8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
1/266
3GPP2 C.R1002-B
Version 1.0
Date: December, 2009
cdm a2000 Evaluat ion Methodolog y1
Revis ion B2
3
COPYRIGHT 2009
3GPP2 and its Organizational Partners claim copyright in this document and individual
Organizational Partners may copyright and issue documents or standards publications in
individual Organizational Partners name based on this document. Requests for reproductionof this document should be directed to the 3GPP2 Secretariat at [email protected].
Requests to reproduce individual Organizational Partners documents should be directed to
that Organizational Partner. Seewww.3gpp2.org for more information.
4
mailto:[email protected]://www.3gpp2.org/http://www.3gpp2.org/mailto:[email protected]8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
2/266
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
3/266
3GPP2 C.R1002-B v1.0
CONTENTS
i
FOREWORD ...................................................................................................................xiii1
REFERENCES .................................................................................................................xv2
1 Introduction ................................................................................................................ 131.1Study Objective and Scope ........................................................................................ 141.2Simulation Description Overview .............................................................................. 15
2 Evaluation Methodology for the Forward Link .............................................................. 362.1System Level Setup ................................................................................................... 37
2.1.1 Antenna Pattern ............................................................................................ 382.1.2 System Level Assumptions ............................................................................ 392.1.3 Dynamical Simulation of the Forward Link Overhead Channels ................... 10102.1.4 Reverse Link Modeling in Forward Link System Simulation ......................... 11112.1.5 Signaling Errors .......................................................................................... 11122.1.6 Fairness Criteria .......................................................................................... 1213
2.1.6.1 Fairness Criterion with the Normalized CDF of the User14Throughput ....................................................................................................... 1215
2.1.6.1.1 A Generic Proportional Fair Scheduler .............................................. 14162.1.6.2 Fairness Criterion with Geometric Mean and Harmonic Mean ............... 1517
2.1.7 C/I Predictor Model for System Simulation .................................................. 16182.2Link Level Modeling ................................................................................................ 1619
2.2.1 Link to System FER mapping ...................................................................... 16202.2.1.1 Quasi-Static Approach with Fudge Factors: .......................................... 17212.2.1.2 Quasi-Static Approach with Short Term FER: ....................................... 17222.2.1.3 Equivalent SNR Approach: .................................................................... 1923
2.2.2 Channel Models ........................................................................................... 19242.2.2.1 Channels model based on ITU channel model ....................................... 19252.2.2.2 Channels model based on SCM ............................................................. 2126
2.2.2.2.1 Channel model for system level simulations ...................................... 21272.2.2.2.2 Channel model for link level simulations ........................................... 22282.2.2.2.3 Channel model for virtual decoder generation and verification ........... 2329
2.2.3 C/I modeling for system simulation ............................................................. 23302.3Simulation Flow and Output Matrices ..................................................................... 2731
2.3.1 Simulation Flow for the Center Cell Method ................................................. 2732
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
4/266
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
5/266
3GPP2 C.R1002-B v1.0
CONTENTS
iii
3.1.8 Fairness Criteria .......................................................................................... 5413.1.9 FER Criterion .............................................................................................. 5423.1.10 IoT Criterion ................................................................................................ 543
3.2Link Level Modeling ................................................................................................ 5543.2.1 Link Level Parameters and Assumptions ..................................................... 555
3.2.1.1 Frame Erasures .................................................................................... 5563.2.1.2 Target FER ............................................................................................ 5673.2.1.3 Channel Models .................................................................................... 578
3.2.2 Forward Link Loading .................................................................................. 5893.2.3 Reverse Link Power Control ......................................................................... 5810
3.3Simulation Requirements ........................................................................................ 59113.3.1
Simulation Flow .......................................................................................... 59
12
3.3.1.1 Soft and Softer Handoff ......................................................................... 59133.3.1.2 Simulation Description ......................................................................... 60143.3.1.3 Layout Files .......................................................................................... 6115
3.3.2 Outputs and Performance Metrics ............................................................... 62163.3.2.1 General Output Matrices ...................................................................... 62173.3.2.2 Data Services and Related Output Matrices .......................................... 62183.3.2.3 1xEV-DV Systems Only ......................................................................... 6419
3.3.2.3.1 Voice Services and Related Output Matrices ...................................... 64203.3.2.3.2 Mixed Voice and Data Services .......................................................... 6421
3.3.2.4 Mixed Rev. 0 and Rev. A Mobiles (1xEV-DO Systems Only) ................... 65223.3.2.5 UMB Systems Only ............................................................................... 65233.3.2.6 Link Level Output ................................................................................. 6624
3.4Calibration Requirements ....................................................................................... 66253.4.1 Link Level Calibration .................................................................................. 66263.4.2 System Level Calibration ............................................................................. 6627
3.4.2.1 1xEV-DV System Calibration ................................................................ 66283.4.2.2 1xEV-DO System Calibration ................................................................ 66293.4.2.3 UMB System Calibration ....................................................................... 6730
3.4.2.3.1 Scheduler .......................................................................................... 68313.51xEV-DO Baseline Simulation Procedures .............................................................. 7132
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
6/266
3GPP2 C.R1002-B v1.0
CONTENTS
iv
3.5.1 Access Terminal Requirements and Procedures: .......................................... 7113.5.2 Access Network Requirements and Procedures: ........................................... 7223.5.3 Simulation Procedures ................................................................................ 723
4 Traffic Service Models ................................................................................................ 7544.1Forward Link Services............................................................................................. 755
4.1.1 Service Mix (1xEV-DV Systems Only) ........................................................... 7564.1.2 TCP Model ................................................................................................... 7574.1.3 HTTP Model ................................................................................................. 818
4.1.3.1 HTTP Traffic Model Characteristics ....................................................... 8194.1.3.2 HTTP Traffic Model Parameters ............................................................. 8310
4.1.3.2.1 Packet Arrival Model for HTTP/1.0-Burst Mode ................................. 85114.1.3.2.2
Packet Arrival Model for HTTP/1.1-Persistent Mode .......................... 87
12
4.1.4 FTP Model ................................................................................................... 90134.1.4.1 FTP Traffic Model Characteristics .......................................................... 90144.1.4.2 FTP Traffic Model Parameters................................................................ 9015
4.1.5 WAP Model .................................................................................................. 92164.1.6 Near Real Time Video Model ........................................................................ 93174.1.7 Voice Model (1xEV-DV Systems Only) .......................................................... 95184.1.8 Delay Criteria .............................................................................................. 9519
4.1.8.1 Performance Criteria for Near Real Time Video ...................................... 96204.1.8.2 Delay Criterion for WAP Users............................................................... 9621
4.2Reverse Link Services ............................................................................................. 96224.2.1 Service Mix (1xEV-DV Systems Only) ........................................................... 9623
4.2.1.1 Data Model ........................................................................................... 97244.2.1.2 Traffic Model ......................................................................................... 9725
4.2.2 TCP Modeling .............................................................................................. 98264.2.3 FTP Upload / Email ................................................................................... 103274.2.4 HTTP Model ............................................................................................... 10428
4.2.4.1 HTTP Traffic Model Parameters ........................................................... 105294.2.4.2 Packet Arrival Model for HTTP ............................................................. 107304.2.4.3 Forward Link Delay Model for HTTP Users .......................................... 10931
4.2.5 WAP Users ................................................................................................. 11032
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
7/266
3GPP2 C.R1002-B v1.0
