D4.2 Final conclusions on end-to-end performance and sensitivity analysis

(1)

Project Number: CELTIC / CP5-026

Project Title: Wireless World Initiative New Radio – WINNER+

Document Type: P (Public)

Document Identifier: D4.2

Document Title:

D4.2 Final conclusions on end-to-end performance

and sensitivity analysis

Source Activity: WP4

Editor: Marc Werner

Authors: Jorge Cabrejas, Valeria D’Amico, David Martín-Sacristán, Jose F. Monserrat, Ahmed Saadani,Thierry Clessienne, Pascal Chauveau, Krystian Safjan, Daniel Bültmann, Johan Nyström, Per Skillermark, Jouko Leinonen, Krzysztof Bąkowski, Marc Werner

Status / Version: final version / v1.0

Date Last changes: 30.06.2010

File Name: D4.2.doc

Abstract: This document contains an end-to-end performance analysis of the

LTE-Advanced system as submitted as IMT-Advanced candidate technology to ITU-R. WINNER+ has evaluated this technology proposal as an official ITU External Evaluation Group. This deliverable describes the calibration, system configuration, and deployment environments that were used by WP4 for assessing LTE-Advanced. The set of simulation results given in this document provides an extensive insight into the achieved end-to-end system performance. The final conclusion of WINNER+ is that the LTE-Advanced proposal meets all the IMT-LTE-Advanced performance requirements. A sensitivity analysis of system performance with respect to selected system parameters is presented.

Furthermore, this document also provides end-to-end performance results for complementary deployment scenarios, as well as spectral efficiency results for the 802.16-based IMT-Advanced proposal.

Keywords: End-to-end performance evaluation, system-level simulation,

link-level simulation, calibration, IMT-Advanced, LTE-Advanced, ITU-R WP5D, Evaluation Group, Evaluation Report

Document History:

2010-05-17 Document created by Marc Werner, based on the chapter and editor assignments discussed at Valencia meeting.

2010-06-02 Preliminary version finalized for submission to CELTIC review.

(2)

Channel Model Calibration ... 18

4.1 Calibration of large scale parameters ... 18

4.1.1 Path gain and wideband SINR for InH ... 19

4.1.2 Path gain and wideband SINR for UMi ... 19

4.1.3 Path gain and wideband SINR for UMa ... 19

4.1.4 Path gain and wideband SINR for RMa ... 20

4.1.5 Comparison of calibration results ... 20

4.2 Calibration of small scale parameters ... 20

4.2.1 Small-scale fading in InH ... 21

4.2.2 Small-scale fading in UMi ... 22

4.2.3 Small-scale fading in UMa ... 23

4.2.4 Small-scale fading in RMa... 24

5

LTE Rel-8 Baseline Configuration & Calibration ... 25

5.1 Calibration results in uplink ... 25

5.2 Calibration results in downlink ... 28

5.3 Quality of calibration – quantitative assessment ... 30

6

Simulation Results ... 31

6.1 Cell and cell edge spectral efficiency ... 31

6.1.1 Summary of results ... 31

6.1.2 Simulation assumptions & detailed results for organization 1 ... 35

6.1.4 Simulation assumptions for organization 3 ... 37

6.1.5 Simulation Assumptions for organization 4 ... 38

6.2 Mobility ... 41

6.2.2 Organization 1 system and simulator setup ... 41

(3)

7

Sensitivity Analysis ... 47

7.1 Introduction to sensitivity analysis... 47

7.2 Goals of WINNER+ sensitivity analysis ... 47

7.3 Scope of analysis ... 47

7.4 Outlook on the wireless network model sensitivity analysis process ... 48

7.5 Methodology used for sensitivity analysis ... 49

7.6 Conclusions ... 50

8

Conclusions ... 51

9

Annex: Complementary Simulation Results ... 52

9.1 Simulation results for organization 3 ... 52

9.2 Simulation results for organization 8 ... 52

9.3 Complementary evaluation of use case proposed by TCOE India ... 53

9.3.1 Introduction ... 53

9.3.2 Models and assumptions ... 53

9.3.3 Results... 53

9.3.4 Summary ... 56

9.4 Complementary simulation results for IEEE 802.16 based RIT ... 56

9.4.1 Introduction ... 56

9.4.2 Models and assumptions ... 56

9.4.3 Results... 57

9.4.4 Summary ... 58

(4)

