antenna pair is a
superposition of oscillators
– Only one quantity is time evolving.
– All other quantities are fixed for duration of the drop at initialization
h u,s,n (t ) 
Pn
 SF
M
Pn SF
M

 G BS n ,m , AoD  exp j kd s sinn ,m , AoD    n ,m

M

 G MS n ,m , AoA  exp jkd u sinn ,m , AoA  
m 1
 exp jk v cos n ,m , AoA  v t




  





is the power of the nth path (Step 5).
is the lognormal shadow fading (Step 3)
is the number of subpaths per path.
G BS ( n ,m ,AoD ) is the BS antenna array gain (Step 12).
G MS ( n ,m ,AoA ) is the MS antenna array gain (Step 12).
k
is the wave number 2/  where  is the carrier wavelength in meters.
ds
is the distance in meters from BS antenna element s from the reference (s = 1) antenna. For the
reference antenna s = 1, d1 =0.
du
is the distance in meters from MS antenna element u from the reference (u = 1) antenna. For the
reference antenna u = 1, d1 =0.
Page
n,m22
is the phase of the mth subpath of the nth path (Step 8).
Model for Polarized Antennas
IEEE C802.20-04-79
Model defined to allow any type of polarized antennas on the z-plane
by decomposition into vertical & horizontal polarizations
– Power mixing between Vertical (P1) and Horizontal (P2) pols is defined by a
discrimination function: XPD=P1/P2
– For urban channels: P2 = P1 - A - B*N(0,1) (B is the St.Dev of XPD) where:
» Urban macro: A=0.34*(mean relative path power)+7.2 dB, and B=5.5dB
» Urban micro: A=8 dB, and B=8dB
– Fast Fading between the two pols is independent
– For each subpath on one pol a corresponding subpath is generated on the other pol
with a different initial phase
– The propagation characteristics of V-to-V paths are assumed to be similar to the
propagation characteristics of H-to-H paths.
– Example (X-pol transmit, X-pol receive): If BS transmits in X-pol then all path
components are decomposed to V and H ones, a mixing is generated, and reassembled at the Xpol receiver
Page 23
SCM Polarized Generation
IEEE C802.20-04-79
A matrix describes the amplitude mixing
– Definitions follow the single polarization case ones
h u ,s,n (t ) 
Pn SF
M
T
(v ,h )
  (v ) (

 exp  j n(v,,mv ) 
 (v )
r
exp
j



)
n
1
n
,m   MS ( n ,m ,AoA )
  BS n ,m ,AoD  

  (h )


(h )
M

m 1  BS (n ,m ,AoD )  rn 2 exp  j n(h,m,v )  exp  j n(h,m,h )   MS (n ,m ,AoA )

 exp jkd sin( 

)

exp
jkd
sin(

)

exp
jk
v
cos(
θ


)
t






s
n
,
m
,
AoD
u
n
,
m
,
AoA
n
,
m
,
AoA
v


v)
 (BS
( n ,m , AoD ) is the BS antenna complex response for the V-pol component.
h)
 (BS
( n ,m , AoD ) is the BS antenna complex response for the H-pol component.
v)
(MS
( n ,m , AoA ) is the MS antenna complex response for the V-pol component.
h)
(MS
( n ,m , AoA ) is the MS antenna complex response for the H-pol component.
(.)(.)
rn1
2
is the antenna gain
is the random variable representing the power ratio of waves of the nth path leaving the BS in the vertical direction
and arriving at the MS in the horizontal direction (v-h) to those leaving in the vertical direction and arriving in the
vertical direction (v-v).
rn 2
is the random variable representing the power ratio of waves of the nth path leaving the BS in the horizontal
direction and arriving at the MS in the vertical direction (h-v) to those leaving in the vertical direction and arriving
in the vertical direction (v-v). The variables rn1 andrn 2 are i.i.d.
 n( x,m,y )
Page 24
phase offset of the mth subpath of the nth path between the x component (either the horizontal h or vertical v) of
the BS element and the y component (either the horizontal h or vertical v) of the MS element.
Antenna Patterns
IEEE C802.20-04-79
SCM Model allows any mix of antenna patterns in the MS or BS array
Only single polarization antennas defined in SCM
3 Sector Antenna Pattern
0
Base Station:
-5
   