CONTENTS
v
4.2.6 Reverse Link Delay Criteria for HTTP/WAP ................................................ 11214.2.7 Mobile Network Gaming Model .................................................................. 11324.2.8 Voice Model (1xEV-DV Systems Only) ........................................................ 1143
4.3Common Traffic Models Applicable for Both Forward Link and Reverse Link4Services ................................................................................................................ 1145
4.3.1 Voice over IP(VoIP) ..................................................................................... 11464.3.1.1 Source Configuration Files .................................................................. 11474.3.1.2 Source Files ........................................................................................ 11484.3.1.3 Simulation Specifics ........................................................................... 11594.3.1.4 VoIP Statistics..................................................................................... 116104.3.1.5 Source Mix .......................................................................................... 117114.3.1.6 Scheduler Statistics ............................................................................ 11712
4.3.2 Video Telephony(VT) .................................................................................. 117134.3.2.1 Source Configuration Files .................................................................. 117144.3.2.2 Source Files ........................................................................................ 117154.3.2.3 Simulation Specifics ........................................................................... 118164.3.2.4 VT Statistics ....................................................................................... 119174.3.2.5 Source Mix .......................................................................................... 120184.3.2.6 Scheduler Statistics ............................................................................ 12019
4.3.3
Data Demand Spatial Distribution Models And Methodology To20 Determine Network Capacity .............................................................................. 120214.3.3.1 Charactersitics of Data Demand in Wireless Wide-Area Networks ....... 120224.3.3.2 Capacity Modeling .............................................................................. 120234.3.3.3 Proposed Data Demand Spatial Distribution Model ............................. 12124
4.3.3.3.1 Central cell demand distribution pattern ......................................... 122254.3.3.3.2 Time varying demand distribution pattern ...................................... 122264.3.3.3.3 Micro demand distribution model .................................................... 12327
5 heterogeneous deployment Modeling ........................................................................ 125285.1Heterogeneous deployment modeling .................................................................... 12529
5.1.1 Cost231-Walfisch-Ikegami-NLOS model parameter values ......................... 125305.2Macro-cellular deployment modeling ..................................................................... 12531
Appendix A: Lognormal description ............................................................................... 12732Appendix B: Antenna Orientation ................................................................................. 12833
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
8/266
3GPP2 C.R1002-B v1.0
CONTENTS
vi
Appendix C: Definition of System Outage and Voice Capacity ........................................ 1301Appendix D: Formula to define various throughput and Delay Definitions ..................... 1312Appendix E: Link budget ............................................................................................... 1343Appendix F: Quasi-static Method for Link Frame Erasures Generation and Dynamically4
Simulated Forward Link Overhead Channels ................................................................. 1415Appendix G: Equalization .............................................................................................. 1526Appendix H: Max-Log-Map Turbo Decoder Metric .......................................................... 1557Appendix I: 19 Cell Wrap-Around Implementation ......................................................... 1578Appendix J: Link Level Simulation Parameters .............................................................. 1619Appendix K: Joint Technical Committee (JTC) Fader ..................................................... 16510Appendix L: Largest Extreme Value Distribution ........................................................... 16911Appendix M: Reverse Link Output Matrices ................................................................... 17012
M.1 Output Matrix ..................................................................................................... 17013M.1.1 1xEV-DV Systems ............................................................................................. 17014M.1.2 1xEV-DO Systems ............................................................................................ 17715M.2 Definitions ........................................................................................................... 18216
Appendix N: Link Prediction Methodology for Uplink System Simulations ...................... 18517N.1 Definition of Required Terms ................................................................................ 18818
N.1.1 Channel Estimation SNR, ,i p ........................................................................ 18819N.1.2 Non-Gaussian Penalty, NG ............................................................................ 18920N.1.3 Reference Curves ............................................................................................ 18921
Appendix O: Reverse Link Hybrid ARQ: Link Error Prediction Methodology Based on Convex22
Metric............................................................................................................................ 19123O.1 Convex Metric based on Channel Capacity Formula ............................................. 19124O.2 Equivalent SNR Method based on Convex Metric (ECM) ....................................... 19425
O.2.1 Overview of the Procedure .............................................................................. 19426O.2.2 Detailed Procedure ......................................................................................... 19527O.2.3 Combining Procedure for H-ARQ .................................................................... 19728
Appendix P: Pilot SINR Estimation For Power-Control Command Update in Link-Level29
Simulations ................................................................................................................... 19830Appendix Q: 1xEV-DV Reverse Link Simulation and Scheduler Procedures ................... 19931
Q.1.1 Mobile station Requirements and Procedures .................................................... 19932
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
9/266
3GPP2 C.R1002-B v1.0
CONTENTS
vii
Q.1.2 Base Station Requirements and Procedures ...................................................... 2021Q.1.3 Scheduler Requirements and Procedures .......................................................... 2042
Q.1.3.1 Scheduling, Rate Assignment and Transmission Timeline ........................... 2053Q.1.3.2 Scheduler Description and Procedures ........................................................ 2074
Q.2 Baseline specific simulation parameters ............................................................... 2105Appendix R: Modeling of D_RL(request) and D_FL(Assign) ............................................. 2136Appendix S: Symbol SNR Modeling for CDM transmission with Rake demodulation ...... 2197Appendix T: Equivalent SNR Approach for OFDM Transmission and Demodulation ...... 2228T.1 Coherence Loss due to Doppler ............................................................................. 2229T.2 Inter-tone Interference (ITI) due to Doppler ........................................................... 22210T.3 Channel Estimation Loss and Pilot Weighted Combining (PWC) ............................ 22211T.4 Antenna Combining with Receive Diversity ........................................................... 223
12
T.5 Averaging SNR in the constrained capacity domain .............................................. 22413Appendix U: File formats for VoIP and VT model ............................................................ 22614
U.1 Source Configuration File Format......................................................................... 22615U.2 Source File Format ............................................................................................... 22616U.3 Per-AT Data Reporting Format for VoIP ................................................................ 22617U.4 Network Statistics for VoIP ................................................................................... 22718U.5 Per-AT Data Reporting Format for VT ................................................................... 22719U.6 Network Statistics for VT ...................................................................................... 22720
Appendix V: Channel parameters for furge factor .......................................................... 22921Appendix W: link level statistics for generating the short-term FER curves for link-to-22
system mapping ............................................................................................................ 24123W.1 Terminology ......................................................................................................... 24124W.2 PER and SINR Definitions .................................................................................... 24225W.3 Examples ............................................................................................................ 24326
Appendix X: MTD Antenna Pattern model ...................................................................... 2452728
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
10/266
3GPP2 C.R1002-B v1.0
CONTENTS
viii
This page intentionally left blank.1
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
11/266
3GPP2 C.R1002-B v1.0
FIGURES
ix
Figure 2.1.1-1 Antenna Pattern for 3-Sector Cells .............................................................. 31Figure 2.3.1-1 Simulation Flow Chart .............................................................................. 292Figure 3.1.3-1: Simplified Call Setup Timeline for 1xEV-DV. (Timeline for 1xEV-DO is the3
same by modifying 320 ms to 427 ms) ....................................................................... 504Figure 4.1.2-1 Control Segments in TCP Connection Set-up and Release.......................... 765Figure 4.1.2-2 TCP Flow Control During Slow-Start; l= Transmission Time over the Access6
Link; rt= Roundtrip Time .......................................................................................... 777Figure 4.1.2-3 Packet Arrival Process at the Base Station for the Download of an Object8
Using TCP; PW = the Size of the TCP Congestion Window at the End of Transfer of the9
Object; Tc=c(Described in Figure 4.1.2-2) .................................................................. 8010Figure 4.1.3-1 Packet Trace of a Typical Web Browsing Session ....................................... 8111Figure 4.1.3-2 Contents in a Packet Call .......................................................................... 8212Figure 4.1.3-3 A Typical Web Page and Its Content .......................................................... 82