Authors

Partner

Name

Phone / Fax / e-mail

EAB Per Skillermark Phone: +46 10 7131922

Fax: +46 10 7172092

e-mail: [email protected]

EAB Johan Nyström Phone: +46 10 7170586

Fax: +46 10 7131480

e-mail: johan.nystrom @ericsson.com

iTEAM Jorge Cabrejas Phone: +34 963877007

Fax:+34 963879583

iTEAM Jose F. Monserrat Phone: +34 963877007

Fax:+34 963879583

iTEAM David Martín-Sacristán Phone: +34 963877007

Fax:+34 963879583

NSNP Krystian Safjan Phone: +48 728 361 374

Fax: +48 71 777 3873

RWTH Daniel Bültmann Phone: +49 241 80 27918

Fax: +49 241 80 22242

TILab Valeria D’Amico Phone: +39 011 2287544

Fax: +39 011 2285224

(5)

QUA Marc Werner Phone: +49 911 54013130 Fax: +49 911 54013190

FT Ahmed Saadani Phone: +33 1 45 29 54 05

Fax: +33 1 45 29 41 94

FT Thierry Clessienne Phone: +33 1 45 29 48 81

Fax: +33 1 45 29 41 94

FT Pascal Chauveau Phone: +33 1 45 29 44 39

Fax: +33 1 45 29 41 94

UOULU Jouko Leinonen Phone: +358 8 553 2968

Fax: +358 8 553 2845

PUT Krzysztof Bąkowski Phone: +48 61 655 3936

Fax: +48 61 655 3930

(6)

Executive Summary

This document contains an end-to-end performance analysis of the LTE-Advanced system as submitted as IMT-Advanced candidate technology to ITU-R. WINNER+ has evaluated this technology proposal as an official ITU-R External Evaluation Group. This deliverable first outlines LTE-Advanced features, as well as the IMT-Advanced evaluation scenarios and performance criteria. It then describes the calibration, system configuration, and deployment environments that were used by WP4 for assessing LTE-Advanced. The set of simulation results given in this document provides an extensive insight into the achieved end-to-end system performance.

Among WINNER+ WP4 organizations, the channel model calibration in the four IMT-Advanced deployment scenarios Indoor Hotspot (InH), Urban micro-cell (UMi), Urban macro-cell (UMa), and Rural macro-cell (RMa), was successfully carried out as a first step towards a consolidated set of simulation tools. The next step was the system performance calibration for a selected LTE Rel-8 baseline configuration. Due to simulator differences, a small variation in performance results was observed as expected.

The LTE-Advanced simulations themselves were carried out for a set of different system configurations (e.g. MIMO schemes and schedulers), all meeting the IMT-Advanced requirements. In parallel, WINNER+ WP3 has established that all the “analytical” and “inspection” performance criteria were also met. Therefore the final conclusion of WINNER+ is that the LTE-Advanced proposal meets all the IMT-Advanced performance requirements, and hence should qualify as an IMT-IMT-Advanced system.

In the analysis of LTE-Advanced, this document goes beyond the core results that were submitted by WINNER+ in the IMT-Advanced Evaluation Report [5D/769]. Additional simulation results for a wider set of assumptions are given, and the results are supported by a larger amount of simulation data (e.g., throughput distribution curves). The document also provides a sensitivity analysis of system performance with respect to a selected set of simulation assumptions..

(7)

List of Acronyms and Abbreviations

3GPP 3rd Generation Partnership Project

ARRM Advanced Radio Resource Management

AWGN Additive White Gaussian Noise

BCH Broadcast Channel

BER Bit Error Rate

B-LDPCC Block-Type Low-Density Parity Check Code

BLER Block Error Rate

BS Base Station

C/I Carrier to Interference (Ratio)