– A      min 12 
 , Am 
  3dB 

2
– 3-sector:
– 6 sector:
3dB =
70o,
3dB =
35o,
where  180    180
Gain in dB.
– 3-sector and 6-sector antennas defined
-10
-15
-20
-25
-120 -100 -80 -60 -40 -20
Am  20dB
0
20
40
60
80 100 120
Azim uth in Degrees
6 Sector Antenna Pattern
Am  23 dB
0
Mobile:
– Omnidirectional antenna at the X-Y plane at –1 dBi
Gain in dB
-5
-10
-15
-20
-25
-60 -50 -40 -30 -20 -10
0
10
20
Azim uth in Degrees
Page 25
30
40
50
60
IEEE C802.20-04-79
Far Scattering Cluster
Bad Urban Model
FS
The modified procedure includes:
L2
L1
1. Drop MS within test cell as usual.
2. Drop three FS clusters uniformly across the cell
hexagon, with a minimum radius of R = 500m.
3. Choose the FS cluster to use for the mobile that is
closest to the mobile.
4. Main cluster is assigned 4 paths, FSC assigned 2
paths
5. Assign Powers, and delays based with the FSC
components modified by:
L3
MS

FS
BS

R
FS
Excess delay due to path length
1dB/uS additional attenuation
6. Independent shadow fading per cluster.
50% STS correlation applied
7. Normalize powers of the 6 paths to unity power.
Page 26
Result:
Angle Spread and Delay
Spread is increased as in
a bad urban environment
Line of Sight Model
IEEE C802.20-04-79
Applied to Micro-cell
– Probability of occurrence is ~ 15%
– Cost 231 Walfisch-Ikegami for NLOS
Path Loss versus Distance
160
2.0 GHz: 34.53 + 38*log10(d), d in meters
Log Normal = 10 dB
2.0 GHz: 30.18 + 26*log10(d), d in meters
Log Normal = 4 dB
120
Path Loss in dB
– Cost 231 Walfisch-Ikegami Street
Channeling Model for LOS
140
100
80
Mixing function (Probability of occurrence):
(300  d )/300, 0  d  300m
P (LOS )  
0, d  300m

60
40
10
1
10
2
Distance in meters
K-factor for LOS component:
K = 13.0 – 0.03*d, K in dB, d < 300m
A direct component is added to first arriving
path, normalized with 6 other paths to
unity power and ratio K (dB)
Page 27
Result:
Angle Spread, Delay Spread,
and Path Loss statistics
affected by LOS component
Urban Canyon Model
IEEE C802.20-04-79
No building grid required
RMS Angle Spread UE CDF
1
 = 90%
0.9
0.8
0.7
Pr(RMS AS<=abscissa)
1. Select a random street orientation
which equals the direction of UE
movement.
2. Select a random orientation for the
subscriber antenna array
3. If  = U(0,1) <= 0.9 Select the UE AoAs
for all arriving paths to be equal
4. If  > 0.9
Select the directions of
arrival for all paths using the standard
SCM MS AoA model.
0.6
0.5
0.4
0.3
0.2
0.1
0
30
40
50
60
70
80
RMS Angle Spread degrees
90
100
This model emphasizes the case of all paths arriving from a common direction.
Each Path has a AS = 35°, producing this minimum value for 90% of the mobiles.
The remaining 10% of cases represent a mix of other cases, with more random
paths.
Page 28
110
Calibration – Output Statistics
Micro
IEEE C802.20-04-79
• Composite Angle Spead at BS and Delay Spead distributions in NLOS and LOS
cases of urban micro
Delay Spread
Base Angle Spread
1
1
K = 13 dB
0.9
0.9
0.8
0.8
Probability DS < Absissa
Probability AS < Absissa
K = 13 dB
0.7
0.6
0.5
Mix
0.4
0.3
0.7
0.6
0.5
0.4
Mix
0.3
K = -inf dB
0.2
0.2
0.1
0.1
K = -inf dB
0
0
Page 29
5
10
15
20
25
Angle Spread in Degrees
30
35
40
0
0
1
2
3
4
5
Delay Spread in Seconds
6
7
8
x 10
-7
Calibration – Output Statistics
Micro (II)
IEEE C802.20-04-79
• Composite Angle Spread distribution in NLOS and LOS cases of urban micro at
the MS
P