13
Figure 4.1.3-4 Modeling a Web Page Download ................................................................. 8514Figure 4.1.3-5 Download of an Object in HTTP/1.0-Burst Mode ....................................... 8715Figure 4.1.3-6 Download of Objects in HTTP/1.1-Persistent Mode .................................... 8916Figure 4.1.4-1 Packet Trace in a Typical FTP Session ....................................................... 9017Figure 4.1.4-2 Model for FTP Traffic ................................................................................. 9218Figure 4.1.5-1 Packet Trace for the WAP Traffic Model ...................................................... 9319Figure 4.1.6-1 Video Streaming Traffic Model ................................................................... 9420Figure 4.2.2-1: Modeling of TCP three-way handshake ..................................................... 9921Figure 4.2.2-2: TCP Flow Control During Slow-Start; l = Transmission Time over the22
Access Link (RL); rt= Roundtrip Time ...................................................................... 10023Figure 4.2.2-3 Packet Arrival Process at the mobile Station for the Upload of a File Using24
TCP .......................................................................................................................... 10325Figure 4.2.4-1: Example of events occurring during web browsing. ................................. 10526Figure 4.2.5-1: Packet Trace for the WAP Traffic Model ................................................... 11127Figure B-1 Center Cell Antenna Bearing Orientation diagram ......................................... 12828Figure B-2 Orientation of the Center Cell Hexagon ......................................................... 12829Figure B-3 Mobile Bearing orientation diagram example. ................................................ 12930Figure D-1: Description of arrival and delivered time for a packet and a packet call. ...... 13331Figure F-1 Flowchart for QPSK modulation .................................................................... 14532Figure F-2 Prediction methodology for higher order modulations without pure Chase33
combining ................................................................................................................ 14634
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
12/266
3GPP2 C.R1002-B v1.0
FIGURES
x
Figure F-3 Prediction methodology for higher order modulations with pure Chase1
combining (this corresponds to Case 1) .................................................................... 1472Figure F-4 Obtaining sample values of
s tE /N and indicators of packet errors ............... 1483
Figure F-5 Determining the Doppler penalty by evaluating the predictor performance .... 1494Figure H-1 QAM Receiver Block Diagram ........................................................................ 1565Figure I-1 Wrap-around with 9 sets of 19 cells showing the toroidal nature of the wrap-6
around surface......................................................................................................... 1587Figure I-2: The antenna orientations to be used in the wrap-around simulation. The arrows8
in the Figure show the directions that the antennas are pointing. ............................ 1609Figure K-1: I and Q Fade Multiplier Generation .............................................................. 16510Figure N-1: Outline of Equivalent SNR Method. .............................................................. 18511Figure O-1. Approximate channel capacities at BPSK and QPSK signaling (Gaussian12
signaling case is plotted as a reference). ................................................................... 193
13Figure Q-1: Set point adjustment due to rate transitions on R-SCH ................................ 20414Figure Q-2 Scheduling Delay Timing .............................................................................. 20515Figure Q-3: Parameters associated in mobile station scheduling on RL ........................... 20616Figure R-1: PDF of FL transmission delays of ESCAM on F-PDCH .................................. 21617Figure R-2: ESCAM delays on F-PDCH ........................................................................... 21718Figure T-1 Constrained Capacity Curve for 16-QAM ....................................................... 22519
20
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
13/266
3GPP2 C.R1002-B v1.0
TABLES
xi
Table 2.1-1 Forward System Level Simulation Parameters .................................................. 41Table 2.1-2 Details of Self-Interference Values Resulting in 13.5 dB of Maximum C/I for2
CDM Transmission with Rake Demodulation ............................................................... 83Table 2.1-3 Details of Self-Interference Values Resulting in 17.8 dB of Maximum C/I for4
CDM Transmission with Rake Demodulation ............................................................... 85Table 2.1-4 Details of Self-Interference Values Resulting in 17 dB of Maximum C/I for6
OFDM Transmission and Demodulation ....................................................................... 97Table 2.1-5 Signaling Errors ............................................................................................. 128Table 2.1-6 Criterion CDF ................................................................................................ 139Table 2.1-7 Web Browsing Model Parameters ................................................................... 1410Table 2.2-1 Channel Models ............................................................................................. 1911Table 2.2-2 Fractional Recovered Power and Fractional UnRecovered Power ..................... 2012Table 2.2-3 Relative Power of each Multipath Model (in dB) .............................................. 2013Table 2.2-4 Delay of each Multipath Model (in ns) ........................................................... 2014Table 2.3-1 Required 1xEV-DV Simulation Evaluation Comparison Cases Table .............. 3415Table 3.1-1 Reverse Link System Level Simulation Parameters ......................................... 4516Table 3.1-2 Backhaul bandwidth used by signaling and measurement messages ............. 5117Table 3.1-3 CDF Criterion for FTP Upload MS .................................................................. 5418Table 3.4-1 Default 1xEV-DO RL MAC Transition Probabilities ......................................... 6719Table 4.1-1 HTTP Traffic Model Parameters ...................................................................... 8420Table 4.1-2 FTP Traffic Model Parameters ........................................................................ 9121Table 4.1-3 WAP Traffic Model Parameters ....................................................................... 9322Table 4.1-4 Video Streaming Traffic Model Parameters ..................................................... 9523Table 4.2-1: Traffic Configurations ................................................................................... 9724Table 4.2-2 Delay components in the TCP model for the RL upload traffic ...................... 10125Table 4.2-3: FTP Characteristics ..................................................................................... 10426Table 4.2-4: HTTP Traffic Model Parameters ................................................................... 10727Table 4.2-5 Points to obtain the average transmission rate (ATR) given the geometry and28
channel model of a user ........................................................................................... 11029Table 4.2-6: WAP Traffic Model Parameters .................................................................... 11230Table 4.2-7 Reverse link delay criteria for HTTP request ................................................. 11231Table 4.2-8 Mobile network gaming traffic model parameters ......................................... 11332Table E-1 Link-Budget Template for the Reverse Link..................................................... 13533
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
14/266
3GPP2 C.R1002-B v1.0
TABLES
xii
Table E-2 Link-Budget Template for the Forward Link .................................................... 1371Table E-3 Propagation Index and Log-Normal Sigma Values from [20] ............................ 1402Table J-1 Link Level Simulation Parameters for Forward Link ........................................ 1613Table J-2 Link Level Simulation Parameters for Reverse Link ......................................... 1634Table K-1: Coefficients of the 6-tap FIR Filter ................................................................. 1655Table K-2: Coefficients of the 8-tap FIR Filter ................................................................. 1656Table K-3: Coefficients of the 11-tap Filter ...................................................................... 1667Table K-4: Coefficients of the 28-tap Filter ...................................................................... 1668Table K-5: Jakes (Classic) Spectrum IIR Filter Coefficients ............................................. 1679Table M-1 Required statistics output in excel spread sheet for the base station side ....... 17010Table M-2 Required statistics output in excel spread sheet for the base station side ....... 17611Table M-3 Required statistics output in excel spread sheet for the base station side ....... 177
12
Table N-1. Notations used. ............................................................................................. 18613Table R-1: D_RL(request) delay for Method a .................................................................. 21314Table R-2: D_RL(request) delay for Method b .................................................................. 21415Table R-3: D_FL(assign) delay for Method a .................................................................... 21416Table R-4: D_FL(assign) delay for Method b (excluding F-PDCH scheduling delay) .......... 21517Table R-5: Reference table for mean transmission times vs Geometry ............................. 21718
19
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
15/266
3GPP2 C.R1002-B v1.0
FOREWORD
xiii
This document was prepared by Technical Specification Group C of the Third Generation 31
Partnership Project 2 (3GPP2).2
Revision 0 of this document was used for the evaluation and analysis leading to the3
development of the following cdma2000 systems specifications: cdma20001 Revision C4
(1xEV-DV), cdma2000 Revision D (1xEV-DV), and cdma2000 High Rate Packet Data Air5 Interface Revision A (1xEV-DO).6
Revision A of this document also includes evaluation methodology for cdma2000 High Rate7
Broadcast-Multicast Packet Data Air Interface (1xEV-DO BCMCS) and UMB (Ultra Mobile8
Broadband)2.9
Revision B of this document further includes evaluation methodology to handle more10
practical deployment scenarios such as using actual traffic models, data demand11
differential in networks, heterogeneous deployment, etc, and evaluation methodology for12
reverse link transmit diversity for cdma2000 High Rate Packet Data Air Interfacce (1xEV-13
DO).14
15
1 cdma2000 is the trademark for the technical nomenclature for certain specifications and
standards of the Organizational Partners (OPs) of 3GPP2. Geographically (and as of the date of
publication), cdma2000 is a registered trademark of the Telecommunications Industry Association
(TIA-USA) in the United States.
2 Ultra Mobile Broadband and (UMB) are trade and service marks owned by the CDMA
Development Group (CDG).