CAS Coordinated Antenna Systems

CDF Cumulative Distribution Function

CoMP Coordinated MultiPoint

CQI Channel Quality Indicator

DL Downlink

EEG External Evaluation Group

EESM Exponential Effective SIR Mapping

E-UTRA Evolved-UMTS Terrestrial Radio Access

FDD Frequency Division Duplex

FER Frame Error Rate

H-ARQ Hybrid Automatic Repeat Request

IAT Inter Arrival Time

IEG Israeli Evaluation Group

IMS IP Multimedia Subsystem

IMT International Mobile Telecommunications

IMT-A International Mobile Telecommunications – Advanced

IP Internet Protocol

ISD Inter Site Distance

ITU International Telecommunication Union

ITU-R International Telecommunication Union - Radiocommunications

KPI Key Performance Indicator

LA Link Adaptation

LMS Least Mean Squares

LOS Line of Sight

LTE Long Term Evolution

LTE-A Long Term Evolution - Advanced

MAC Medium Access Control

MCS Modulation and Coding Scheme

MIESM Mutual Information Effective SINR Mapping

MIMO Multiple-Input Multiple-Output

NLOS Non Line of Sight

OFDM Orthogonal Frequency-Division Multiplexing

OAT One Factor at Time

PCH Paging Channel

PDU Packet Data Unit

PER Packet Error Rate

(8)

PFS Proportional Fair Scheduler

PRB Physical Resource Block

PS Packet Size

P-SCH Primary-SCH

QAM Quadrature Amplitude Modulation

QoS Quality of Service

QPSK Quadrature Phase Shift Keying

RACH Random Access Chanel

RAN1 Radio Access Network (group 1)

RAP Radio Access Point

RAT Radio Access Technology

RIT Radio Interface Technology

RLC Radio Link Control

RN Relay Node

RP Resource Partitioning

RRM Radio Resource Management

RSCP Reference Signal Received Power

RX Receive

SCH Synchronization Channel

SINR Signal to Interference plus Noise Ratio

SNR Signal to Noise Ratio

S-PARC Selective Per Antenna Rate Control SRIT Set of Radio Interface Technologies S-SCH Secondary Synchronization Channel TCOE Telecom Centres of Excellence (India)

TDD Time Division Duplex

Tdoc Technical Document

TS Technical Specification

TX Transmit

UE User Equipment

UL Uplink

UMTS Universal Mobile Telecommunications System

UT User Terminal

VoIP Voice over IP

(9)

1

Introduction

This document summarizes the performance assessment activities carried out in Year 2 of the WINNER+ project in WP4. In this sense, it continues the preparatory work presented in IR4.2 “Interim conclusions on end-to-end performance and sensitivity analysis based on Y1 results”.

The end-to-end performance assessment work in WP4 has been concentrated on system- and link level simulations of LTE-Advanced in the context of the IMT-Advanced candidate proposal assessment which WINNER+ has carried out as an official external evaluation group for ITU-R.

In WP4, the simulation focus has been placed on the following performance requirements:

• Cell spectral efficiency,

• Cell edge spectral efficiency,

• Mobility,

• VoIP capacity.

The simulation results presented in this deliverable contain the results for LTE-Advanced that were officially submitted to ITU-R in the WINNER+ Evaluation Report [5D/769]. Beyond that, this document contains additional simulation results of LTE-Advanced system which go beyond the studied configurations for ITU.

In Chapter 2, the LTE-Advanced system is shortly characterized by outlining its main improvements over LTE Rel-8. Chapter 3 reflects on the IMT-Advanced evaluation scenarios and performance criteria that had been set up by ITU-R WP 5D for the IMT-Advanced evaluation activity.

The simulator calibration activities of WP4 are described in Chapters 4 and 5. Chapter 4 presents the results of the IMT-Advanced channel model calibration, while in Chapter 5, the performance results for a baseline LTE Rel-8 system by different contributing partners are described.

Chapter 6 then presents the core results of the simulative assessment of the LTE-Advanced system in different deployment scenarios, for TDD and FDD as well as for Uplink and Downlink. Different system configurations were studied e.g. with respect to multi-antenna techniques.

Chapter 7 provides a short sensitivity analysis of a selected subset of simulation results for a number of system parameters. In Chapter 8, the main conclusions of the deliverable are summarized, and in Chapter 9, an Annex with complementary simulation results (e.g. for the 802.16 base IMT-Advanced proposal is given.