10  log10  max
• Complementary CDF of resolvable path power in dBr:
Pmin 

CDF of UE AS
10
K = 13 dB
0.9
Probability AS < Absissa
0.8
0.7
0.6
0.5
0.4
0.3
Mix
0.2
K = -inf dB
0.1
0
0
Page 30
20
40
60
80
100
120
Angle Spread in Degrees
140
160
180
CCDF Probability path power dyn range > Absissa
1
CCDF Dynamic Range of individual channel realizations
0
1x
10
10
10
delays: U(0, 1200nS)
path sigma: 3dB
3x
-1
-2
Individual Path Powers
-3
0
5
10
15
20
25
Power in dBr
30
35
40
Path Statistics
IEEE C802.20-04-79
• Number of resolvable paths (fingers) as determined by the finger assignment
procedure
• Receiver-based finger assignment assumes 1.2288Mcps and root raised cosine
filters with roll-off factor beta=0.22
No of finger probabilities - Flexible Finger Assingment Method
No of finger probabilities - Flexible Finger Assingment Method
0.9
0.45
0.8
0.4
Urban Macro
0.35
0.6
0.3
0.5
0.25
PDF
PDF
Suburban Macro
0.7
0.4
0.2
0.3
0.15
0.2
0.1
0.1
0.05
0
Page 31
0
1
2
3
4
5
6
7
8
9
Number of fingers
10
11
12
1
2
3
4
5
6
7
8
9
Number of fingers
10
11
12
Path Statistics (II)
IEEE C802.20-04-79
• Number of resolvable paths (fingers) as determined by the finger assignment
procedure
• Assumes NLOS case.
No of finger probabilities - Flexible Finger Assingment Method
0.9
0.8
Urban Micro
0.7
PDF
0.6
0.5
0.4
0.3
0.2
0.1
0
Page 32
1
2
3
4
5
6
7
8
9
Number of fingers
10
11
12
Evaluation Methodology
IEEE C802.20-04-79
Ioc Modeling
– Explicit Spatial modeling of some of the interfering sources.
Channel Metric to FER mappings
– Proposal proponent should provide metric.
– An MMSE space-time receiver can be specified as a reference example design for advanced
receivers.
– For most SIMO/MISO schemes the current methodology is sufficient
– For SIMO/MIMO schemes the use of the MMSE receiver offers a reference design
Ray Mapping
– Mapping of SCM paths into fingers for RAKE receivers
Page 33
Ioc Modeling
IEEE C802.20-04-79
–
Sophisticated receivers account for spatial characteristics of signals from
interfering bases.
–
Need to model the spatial characteristics of “strongest” B bases. (B = 8 for
120 degree sectors, B = 12 for 60 degree sectors.)
–
Remaining bases are modeled as spatially white to reduce complexity.
–
The impact of the Ioc components depends heavily on the receiver algorithm
–
Methodology:
1. Determine the pathloss (including antenna patterns) and shadowing of all
bases.
2. Rank bases in order of received power.
3. Assign the strongest base as the serving base.
4. Model the next strongest B bases as spatially correlated Gaussian noise
processes whose covariances are determined by their channel matrices.
5. Model the remaining bases as spatially white Gaussian noise processes
whose variances are based on a flat Rayleigh fading process.
Page 34
Ray Mapping
IEEE C802.20-04-79
•
Path delays appear at a resolution of 1/16 chip interval
•
A ray mapping method is needed for RAKE receivers to map the paths into
resolvable paths.
•
The SCM AHG is recommending a finger assignment procedure:
–
Determine F number of finger positions common to all receiver antennas.
Calculate the local maxima of the channel convolution that include the transmit and
receive filters. Valid local maxima are only those who satisfy both of the constraints
below:
(a) A local maximum must be within 14dB from the highest local maximum.
(b) No two consecutive local maxima can be temporally located less than a chip interval apart (i.e.
local maxima within a chip interval away from a previously selected maximum are ignored)
–
Page 35
Calculate the power contribution of each of the N paths to each finger assigned.
SCM Complexity in
System Simulations
IEEE C802.20-04-79
Initialization:
– Performed once per drop. Stores geometrical, temporal, and spatial variables
– No computational impact in system level simulations
Runtime:
– With M=20 sub-paths per path the number of calculations per Tx-Rx antenna pair and
channel sample (e.g. once per slot) are comparable to those in typical channel models
– Complexity increases linearly with the number of Tx - Rx antennas
– Polarization doubles the amount of calculations
– LOS, FSC, Urban Canyon do not impact complexity
– Ioc modeling: modeling the spatial parameters of B sectors per MS multiplies the
complexity (factor of B)
– Ray mapping procedure executed per Tx-Rx antenna pair is a linear transformation
(minimal impact)
– Overhead from Advanced Receiver (e.g. MMSE) is additional but minor
Page 36
Link Level Models
IEEE C802.20-04-79
• Link Level are considered. Models are a representative set of SCM
implementations with fixed parameters.
• Link Level models can be used for calibration purposes.
Page 37
Link Level SCM Calibration (1)
IEEE C802.20-04-79
Model
Case I
Case II
Case III
Case IV
Corresponding
Case B
Case C
Case D
Case A
Model A, D, E
Model C
Model B
Model F
Modified Pedestrian A
Vehicular A
Pedestrian B
Single
3GPP Designator*
Corresponding
3GPP2
Designator*
PDP
Path
# of Paths
1)
4+1 (LOS on, K =
6
6
1
6dB)
Delay (ns)
Relative Path Power (dB)
2)
Speed (km/h)
Page 38
4 (LOS off)
1)
0.0
2)
-Inf
1)
-6.51
2)
0.0
1)
-16.21
2)
-9.7
1)
-25.71
2)
–19.2
1)
-29.31
2)
-22.8
1)
3
2)
30, 120
0
0,0
0
0.0
0
0
-1.0
310
-0.9
200
110
-9.0
710
-4.9
800
190
-10.0
1090
-8.0
1200
410
-15.0
1730
-7.8
2300
-20.0
2510
-23.9
3700
3, 30, 120
3, 30, 120
0
0
3
Link Level SCM Calibration (2)
Topology
UE/Mobile Station
PAS
Reference 0.5λ
1)
2)
IEEE C802.20-04-79
Reference 0.5λ
LOS on: Fixed AoA for
Reference 0.5λ
RMS angle
RMS angle spread
LOS component,
spread of 35
of 35 degrees per
remaining power has
degrees per path
path with a
360 degree uniform PAS.
with a Lapacian
Lapacian
distribution
distribution
LOS off: PAS with a
Laplacian distribution,
Or 360 degree
RMS angle spread of 35
uniform PAS.
N/A
N/A
degrees per path
DoT
0
22.5
-22.5
N/A
(degrees)
AoA
(degrees)
22.5 (LOS component)
67.5 (all paths)
22.5 (odd
N/A
numbered paths),
67.5 (all other paths)
-67.5 (even
numbered paths)
Node B/ Base Station
Topology
Page 39
Reference: ULA with
0.5λ-spacing
PAS
or
4λ-spacing
or
N/A
10λ-spacing
Lapacian distribution with RMS angle spread of
2 degrees
or
N/A
5 degrees,
per path depending on AoA/AoD
AoD/AoA
50 for 2 RMS angle spread per path
(degrees)
20 for 5 RMS angle spread per path
N/A
Summary
IEEE C802.20-04-79
• The SCM recommendation is a comprehensive system level framework for
evaluation of multiple antenna techniques.
• The MBWA has already adopted most of the SCM modeling principles and has
also extended the specification for additional (indoor) environments
Page 40