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
16/266
3GPP2 C.R1002-B v1.0
FOREWORD
xiv
This page intentionally left blank.1
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
17/266
3GPP2 C.R1002-B v1.0
REFERENCES
xv
Normative References1
The following specifications contain provisions which, through reference in this text,2
constitute provisions of this Specification. At the time of publication, the editions indicated3
were valid. If the specification version number is included, the reference is specific. Parties4
implementing this Specification should use the specific versions of the indicated5
specification. If the specification version number is not included, the reference is non-6specific. Parties implementing this Specification are encouraged to investigate the7
possibility of applying the most recent editions of the indicated specifications.8
[1] ETSI TR 101 12, Universal Mobile Telecommunications System (UMTS); Selection9
procedures for the choice of radio transmission technologies of the UMTS (UMTS 30.0310
v3.2.0)11
[2] ITU-RM.1225, Guidelines for Evaluation of Radio Transmission Technologies for IMT-12
2000.13
[3] A. Viterbi, Principles of Spread Spectrum Communication, Addison-Wesley, 1995.14
[4] F. Ling, Optimal Reception, performance bound, and cutoff rate analysis of reference-15assisted coherent CDMA communications with applications, IEEE Trans. on Commun.,16
46(10), pp. 15831592, October 1999.17
[5] Motorola, "HTTP Traffic Models for 1xEV-DV Simulations", Contribution 3GPP2-C50-18
Eval-20010212-004.19
[6] Motorola, " HTTP Traffic Models for 1xEV-DV Simulations (v2)", Contribution 3GPP2-20
C50-Eval-20010321-002-HTTP-traffic.21
[7] R. Fielding, J. Gettys, J. C. Mogul, H. Frystik, L. Masinter, P. Leach, and T. Berbers-Lee,22
"Hypertext Transfer Protocol - HTTP/1.1", RFC 2616, HTTP Working Group, June 1999.23
ftp://ftp.Ietf.org/rfc2616.txt.24
[8] B. Krishnamurthy and M. Arlitt, "PRO-COW: Protocol Compliance on the Web",25
Technical Report 990803-05-TM, AT&T Labs, August 1999,26
http://www.research.att.com/~bala/papers/procow-1.ps.gz.27
[9] Lucent, "Comments on HTTP traffic model", Contribution 3GPP2-C50-Eval-20010323-28
001-traffic-comments.29
[10] J. Cao, William S. Cleveland, Dong Lin, Don X. Sun., "On the Nonstationarity of30
Internet Traffic", Proc. ACM SIGMETRICS 2001, pp. 102-112, 2001.31
[11] B. Krishnamurthy, C. E. Wills, "Analyzing Factors That Influence End-to-End Web32
Performance", http://www9.org/w9cdrom/371/371.html33
[12] H. K. Choi, J. O. Limb, "A Behavioral Model of Web Traffic", Proceedings of the seventh34
International Conference on Network Protocols, 1999 (ICNP '99), pages 327-334.35
[13] F. D. Smith, F. H. Campos, K. Jeffay, D. Ott, "What TCP/IP Protocol Headers Can Tell36
Us About the Web", Proc. 2001 ACM SIGMETRICS International Conference on37
Measurement and Modeling of Computer Systems, pp. 245-256, Cambridge, MA June38
2001.39
http://www.research.att.com/~bala/papers/procow-1.ps.gzhttp://www.research.att.com/~bala/papers/procow-1.ps.gzhttp://www.research.att.com/~bala/papers/procow-1.ps.gz8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
18/266
3GPP2 C.R1002-B v1.0
REFERENCES
xvi
[14] P. Barford and M Crovella, "Generating Representative Web Workloads for Network and1
Server Performance Evaluation" In Proc. ACM SIGMETRICS International Conference on2
Measurement and Modeling of Computer Systems, pp. 151-160, July 1998.3
[15] S. Deng. Empirical Model of WWW Document Arrivals at Access Link. In Proceedings4
of the 1996 IEEE International Conference on Communication, June 19965
[16] W. R. Stevens, "TCP/IP Illustrated, Vol. 1", Addison-Wesley Professional Computing6
Series, 1994.7
[17] Motorola, "Comments on Data Traffic Mix", Contribution 3GPP2-C50-Eval-20010321-8
006-Mot-traffic-mix.9
[18] UMTS 30.03 V3.2.0 "Universal Mobile Telecommunications Systems (UMTS); Selection10
procedures for the choice of radio transmission technologies of the UMTS, 1998-04, pg 33-11
35.12
[19] K. C. Claffy, "Internet measurement and data analysis: passive and active13
measurement", http://www.caida.org/outreach/papers/Nae/4hansen.html.14
[20] ITU, Guidelines for Evaluation of Radio Transmission Technologies for IMT-2000,15
Recommendation ITU-R M.1225, 1997.16
[21] Bob Love and Frank Zhou, Comments on Qualcomm Link Budget for 1xEV -DV,17
Motorola contribution 3GPP2-C50-Eval-20001205-001 to the Evaluation Ad Hoc, December18
5, 2000.19
[22] Tao Chen, Link Budget Examples for 1xEV-DV Proposal Evaluation, Rev. 2,20
QUALCOMM contribution 3GPP2-C50-Criteria Ad Hoc-20001115-002 to the WG5 Criteria21
Ad Hoc, November 15, 2000.22
[23] Louay Jalloul, Comments on Path Loss Models for System Simulations, Motorola23
contribution 3GPP2-C50-WG5-20001116-003 to the Simulation Ad Hoc, November 16,242000.25
[24] Steve Dennett, The cdma2000 ITU-R RTT Candidate Submission (0.18), July 27,26
199827
[25] Working Group 5 evaluation ad hoc chair, 1xEV -DV Evaluation Methodology28
Addendum, version 6, 3GPP2 TSG-C contribution to Working Group 5 in the Portland,29
Oregon meeting, C50-20010820-026, August 20, 2001.30
[26] 3GPP2/TSG-C - C50-Eval-20010329-001, Link Error Prediction Methodology, Lucent31
Technologies, March 2001.32
[27] 3GPP2/TSG-C C30-20030217-010, Link Prediction Methodology for Reverse Link33 System Simulations, Lucent Technologies, February 2003.34
[28] 3GPP2/TSG-CC30-20030217-010A Link Prediction Methodology for Uplink System35
Level SimulationsAnalysis, Lucent Technologies, February 2003.36
[29] TIA/EIA-IS-2000.2, Mobile Station-Base Station Compatibility Standard for Dual-37
Mode Wideband Spread Spectrum Cellular System, June, 2002.38
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
19/266
3GPP2 C.R1002-B v1.0
REFERENCES
xvii
[30] Ramakrishna, S., Holtzman, J.M., A scheme for throughput maximization in a dual-1
class CDMA system, Selected Areas in Communications, IEEE Journal on, Vol. 16 Issue:2
6, page 830844, Aug. 1998.3
[31] F. Kelly, Charging and rate control for elastic traffic, European Trans. On4
Telecommunications, vol. 8, pp. 33-37, 1997.5
[32] L3NQS, Results of L3NQS Simulation Study, 3GPP2-C50-20010820-011, August6
2001.7
[33] Doug Reed, Modified Hata path loss model used in 3GPP2 , 3GPP2 Contribution8
C30-20040920-012, Motorola.9
[34] 3GPP2/TSG-CC30-20040823-063R1 Qualcomm, Confidence Interval, August 23rd,10
2004.11
[35] Ye Li, Leonard Cimini, Bounds on the Interchannel Interference of OFDM in Time-12
Varying Impairments, IEEE Transactions on Communications, Vol 49, No. 3, March 2001.13
[36] 3GPP2/TSG-C WG3 contribution, C30-20060413-006, SampleSrcConfigFile_VTMix614
AT, April 2006.15
[37] 3GPP2/TSG-C WG3 contribution, C30-20060413-005, MSOWithBlankingSourceFile,16
April 2006.17
[38] 3GPP2/TSG-C WG3 contribution, C30-20060327-030A, FixedRateFixedQualityVideo18
SourceFile, April 2006.19
[39] 3GPP2 C.S0076-0 v1.0, Discontinuous Transmission (DTX) of Speech in cdma200020
Systems, December, 2005.21
[40] 3GPP2/TSG-C WG1 contribution, C12-20051012-006, Video Database for 3GPP222
multimedia services, October 2005.23
[41] 3GPP2/TSG-C WG3 contribution, C30-20030915-006, SCM-135 Channel Model Text,24
September 2006.25
[42] 3GPP2/TSG-C WG3 contribution, C30-20060823-005, FL VoIP packet arrival with26
jitter.dat,, August 2006.27
[43] 3GPP2/TSG-C WG3 contribution, C30-20080114-030, Updated location files for28
calibration,January 2008.29
[44] 3GPP2/TSG-C WG3 contribution, C30-20080114-029, Effective SNR AWGN Curves,30
January 2008.31
[45] 3GPP2/TSG-C WG3 contribution, C30-20080331-012R3, Performance Evaluation32
Parameters,March 2008.33
[46] 3GPP2/TSG-C WG3 contribution, C30-20080114-021R2, Calibration output metrics,34
January 2008.35
[47] 3GPP2/TSG-C WG3 contribution, C30-20071203-020R2, SNR to CQI mapping in36
calibration,December 2007.37
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
20/266
3GPP2 C.R1002-B v1.0
REFERENCES
xviii
[48] 3GPP2/TSG-C WG3 contribution, C30-20090615-025A, MTD antenna pattern files,1
June 2009.2
3
Informative References4
The following documents do not contain provisions of the Specification. They are listed to5 aid in better understanding this Specification.6
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
21/266
3GPP2 C.R1002-B v1.0
1
1 INTRODUCTION1
1.1 Study Objective and Scope2
The objective of this document is to explain the set of definitions, assumptions, and a3
general framework for simulating cdma2000 systems (e.g., 1xEV-DV and 1xEV-DO) and4UMB(Ultra Mobile Broadband) systems to arrive at system wide voice, data, or both voice5
and data performance on the forward and reverse links.6
This document was used in the evaluation and analysis leading to the development of the7
following specifications: cdma2000 Revision C (1xEV-DV), cdma2000 Revision D (1xEV-DV),8
cdma2000 High Rate Packet Data Air Interface Revision A (1xEV-DO), cdma2000 High Rate9
Broadcast-Multicast Packet Data Air Interface (1xEV-DO BCMCS) and UMB.10
This document also defines the necessary framework for simulating the performance of11
cdma2000 and UMB systems with proposed enhancements that are not part of the current12
cdma2000 and UMB family of specifications. The proponent(s) of any proposal shall provide13
the details required so that other companies can evaluate the proposal independently. The14proponent(s) of any simulation results shall provide the details required so that other15
companies can repeat the simulation independently. The information about the simulations16
will include the predictors being used, and the reported results will include the prediction17
errors (bias and standard deviation).18
1.2 Simulation Description Overview19
Determining voice and high rate packet data system performance requires a dynamic20
system simulation tool to accurately model feedback loops, signal latency, protocol21
execution, and random packet arrival in a multipath-fading environment. The packet22
system simulation tool will include Rayleigh and Rician fading and evolve in time with23
discrete steps (e.g., time steps of 1.25 ms or 1.67 ms). The time steps need to be small24enough to correctly model feedback loops, latencies, scheduling activities, and25
measurements of the proposed system.26
27
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
22/266
3GPP2 C.R1002-B v1.0
2
This page intentionally left blank.1
2
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
23/266
3GPP2 C.R1002-B v1.0
3
2 EVALUATION METHODOLOGY FOR THE FORWARD LINK1
2.1 System Level Setup2
2.1.1 Antenna Pattern3
The antenna pattern used for each sector, reverse link and forward link, is plotted inFigure4
2.1.1-1 and is specified by5
2
3
min 12 , , where 180 180.mdB
A A
2.1-16
dB3 is the 3 dB beamwidth, and dBAm 20 is the maximum attenuation.7
-25
-20
-15
-10
-5
0
-180 -150 -120 -90 -60 -30 0 30 60 90 120 150 180
Horizontal Angle - Degrees
Gain-dB
8
Figure 2.1.1-1 Antenna Pattern for 3-Sector Cells9
2.1.2 System Level Assumptions10
The parameters used in the simulation are listed in Table 2.1-1. Where values are not11
shown, the values and assumptions used shall be specified in the simulation description.12
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
24/266
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
25/266
3GPP2 C.R1002-B v1.0
5
Parameter Value Comments
Forward Link
Overhead
Channel
Resource
Consumption
Circuit
switched
and packet
switched
data
systems
(e.g., 1xEV-
DV)
Pilot, Paging and Sync
overhead: 20%.