Since the contributions of different WINNER+ organizations were presented in an anonymous way in the WINNER+ Evaluation Report to ITU-R, anonymity has been preserved in the present public deliverable as well.

(10)

2

LTE-Advanced Features

In 2008, 3GPP held two “3GPP IMT-Advanced Workshops” [WS1], [WS2]. The goal of these workshops was to investigate what were the main changes that could be brought forward to evolve the eUTRA Radio Interface as well as the eUTRAN in the context of IMT-Advanced.

In particular, the LTE-Advanced Study Item (SI) was initialized in order to study the evolution of Long Term Evolution (LTE), based on a new set of requirements. This initiative has been collecting operators’ and manufacturers’ views in order to develop and test innovative concepts that will satisfy the needs of next-generation communications. The requirements have been gathered in 3GPP TR 36.913 “Requirements for Evolved UTRA (E-UTRA) and Evolved UTRAN (E-UTRAN)” [36.913]. The resulting Technical Report was published in June 2008 and a liaison was sent to ITU-R covering the work in 3GPP RAN on LTE-Advanced towards IMT-Advanced. Finally 3GPP has contributed to the ITU-R towards IMT-Advanced via its proposal “3GPP LTE Release 10 & Beyond (LTE-Advanced)” [IMT-ADV/8].

The new technical features of LTE-Advanced are defined in 3GPP TR 36.814 “Further advancements for E-UTRA physical layer aspects” [36.814]. The main technical features under discussion are the following:

• Support of wider bandwidth

Carrier aggregation, where two or more component carriers, each with a bandwidth up to 20 MHz, are aggregated, is considered for LTE-Advanced in order to support downlink transmission bandwidths larger than 20 MHz, e.g. up to 100 MHz. A terminal may simultaneously receive or transmit one or multiple component carriers depending on its capabilities.

• Extended multi-antenna configurations

Extension of LTE downlink spatial multiplexing to up to eight layers is considered. For the uplink spatial multiplexing to up to four layers is considered.

• Coordinated Multiple Point (CoMP) transmission and reception

This feature is considered as a tool to improve the coverage of high data rates, the cell-edge throughput and/or to increase system throughput

• Relaying functionality

Relaying is considered for LTE-Advanced as a tool to improve e.g. the coverage of high data rates, group mobility, temporary network deployment, the cell-edge throughput and/or to provide coverage in new areas.

In particular, in the following subsections we will provide details on the LTE-Advanced features that are relevant for the assessment of the 3GPP proposal as candidate IMT-Advanced technology also highlighting the approach followed by the WINNER+ Evaluation Group.

2.1

Focus on LTE-Advanced features for IMT-Advanced assessment

As mentioned above, 3GPP has been working on Advanced since early 2008. In June 2008 the LTE-Advanced targets were set and the dedicated Study Item was created. Subsequently in December 2009 the following Work Items were created: Carrier aggregation, Enhanced DL multi-antenna transmission, UL Multi antenna transmission, and Relaying. In March 2010 the LTE-Advanced Study Item was closed and a Work Item on extended Inter-Cell Interference Coordination (eICIC) for co-channel deployments of heterogeneous networks was started. It should be noted that the Study Item on Coordinated Multiple Point was placed on hold until September 2010. The first decisions have been taken and will form the basis for LTE-Advanced standardization in Release 10 that are being reflected in the Technical Report 36.814 [36.814] that is still to be finalized.

In the following we will focus on the LTE-A features that are relevant for the assessment of the 3GPP proposal as candidate IMT-Advanced technology.

For the uplink direction, LTE-Advanced supports spatial multiplexing of up to four layers. In uplink single user spatial multiplexing, up to two transport blocks can be transmitted from a User Terminal (UE) in a subframe per uplink component carrier and each transport block has its own Modulation and Coding Scheme (MCS). Based on the number of transmission layers, the modulation symbols of each transport block are mapped onto one or two layers similar as in LTE Rel-8 downlink spatial multiplexing. Moreover the transmission rank can be adapted dynamically. The uplink single user spatial multiplexing transmission can be configured with or without the layer shifting. In case of the layer shifting, shifting in time domain is supported. For FDD and TDD, precoding is performed using predefined codebooks. For uplink spatial multiplexing with four transmit antennas, a 6-bit precoding codebook is used. The employed precoding codebooks are defined in [36.814].