Any additional overhead
needed to support other
control channels (dedicated
or common) must be
specified and accounted for
in the simulation
Packet
switched
data
systems
(e.g., 1xEV-
DO)
FL MAC, Preamble and Pilot
channel overhead shall be
considered. The portion of
times the Control Channel
(CC) (38.4 kbps or 76.8
kbps) is sent shall be set as
a fixed TDM overhead.
CC portion is assumed to be
6.25% of the total time. Any
additional overhead must be
specified and accounted for
in the simulation.
Mobile Noise Figure 10.0 dB
Thermal Noise Density -174 dBm/Hz
Carrier Frequency 2 GHz
BS Antenna Gain with Cable
Loss
15 dB 17 dB BS antenna gain; 2
dB cable loss
MS Antenna Gain -1 dBi
Other Losses 10 dB Applicable to all fading
models
Fast Fading Model Based on Speed SeeTable 2.2-1.The fadingmodel is specified in
Appendix K. With dual
antenna receiver, the fading
processes on the paths from
a given BS to the MS receive
antennas are mutually
independent.
Active SetMembership
Circuit
switched and
packet
switched
data systems
(e.g., 1xEV-
DV)
Up to 3 members are in the
Active Set if the pilot Ec/Io
is larger than T_ADD = -18
dB (=9 dB below the FL pilot
Ec/Ior) based on the FL
evaluation methodology
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
26/266
3GPP2 C.R1002-B v1.0
6
Parameter Value Comments
Packet
switched
data systems
(e.g., 1xEV-
DO)
Up to 3 members are in the
Active Set if the pilot Ec/Io
is larger than T_ADD = -9 dB
based on the FL evaluation
methodology
Finger Assignment
Threshold (T_PATH)
-12 dB A finger may be assigned to
a multipath component only
if its (Ec/Io) exceeds the
finger assignment threshold.
This parameter should be
used only for 1xEV-DO
BCMCS.
(SeeAppendix S)
Maximum Number of Paths
assigned to Rake fingers(MAX_NUM_PATHS)
8 for single RX-antenna
receivers, and
4 for dual RX-antenna
receivers
This is the maximum
number of paths to whichRake fingers may be
assigned at the MS.
Delay Spread Model SeeTable 2.2-1 andTable
2.2-2
Fast Cell Site Selection Disable. The overhead shall
be accounted for if it is used
in the proposal.
Forward Link
Power
Control
Circuit
switched and
packet
switched
data systems
(e.g., 1xEV-
DV)
(If used on
dedicated
channel)
Power Control loop delay:
two PCGs1
Update Rate: Up to 800Hz
PC BER: 4%
1One PCG/slot delay in link level modeling (measured from the time that the SIR is sampled to the
time that the BS changes TX power level.)
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
27/266
3GPP2 C.R1002-B v1.0
7
Parameter Value Comments
Packet
switched
data systems
(e.g., 1xEV-
DO)
(If used on
MAC
channels)
Power Control loop delay:
two slots1
Update Rate: Up to 600 Hz
PC BER: 4%,
Based on DRC feedback
BS Maximum PA Power 20 Watts
Site to Site distance 2.5 km
1.9 km Determined by RL Link
Budget of 1xEV-D0 Rev-A
2.0 km Default site to site distance
for UMBTotal Path Loss Threshold 140 dB This term includes the MS
and BS antenna gains, cable
and connector losses, other
losses, and shadowing, but
not fading. A subscriber
whose total path loss on the
best forward link exceeds the
Total Path Loss Threshold is
redropped. This value
should be applied when site-
to-site distance is 1.9km and2.0km.
Maximum C/I achievable,
where C is the
instantaneous total received
signal from the serving base
station(s) (usually also
referred to as rx_Ior(t), or
or(t)), and I is the
instantaneous total
interference level (usually
13.5 dB and 17.8 dB for
CDM transmission with
Rake demodulation
17 dB 18.1 dB1, and 28dB2
for OFDM transmission and
demodulation
13.5 dB for typical current
subscriber designs for IS-95
and cdma2000 1x systems;
17.8 dB for improved
subscriber designs for 1xEV-
DV and 1xEV-DO systems;
18.1dB and 28dB for
improved subscriber designs
for UMB. The details on howthese values were derived
1The max AN C/I of 18.5 dB and the max AT C/I of 29dB are assumed. 18.1dB is the geometric
mean of those two values.
2C/I required to meet 1% FER for the packet of 64QAM, code rate 3/4, and 2x2 SCW in channel of
PedB 3km/h is 26dB. 28dB max C/I is obtained by adding 2 dB margin.
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
28/266
3GPP2 C.R1002-B v1.0
8
Parameter Value Comments
also referred to as Nt(t)). are inTable 2.1-2, Table
2.1-3,andTable 2.1-4.(C/I)maxfor other
transmission/demodulation
schemes shall be providedwith justification, by the
system proponents.
1
Table 2.1-2 Details of Self-Interference Values Resulting in 13.5 dB of Maximum C/I2
for CDM Transmission with Rake Demodulation3
Contribution of Self-Interference )(/ iselfor II Note
Base-band pulse shaping waveform 16.5dB IS-95 Tx filter and matched
Rx filter
Radio noise floor 20dB With improved Tx-Rx. Noise
performance
ADC quantization noise 20dB 4-bit A/D converter
Adjacent channel interference 27dB 1.25 MHz spacing
4
Table 2.1-3 Details of Self-Interference Values Resulting in 17.8 dB of Maximum C/I5
for CDM Transmission with Rake Demodulation6
Contribution of Self-Interference )(/ iselfor II Note
Base-band pulse shaping waveform 24 dB IS-95 Tx filter with 64-tap
Rx filter
Radio noise floor 20 dB For Tx RHO increased to
99%
ADC quantization noise 31.9 dB 6-bit A/D converter
Adjacent channel interference 27 dB 1.25 MHz spacing
7
8
9
10
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
29/266
3GPP2 C.R1002-B v1.0
9
Table 2.1-4 Details of Self-Interference Values Resulting in 17 dB of Maximum C/I for1
OFDM Transmission and Demodulation2
Contribution of Self-Interference )(/ iselfor II Note
Base-band pulse shaping waveform Not Applicable OFDM transmission anddemodulation to eliminate
pulse ISI
Radio noise floor 20dB With improved Tx-Rx. Noise
performance
ADC quantization noise 20 dB 4-bit A/D converter
Adjacent channel interference Not applicable Guard tones to eliminate
ACI
3
The maximum C/I achievable in the subscriber receiver is limited by several sources,4including inter-chip interference induced by the base-band pulse shaping waveform, the5
radio noise floor, ADC quantization error, and adjacent carrier interference. For 1xEV-DO6
BCMCS, the noise floor associated with the maximum C/I limitation is modeled as7
described in Section2.2.2 andAppendix S.8
In the system level simulation, the noise floor associated with the maximum C/I limitation9can be characterized by the parameter , given by10
ma x/1
IC 2.1-211
where ma xIC denotes the maximum achievable C/I for the subscriber receiver. As12indicated in Table 2.1-1,
ma xIC is assumed to be 13.5 dB for the current IS-95 and13
cdma2000 1X subscriber receivers, and 17.8 dB for improved 1xEV-DV/1xEV-DO designs.14Thus, = 0.045 and = 0.0166 for maximum C/I values of 13.5 dB and 17.8 dB,15
respectively.16
In the system level simulation for CDM transmission with Rake demodulation, the effective17C/I shall be given by18
combined
effective
)/(
1
1
IC
IC 2.1-319
where combined
IC is the instantaneous signal-to-interference ratio after pilot-weighted20
combining of the Rake fingers (see section 2.2.2 for detail). The effective signal-to-21
interference ratio, effective
IC , accounts for the interference sources associated with the22
maximum C/I limitation, and shall be used as the C/I observed by the mobile station23
receiver.24
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
30/266
3GPP2 C.R1002-B v1.0
10
The channel between the serving cell and the subscriber is modeled using the channel1
models defined in section 2.2.2. The channel between any interfering cell and the2
subscriber can be modeled as a one-path Rayleigh fading channel, where the Doppler of the3
fading process is randomly chosen based on the velocities specified inTable 2.2-1 and its4
corresponding probabilities.5
If transmit diversity is used in cdma2000 1x and 1xEV-DV systems, the transmit diversity6PA size shall be the same as the main PA size, 12.5% of the main PA power shall be used7
for the Pilot Channel and 7.5% for the Paging Channel and Sync Channel. The Transmit8
Diversity Pilot Channel power is half the power of the Pilot Channel. For example, if the9
main PA size is 20 W, then the transmit diversity PA size is 20 W, 2.5 W of the main PA is10
for the Pilot Channel, 1.5 W of the main PA for the Paging Channel and Sync Channel, and11
1.25 W of the transmit diversity PA is for the Transmit Diversity Pilot Channel.12
2.1.3 Dynamical Simulation of the Forward Link Overhead Channels13
Dynamically simulating the overhead channels for 1xEV-DV or 1xEV-DO systems is14
essential to capture the dynamic nature of power and code space allocation to these15 channels. The simulations shall be done as follows:16
1) The performance of the new overhead channels (other than the Pilot, Sync, and17
Paging Channels for 1xEV-DV systems or the Pilot and control channels for 1xEV-18
DO systems) must be included in the system level simulations. The Pilot Channel,19
Sync Channel, and Paging Channel are taken into account as part of the fixed20
overhead (power and code space) in 1xEV-DV systems. For 1xEV-DO systems, the21
Pilot, preamble, and the total FL MAC shall be transmitted at full BTS power (20 W),22
and the 38.4 kbps and 76.8 kbps Control Channels are taken into account as part23
of the fixed overhead (as a fixed percentage of the total transmission time).24
2) There are two types of these new overhead channels: static and dynamic. A static25
overhead channel requires fixed base station power. A dynamic overhead channel26
requires dynamic base station power.27
3) The system level simulations do not directly include the coding and decoding of28
these new overhead channels. There are two aspects that are important for the29
system level simulation: the required Ec/Ior during the simulation interval (e.g., a30
power control group or slot) and demodulation performance (detection, miss, and31
error probabilitywhatever is appropriate).32
4) The link level performance is evaluated off-line by using separate link-level33
simulations. A quasi-static approach shall be used to conduct the link-level34
simulation. The performance is characterized by curves of detection, miss, false35
alarm, and error probability (whatever is appropriate) versus Eb/No.36
5) For static overhead channels, the system simulation should compute the received37
Eb/No.38
6) For dynamic overhead channels with open-loop control only, the simulations should39
take into account the estimate of the required forward link power that needed to be40
transmitted to the mobile station. For dynamic overhead channels that use closed41
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
31/266
3GPP2 C.R1002-B v1.0
11
loop feedback, the base station allocates forward link power based upon the1
combination of open-loop and closed-loop feedback. During the reception of2
overhead information, the system simulation should compute the received Eb/No.3
7) Once the received Eb/No is obtained, then the various miss error events should be4
determined. The impact of these events should then be modeled. The false alarm5
events are evaluated in link-level simulation, and the simulation results will be6included in the evaluation report. The impact of false alarm, such as delay increases7
and throughput reductions for both the forward and reverse links, will be8
appropriately taken into account in system-level simulation.9
8) The Walsh space utilization shall be modeling dynamically for 1xEV-DV systems.10
9) All new overhead channels shall be modeled.11
10)If a proposal adds messages to an existing channel (overhead or otherwise), the12
proponent shall justify that this can be done without creating undue loading on this13
channel. If a proposal requires an additional overhead channel of the type that is14
already in the system under evaluation, then the proposal shall include the power15 required for this channel. The system level and link level simulation required for16
this modified overhead channel as a result of the new messages shall be performed17
according to 3) and 4), respectively.18
2.1.4 Reverse Link Modeling in Forward Link System Simulation19
The proponents shall only model feedback errors (e.g., power control, acknowledgements,20
rate indication, etc.) and measurements (e.g., C/I measurement) without explicitly modeling21
the reverse link and reverse link channels. In addition to supplying the feedback error rate22
average and distribution, the measurement error model and selected parameters, the23
estimated power level required for the physical reverse link channels will be supplied24
(including those used for fast cell selection even though it is not going to be explicitly25modeled for the 1xEV-DV or 1xEV-DO system simulations).26
2.1.5 Signaling Errors27
Signaling errors shall be modeled and specified as inTable 2.1-5.28
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
32/266
3GPP2 C.R1002-B v1.0
12
Table 2.1-5 Signaling Errors1
Signaling Channel Errors Impact
ACK/NACK channel Misinterpretation, missed
detection, or false detection
of the ACK/NACK message
Transmission (frame or
encoder packet) error or
duplicate transmission
Explicit Rate Indication Misinterpretation of rate One or more transmission
errors due to decoding at a
different rate (modulation
and coding scheme)
User identification channel A user tries to decode a
transmission destined for
another user; a user misses
transmission destined to it.
One or more transmission
errors due to HARQ/IR
combining of wrong
transmissions
Rate or C/I feedbackchannel (DRC or equivalent) Misinterpretation of rate orC/I for DRC feedback
information
Potential transmission errors
Fast cell site selection
signaling, e.g., transmit
sector indication, transfer of
H-ARQ states etc.
Misinterpretation of selected
sector; misinterpretation of
frames to be retransmitted.
Transmission errors
Proponents shall quantify and justify the signaling errors and their impacts in the2
evaluation report. As an example, if an ACK is misinterpreted as a NACK (duplicate3
transmission), the packet call throughput will be scaled down by (1-pACK), wherepACKis the4
ACK error probability.5
2.1.6 Fairness Criteria6
2.1.6.1 Fairness Criterion with the Normalized CDF of the User Throughput7
Because maximum system capacity may be obtained by providing low throughput to some8
users, it is important that all mobile stations be provided with a minimal level of9
throughput. This is called fairness. The fairness is evaluated by determining the10
normalized cumulative distribution function (CDF) of the user throughput, which meets a11