(11)

An uplink single antenna port mode is defined for UEs with multiple transmit antennas. From eNB’s perspective, the UE behavior is then the same as the one with single antenna. The uplink single antenna port mode is the default mode before eNB is aware of the UE transmit antenna configuration.

For the downlink direction, LTE-Advanced supports spatial multiplexing of up to eight layers. In the downlink 8-by-X single user spatial multiplexing, up to two transport blocks can be transmitted to a UE in a subframe per downlink component carrier. Each transport block is assigned its own MCS. A transport block is associated with a codeword. The codeword-to-layer mapping is defined in [36.211] for up to four layers.

LTE-Advanced also supports enhanced Multi-User MIMO (MU-MIMO) transmission. MU-MIMO in LTE-Advanced is that it allows the switching between SU- and MU-MIMO transmission without the need of RRC reconfiguration. The adopted transmission schemes for MU-MIMO are vendor specific and hence implementation dependent.

CoMP transmission/reception is considered for LTE-Advanced as a tool to improve the coverage of high data rates, the cell-edge throughput and/or to increase system throughput. Downlink CoMP transmission implies dynamic coordination among multiple geographically separated transmission points. The serving cell is defined as the cell transmitting PDCCH assignments. This is the serving cell of LTE Release 8. 3GPP currently considers the following two CoMP categories:

• Joint Processing (JP) in which data is available at each point of the CoMP cooperating set. It can be based on:

− Joint Transmission: transmission from multiple points at a time to a single UE to improve the received signal quality and/or cancel actively interference for other UEs.

− Dynamic cell selection: transmission from one point at a time.

• Coordinated Scheduling/Beamforming (CS/CB): data is only available at serving cell (data transmission from that point) but user scheduling/beamforming decisions are made with coordination among cells corresponding to the CoMP cooperating set.

It should be noted that a CoMP cooperating set is a set of (geographically separated) points directly or indirectly participating in transmission towards the UE. CoMP transmission point(s) is a point or set of points actively transmitting towards the UE, thus it is a subset of the CoMP cooperating set. In the JP category for joint transmission the CoMP transmission points are the points in the CoMP cooperating set, whereas for dynamic cell selection a single point is the transmission point at every subframe. This transmission point can change dynamically within the CoMP cooperating set. For CS/CB, the CoMP transmission point corresponds to the serving cell.

Downlink CoMP transmission should include the possibility of coordination between different cells. From a radio-interface perspective, there is no difference from the UE if the cells belong to the same eNB or different eNBs. If inter-eNB coordination is supported, information needs to be signalled between eNBs.

The potential impact which CoMP transmission has on the radio-interface specifications is foreseen in mainly three areas. Firstly, feedback and measurement mechanisms from the UE are involved, so to appropriately select the set of participating transmission points and also dynamically report the channel conditions between the multiple transmission points and the UE . Secondly, preprocessing schemes should be defined, in which joint processing prior to transmission of the signal over the multiple transmission points should be considered together with downlink control signalling to support the transmission scheme. Thirdly, reference signal design is involved, where, depending on the transmission scheme, specification of additional reference signals may be required.

For the time being, for Release 10, there will be no new standardised X2 interface communication for support of inter-eNB CoMP, therefore no additional features are currently specified in Rel-10 to support downlink (DL) CoMP. As mentioned above, the Study Item on DL-CoMP was approved but is on hold until September 2010.

Uplink CoMP reception is expected to have very limited impact on the RAN1 specifications. Uplink CoMP reception can involve joint reception (JR) of the transmitted signal at multiple reception points and/or coordinated scheduling (CS) decisions among cells to control interference and may have some RAN1 specification impact.

Relaying is considered for LTE-Advanced as a tool to improve e.g. the coverage of high data rates, group mobility, temporary network deployment, the cell-edge throughput and/or to provide coverage in new areas. Relay nodes are placed throughout the macro-cell layout, hence modifying the reference layout specified in [M.2135]. Moreover the channel model to be used to model relay backhauling transmission link was not defined in [M.2135]. For these reasons relay nodes have not been considered as advanced feature to be used when assessing IMT-Advanced requirements.