predetermined function in two tests (seven test conditions). The same scheduling12
algorithm shall be used for all simulation runs. That is, the scheduling algorithm is13
not to be optimized for runs with different traffic mixes. The proponent(s) of any14proposal are also to specify the scheduling algorithm.15
Let Tput[k] be the throughput for user k. The normalized throughput with respect to the16
average user throughput for user k, ]k[T~
put is given by17
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
33/266
3GPP2 C.R1002-B v1.0
13
][iTavg
][kT]k[T
~
puti
put
put 2.1-41
The CDF of the normalized throughputs with respect to the average user throughput for all2
users is determined. This CDF shall lie to the right of the curve given by the three points in3
Table 2.1-6.4
Table 2.1-6 Criterion CDF5
Normalized
Throughput w.r.t
average user
throughput
CDF
0.1 0.1
0.2 0.2
0.5 0.5
6
This CDF shall be met for the seven test conditions given in the following two tests:7
Test 1for FTP, six test conditions8
Single path Rayleigh fading9
3, 30, 100 km/h10
All FTP users, with buffers always fullNote that this model differs from the11
FTP traffic model specified in section4.1.412 10, 20 users dropped uniformly in a sector13
80% (for cdma2000 1x and 1xEV-DV systems) or 100% (for 1xEV-DO14systems) of BS power available for data users; max. BS power = 20 w15
Full BS power from other cells16
The 6 test conditions are the combinations 3, 30, and 100 km/h with 10 and1720 FTP users per sector18
Test 2for HTTP, one test condition19
Single path Rayleigh fading20
3 km/h21
HTTP users, with traffic model provided inTable 2.1-7Note that this traffic22model differs from the HTTP traffic model specified in section4.1.323
44 users dropped uniformly in a sector24
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
34/266
3GPP2 C.R1002-B v1.0
14
70% (for cdma2000 1x and 1xEV-DV systems) or 100% (for 1xEV-DO1systems) of BS power available for data users; max. BS power = 20 w2
Full BS power from other cells3
4
Table 2.1-7 Web Browsing Model Parameters5
Process Random Variable Parameters
Packet Calls Size Pareto with cutoff =1.2, k=4.5 Kbytes, m=2
Mbytes, = 25 Kbytes
Time Between Packet
Calls
Geometric = 5 seconds
2.1.6.1.1 A Generic Proportional Fair Scheduler6
Although the proponent of a proposal is free to use any scheduler, a generic proportional-7
fair scheduler [31,32], for full-buffer traffic model, with a priority function Pi(k) is given8
below for reference:9
( )ii
i
R kP k
T k
2.1-510
where k is the slot index, )(kRi is the data rate potentially achievable for the i-th mobile11
station based upon the reported C/I and the power available to the F-PDCH, kTi is the12average fairness throughput of the i-th mobile station up to time k, and is the fairness13exponent factor with the default value chosen as 0.75. Users with the highest priority are14
selected for service. The number of users selected is dependent upon the number of users15to be serviced simultaneously. The average fairness throughput can be calculated as16
follows:17
( 1) - 1
( 1) (1 ) ( 1) - 1
i
i
i i
T k if the i th MS was not scheduled at time k T k
T k N k if the i th MS was scheduled at time k
2.1-618
where19
-3
3
1.25 101 for 1x EV-DV Systems
=1.67 10
1 for 1xEV-DO Systems
t
t
2.1-720
tis set to 1.5s, Ni(k-1)is the number of bits delivered to the MS at time k-1 and Tishould be21
initialized to a small value greater than zero.22
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
35/266
3GPP2 C.R1002-B v1.0
15
2.1.6.2 Fairness Criterion with Geometric Mean and Harmonic Mean1
The fundamental problem of using the fairness criterion with the normalized cdf of the user2
throughput is that this fairness criterion takes the form of an absolute bound on relative3
throughput. This means that it is possible that when there is extra capacity available for4
some users but not for others, the potential extra service is disallowed, as it would violate5
the fairness criteria. Yet it is surely always of positive benefit to be able to improve the6service of some users, while not reducing service on any other users.7
A second problem is that the fairness criterion with the normalized cdf of the user8
throughput is divorced from the final metric, the sum-throughput. The fairness criterion9
effectively defines a set of feasible allocations, and the goal becomes to maximize the sum-10
throughput over this feasible set. This problem is generally non-trivial, and often leads to11
iterative simulation runs searching for this optimal point for each set of simulation12
parameters, especially since it is difficult a priori to know if a given scheduling policy will13
lead to a feasible allocation. The complex tradeoffs made to find this optimal allocation are14
non-trivial and often non-transparent, and in the end are not really the appropriate15
tradeoffs in realistic networks, due to the artificial nature of the fairness bound.16In this criterion a single metric is determined from the set of full-buffer throughputs which17
applies different weighting (in the form of utility) to different levels of throughput. This18
metric is to be optimized for each simulation scenario and used for comparison across19
proposals. Due to the nature of the variable weighting across rates, a limitation on the20
distribution of MS throughputs is a direct result of optimizing the metric. Hence no further21
criterion on throughput fairness is required. This method is to be used both for pure full-22
buffer simulations and for mixed simulations which include full-buffer users. In the latter23
case, only the throughputs of the full-buffer users are used for these comparisons.24
There are two separate metrics that are to be used for full-buffer performance comparison,25
the first using geometric mean (GM) and the second using harmonic mean (HM). We refer26 to these as GMM (geometric mean method) and HMM (harmonic mean method),27
respectively. The comparison methods under GMM and HMM are identical, other than the28
specific metric computation. When using GMM, network resources should be allocated to29
optimize the GM, and when using HMM they should be allocated to optimize the HM.30
Hence, a separate simulation run is performed for the GMM and HMM comparisons.31
Let rbe the vector of throughputs from the simulation run of the full-buffer MSs in the32network, let S be the set of sectors, and let sM be the set of full-buffer MSs for which33
sector s is the serving sector, and which are not in outage. The GMM metric is computed34
as35
Ss
N
Mi
is
t
s
s
rNN
rsec
GMM1:)(U 2.1-836
where tNsec is the number of sectors in the network (57 in the standard layout), and sN is37
the number of mobiles in sM . The HMM metric is computed as38
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
36/266
3GPP2 C.R1002-B v1.0
16
Ss
Mi is
s
t
srN
NN
r11
11:)(U
sec
HMM 2.1-91
The following steps are to be performed for the full-buffer throughput comparison:2
1) Determine if GMM or HMM will be used3
2) Run the simulation (FL or RL, as desired), which includes some full-buffer users,4
optimizing resource allocation as appropriate for GMM or HMM5
3) From the full-buffer throughput rvector determine )(UGMM r or )(UHMM r using6
the appropriate equation above7
4) Report the )(UGMM r or )(UHMM r value as appropriate for comparison across8
proposals9
5) Report MS Relative throughput and average sector throughput as well10
11
2.1.7 C/I Predictor Model for System Simulation12
Each company shall use their own prediction methodology and describe the prediction13
method in enough detail so other companies can replicate the simulations. This shall14
include the timing diagram from measurements at the mobile to scheduling decisions at the15
base station based on those measurements. Furthermore, this delay shall be explicitly16
modeled in the system level simulator.17
2.2 Link Level Modeling18
2.2.1 Link to System FER mapping19
The performance characteristics of individual links used in the system simulation are20
generated a priori from link level simulations. Link level simulation parameters are21
specified in Appendix J.22
Turbo Decoder Metric and Soft Value Generation into Turbo Decoder shall be as specified23
in Appendix H.24
The quasi-static approach with fudge factors or with short term FER shall be used to25
generate the frame erasures for both the 1xEV-DV packet data channel and the 1xEV-DO26
Forward Traffic Channel (FTC), dynamically simulated forward link overhead channels,27
voice and SCH (applicable only to 1xEV-DV), as described below. Equivalent SNR approach28
shall be used to generate the frame erasures for the OFDM-based forward traffic channel29
(FTC) in 1xEV-DO BCMCS, as described below.30
If the BCMCS proposal uses a Reed Solomon (RS) outer code, then the frame erasures31
generated above constitute the erasure events at the input to the outer code. This is used to32
generate packet erasures at the output of the RS decoder as follows:33
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
37/266
3GPP2 C.R1002-B v1.0
17
Suppose the RS code is specified by the ordered triple (N, K, R), where N is number of octets1
in a RS code word, K is the number of data octets in a RS code word, and R=N-K is the2
number of parity octets in a RS code word. Each code word of the RS code contains a data3
octet from N consecutive turbo coded packets. A set of N consecutive turbo coded packets4
(at the input to the RS decoder) spanned by a RS code word is referred to as an error5
control block. An error control block contains K data packets and R parity packets. If an6error control block contains at most R packet erasures, then the RS outer code recovers all7
erasures in the error control block, and all K data packets in the error control block are8
successfully passed on to the higher layer. If the error control block contains more than R9
packet erasures, then the unerased data packets in the error control block are passed on to10
the higher layer, while the remaining data packets constitute erasure events at the output11
of the RS outer code.12
13
2.2.1.1 Quasi-Static Approach with Fudge Factors:14
Quasi-static approach with fudge factors shall be used for 1xEV-DV Packet Data Channel,15 1xEV-DO FTC, and Dynamically Simulated Forward Link Overhead Channel.16
The aggregated Es/Nt is computed over a transmission period and mapped to an FER using17
AWGN curves. The proponent shall select one of two possible methods to determine the18
FER:19
a) Map the aggregated Es/Nt directly to the AWGN curve corresponding to the given20
modulation and coding.21
b) Adjust the aggregated Es/Nt for the given modulation and coding and lookup a22
curve obtained using a reference modulation and coding.23
Furthermore the proponents shall account for an additional Es/Nt loss at higher Dopplers24
for either method.25
Full details of the quasi-static frame error modeling with fudge factors are given in the26
Appendix F.27
2.2.1.2 Quasi-Static Approach with Short Term FER:28
The quasi-static approach with short term FER may be used to generate frame erasures for29
voice, SCH, and F-PDCH for 1xEV-DV systems. The quasi-static approach with short term30
FER may be used to generate frame (i.e. physical-layer packet) erasures for the Forward31
Traffic Channel (FTC) for 1xEV-DO systems.32
A full set of short term FER vs. average Eb/Nt per frame curves is generated as a function33 of radio configurations, transmission diversity schemes (if applicable), channel models,34
different ways of soft hand-off (SHO), different SHO imbalances, and geometries. The35
number of curves should be reduced if possible, provided that this wont unduly affect the36
validity of this quasi-static approach.37
All companies shall use the same set of short term FER vs. average Eb/Nt per frame.38
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
38/266
3GPP2 C.R1002-B v1.0
18
In the system-level simulation, the average Eb/Nt per frame is computed as follows. First,1
the average Eb/Nt is calculated in a PCG (slot). The short-term average Eb/Nt per frame is2defined as the average of the average Eb/Nt for all Ns PCGs (slots) in a frame (physical3
layer packet), i.e.,4
sN
n nt
b
st
b
N
E
NN
E
1
1 2.2-15
where (Eb/Nt)n is the average Eb/Nt in the n-th PCG (slot) in a frame (physical-layer6
packet). Note that finger combining and self-interference are applied before the average7
Eb/Nt in a PCG (slot) is calculated. Once the Eb/Nt is calculated as in the above equation,8
it is used to look up the corresponding link level short term FER vs. average Eb/Nt per9
frame curves for the specific condition (i.e., radio configuration, transmission diversity (if10
applicable) scheme, channel model, way of soft hand-off (SHO), SHO imbalance(s), and11
geometry). A frame erasure event is then generated based on the FER value.12
If a short term FER vs. average Eb/Nt per frame curve is not available for a condition, the13
curve should be computed by interpolating those curves for similar conditions (e.g.,14between the factors for closest geometries available).15
The short term FER vs. average Eb/Nt per frame curves shall be generated as follows:16
1. The link-level simulation is conducted for a specific condition. The average Eb/Nt in a17
frame and the frame erasure indicator for the frame are recorded. For 1xEV-DV18
systems, the average Eb/Nt per frame is computed as follows in the link-level19
simulation20
sN
n
k
kn
t
k
kn
b
t
b
n
Sm
N
E
12),(
2),(
16
1 2.2-221
where n is the index of PCG in a frame and k is the index of symbols within a PCG.22),( kn
bS is the signal component in the k-th received coded symbol in the n-th PCG,),( kn
tn 23
is the noise and interference component in the k-th received symbol in the n-th PCG in24
a frame, and mis the inverse of the code rate (i.e., 4 for RC3 and 2 for RC4, etc). For25
1xEV-DO systems, the average Eb/Nt per slot is computed as follows in the link-level26
simulation27
K
k
L
llkn
t
lkn
s
n
t
b
N
E
MN
E
1 1),,(
),,(1
)( 2.2-328
where M equals to the number of information bits per packet; n is the slot index and29k is the symbol index within a slot; l is the path index, where the total number of30
captured paths is denoted by L; the total number of symbols per slot is K;31),,( lkn
sE denotes the signal energy in the k-th received symbol in the n-th slot in the32
physical-layer packet at the l-th RAKE finger, and),,( lkn
tN denotes the noise and33
interference variance in the k-th received symbol in the n-th slot in the physical-34
layer packet seen at the l-th RAKE finger.35
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
39/266
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
40/266
3GPP2 C.R1002-B v1.0
20
the percentage of users with that channel model in each sector. The JTC fader (see1
Appendix K) shall be used to generate the Jakes fading samples.2
For CDM transmissions with Rake demodulation, each multipath model (Pedestrian A,3
Vehicular A/B etc) is characterized in terms of the number of Rake fingers (resolvable4
paths), the Delay and Fractional Recovered Power (FRP) of each finger, and the Fractional5
UnRecovered Power (FURP). The FRP and FURP are given in Table 2.2-2. FURP shall6contribute to the interference of the finger demodulator outputs as an independent fader.7
The power on all fingers (including FURP) for each channel model shall be normalized so8
that the total power for that channel model adds up to unit one.9
Table 2.2-2 Fractional Recovered Power and Fractional UnRecovered Power10
Model Finger1
(dB)
Delay Finger2
(dB)
Delay (Tc) Finger3
(dB)
Delay (Tc) FURP (dB)
Ped-A -0.06 0.0 -18.8606
Ped-B -1.64 0.0 -7.8 1.23 -11.7 2.83 -10.9151Veh-A -0.9 0.0 -10.3 1.23 -10.2759
11
The delay values given inTable 2.2-2 are for information purposes and do not need to be12
accounted for in the system simulation.13
For OFDM transmission and demodulation of 1xEV-DO BCMCS, each multipath model14
(Pedestrian A, Vehicular A/B etc) is characterized in terms of the total number of paths,15
together with actual power-delay profile of the multipath channel. For each multipath16
model, the power on all paths shall be normalized so that the total power adds up to one.17
The parameters of the multipath models for OFDM transmission and demodulation are18
given inTable 2.2-3 andTable 2.2-4.19
Table 2.2-3 Relative Power of each Multipath Model (in dB)20
Model # Paths 1 2 3 4 5 6
Ped-A 4 0 -9.7 -19.2 -22.8
Ped B 6 0 -0.9 -4.9 -8.0 -7.8 -23.9
Veh-A 6 0 -1.0 -9.0 -10.0 -15.0 -20.0
21
Table 2.2-4 Delay of each Multipath Model (in ns)22
Model # Paths 1 2 3 4 5 6
Ped-A 4 0 110 190 410
Ped B 6 0 200 800 1200 2300 3700
Veh-A 6 0 310 710 1090 1730 2510
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
41/266
3GPP2 C.R1002-B v1.0
21
The channel between the serving sector(s) and the subscriber is modeled using the1
multipath profiles defined above. The channel between any interfering sector and the2
subscriber can be modeled as a one-path Rayleigh fading channel, where the Doppler of the3
fading process is randomly chosen based on the velocities and its corresponding4
probabilities specified inTable 2.2-1.5
2.2.2.2 Channels model based on SCM6
2.2.2.2.1 Channel model for system level simulations7
The 3GPP2/3GPP spatial channel model (SCM) [41] shall be used for all system level8
simulations. The Urban Macro-cellular environment is mandatory and the parameters of9
Table 2.2-5 shall be used for configuring the model.10
Table 2.2-5 Macro-cellular Environment Parameters11
Channel Scenario Urban Macro
Number of paths (N) 6
Number of sub-paths (M) per-path 20
Mean AS at BS E( AS )=15 deg
AS at BS as a lognormal RV
10 ^ , ~ (0,1)AS AS ASx x
15deg
AS = 1.18, AS = 0.210
ASAoDASr / 1.3
Per-path AS at BS (Fixed) 2 deg
BS per-path AoD Distribution standarddistribution
),0( 2AoD where ASASAoD r
Mean AS at MS E(AS, MS)=68 deg
Per-path AS at MS (fixed) 35 deg
MS Per-path AoA Distribution (Pr)),0( 2AoA
Delay spread as a lognormal RV
10 ^ , ~ (0,1)DS DS DSx x
DS = -6.18
DS = 0.18
Mean total RMS Delay Spread E( DS
)=0.65 s
DSdelaysDSr / 1.7
Lognormal shadowing standard deviation,
SF
8.9dB
Pathloss model (dB), d is in meters 28.6 + 35log10(d)
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
42/266
3GPP2 C.R1002-B v1.0
22
The velocity profile is shown in Table 2.2-6. Because of the choice of urban macrocell,1
velocities are biased towards pedestrian speeds.2
3
Table 2.2-6 Quantized Velocity Profile4
Percentage Velocity (km/h)
35% 3
30% 30
20% 60
15% 120
The carrier frequency for all simulations is assumed to be 2.0 GHz.5
2.2.2.2.2 Channel model for link level simulations6
For link-level simulations, that excludes link to system mapping simulations, spatially7
extended ITU profiles will be used.8
Table 2.2-7 ITU Profiles for Link Level Simulations9
ITU Model Velocity (km/h)
AWGN 0
Ped-A 3
Ped-B {3,30}
Veh-A {30,120}For technologies that use a cyclic prefix, the path location will be equal to the closest10
integer sample number, i.e. no FURP modeling is needed.11
12
Table 2.2-8 ITU Profiles Spatial Extension Parameters13
Channel Scenario Urban Macro
AS at BS AS =150
Per-path AS at BS (Fixed) 2 deg
AS at MS AS, MS=680
Per-path AS at MS (fixed) 350
AoDs As specified inTable 2.2-9
AoAs As specified inTable 2.2-9
14
8/12/2019 C.R1002-B v1.0 Evaluation Methodology-IncludeChannelModel
43/266
3GPP2 C.R1002-B v1.0
23
Table 2.2-9 Path power, AoD, AoA1
Path Power Path AOD (rad) Path AoA (rad)
Ped-A 0.889345301 0.346314033 1.737577272
0.095295066 -0.05257642 -1.556450.010692282 -1.817837659 -1.049078459
0.00466735 -0.836999548 0.345571431
Ped-B 0.405688403 -0.13638548 1.319340881
0.329755914 0.302249557 -0.119072067
0.131278194 0.496051618 0.901442565
0.064297279 0.544719913 -1.424448314
0.067327516 0.212670549 -3.062670939
0.001652695 -0.604134536 -1.202289294
Veh-A 0.48500285 -0.46084874 -0.780118399
0.385251458 -0.897480352 -1.729577654
0.061058241 -0.525726742 1.792547973
0.048500285 0.00282531 1.776985779
0.015337137 -1.016095677 1.386034573
0.004850029 0.245512493 3.50389557
The fading coefficient generation will be the one described in section 5.3 of the23GPP2/3GPP SCM [41].3
2.2.2.2.3 Channel model for virtual decoder generati