(12)

2.2

The WINNER+ Evaluation Group approach

The WINNER+ project has produced consistent research work on optimisation of the radio interface concepts for IMT-Advanced systems, based on the heritage of the activities carried out in the former European Union Framework Program 6 project WINNER. In addition to developing enabling technologies for the WINNER+ system concept, in November 2008, WINNER+ has registered as an Independent Evaluation Group towards ITU-R in order to drive the International Mobile Telecommunications (IMT)-Advanced development process.

In particular, the minimum requirements for IMT-Advanced Radio Interface Technologies (RITs) have been identified and are summarized in the document [M.2134] which contains the detailed assumptions and the required values or features of IMT-Advanced relating to technical performance. The requirements have different values for different test environments. Guidelines for evaluation of radio interface technologies for IMT-Advanced containing the detailed simulation assumptions and the evaluation methodologies of IMT-Advanced are specified in [M.2135]. The evaluation procedure is designed in such a way that the overall performance of the candidate RITs/SRITs may be fairly and equally assessed on a technical basis.

Based on the available expertise in IMT-Advanced radio technology concepts and in link- and system-level simulation tools, the WINNER+ Evaluation Group has targeted the 3GPP LTE Release 10 & Beyond (LTE-Advanced) proposal [5D/769]. The WINNER+ group has evaluated all 13 minimum requirements for IMT-Advanced systems according to [M.2134] by means of analytical, inspection and simulation activities in order to perform a full evaluation of the LTE-Advanced candidate technology. For simulation purposes, in order to guarantee the reliability of the results, evaluated characteristics have been assessed by a plurality of partners. During the course of the work, great emphasis has been given to reflect a realistic behavior of the system under consideration, by modeling non-ideal aspects including, e.g., effects of channel estimation errors, CQI measurement errors and feedback delay as well as a correct modeling of the overhead in the system. Simulators of different partner organizations have been calibrated in order to provide consistent results.

The adopted calibration approach together with detailed calibration results are provided in the following chapters of this document. Simulation results, provided in subsequent chapters, have confirmed that the 3GPP LTE Release 10 & Beyond (LTE-Advanced) proposal satisfies all the IMT-Advanced requirements.

(13)

3

IMT-Advanced Evaluation Scenarios & Performance Criteria

According to the evaluation process of ITU-R, IMT-Advanced candidate proposals need to fulfil a set of 13 minimum performance requirements. Some of them shall be assessed analytically, some by inspection and some by simulation. For the latter the ITU-R gives detailed Guidelines for evaluation of radio

interface technologies for IMT-Advanced [M.2135] to ensure comparable simulation results across

evaluation groups. Minimum requirements need to be met in four specific test environments that reflect future use cases of IMT-Advanced systems. Each environment is associated with a deployment scenario that specifies the simulation setup, e.g. inter-site distance, carrier frequency, maximum transmit powers, channel model, etc. This chapter gives an overview of the evaluation scenarios that were used for the simulations and lists the performance criteria imposed by the ITU-R on candidate radio technologies for IMT-Advanced.

3.1

Evaluation scenarios

Evaluation of candidate RIT/SRITs was performed in selected scenarios of the following test environments:

• Indoor: an indoor environment targeting isolated cells at offices and/or in hotspot based on stationary and pedestrian users.

• Microcellular: an urban micro-cellular environment with higher user density focusing on pedestrian and slow vehicular users.

• Base coverage urban: an urban macro-cellular environment targeting continuous coverage for pedestrian up to fast vehicular users.

• High speed: macro cells environment with high speed vehicular and trains.

For each of these test environments deployment scenarios were defined to be used for the performance evaluation of candidate RIT/SRIT (see Table 3.1 and Figure 3.1).

Table 3.1: Deployment Scenarios

Test environment Indoor Microcellular Base covergae urban High Speed Deployment Scenario Indoor Hotspot (InH) Urban micro-cell (UMi) Urban macro-cell (UMa) Rural macro-cell (RMa)

Figure 3.1: Deployment scenarios for Indoor Hotspot scenario (left) and for all other scenarios (middle/right) [M.2135]

Each of the test environments focuses on a specific application for the candidate RIT/SRITs and is accompanied by specific values of the performance criteria to be met by the RIT/SRITs. The deployment scenarios given in [M.2135] are:

• Indoor Hotspot (InH): Small isolated cells at offices or hotspot areas. Targets high user throughput and high user density. All users are pedestrian. Two base stations operating at 3.4 GHz with omni-directional antenna setup are mounted on the ceiling of a long hall with adjacent offices (cell coverage area 3000 m2).

• Urban Micro-cell (UMi): High traffic and user density for city centers and dense urban areas. Outdoor and outdoor to indoor propagation characteristics for pedestrian users are assumed. Continuous hexagonal deployment is used with 3 sectors per cell and below rooftop antenna

(14)

mounting. Base stations operate at 2.5 GHz and have an inter-site distance of 200 m (cell coverage area 0.14 km²)

• Urban Macro-cell (UMa): Targets ubiquitous coverage for urban areas. A similar hexagonal deployment is used with larger inter-site distance of 500 m and antennas mounted clearly above rooftop. Non line-of-sight or obstructed propagation conditions is common for this scenario. Only vehicular users at moderate speed are assumed, suffering from an additional outdoor to in-car penetration loss. Base stations operate at 2 GHz (cell coverage area 0.86 km²).

• Rural Macro-cell (RMa): Similar to UMa, but targets larger cells with support for high-speed vehicular users. Base stations have an inter-site-distance of 1732 m and operate at 800 MHz which is more suitable for large cells (cell coverage are 10.4 km²).

• Suburban Macro-cell (SMa): This is an optional scenario for the same test environment as of the UMa scenario. The key difference is an increased inter-site-distance of 1299 m and a mix of indoor and high-speed vehicular users (cell coverage area 5.85 km²).

• Rural Indian Open Area: Rural Indian Open Area is a large-cell coverage scenario. Some parameters of the scenario may take several values, e.g. the carrier frequency, the terminal antennas height, the inter-site distance (ISD). Inter Site Distance is 30-50 km what corresponds with typical distance between villages in India. In this scenario terminals are in fixed positions with rooftop directional antennas. Base stations operate at 312 MHz-2300MHz (cell coverage are is up to 1256 km²)

3.2

Performance criteria

In [M.2134], ITU-R specifies minimum performance requirements that candidate radio technologies must fulfil. Table 3.2 lists the set of 13 performance criteria that proponents and evaluation groups should consider in their evaluation reports. The last column gives a reference to the respective evaluation of WINNER+.

Table 3.2: List of IMT-Advanced Performance Criteria [M.2134] Characteristic for evaluation Method Evaluation methodology / configurations Related Section in Report ITU-R M.2135 Related section of Reports ITU-R M.2134 and ITU-R M.2133

Results provided in a)IMT-Advanced evaluation report b) Winner+ Deliverable Cell spectral efficiency Simulation (system level) Report ITU-R M.2135 § 7.1.1, Tables 8-2, 8-4 and 8-5 Report ITU-R M.2134, § 4.1

a)Yes, See Part II-D.2 and Annex C

b)Yes, D4.2 section 6.1 Peak spectral

efficiency

Analytical Report ITU-R M.2135

§ 7.3.1, Table 8-3

Report ITU-R M.2134, § 4.2

a)Yes, See Part II-D.2 and Annex A

b)Yes, D3.4 section 3.1 Bandwidth Inspection Report ITU-R

M.2135 § 7.4.1

b)Yes, D3.4 section 3.2 Cell edge user

spectral efficiency Simulation (system level) Report ITU-R M.2135 § 7.1.2, Tables, 8-2, 8-4 and 8-5 Report ITU-R M.2134, § 4.4

a)Yes, See Part II-D.2 and Annex C

b)Yes, D4.2 section 6.1 Control plane

latency

§ 7.3.2, Table 8-2

Report ITU-R M.2134, § 4.5.1

b)Yes, D3.4 section 3.3 User plane

latency

§ 7.3.3; Table 8-2

Report ITU-R M.2134, § 4.5.2

(15)

Characteristic for evaluation Method Evaluation methodology / configurations Related Section in Report ITU-R M.2135 Related section of Reports ITU-R M.2134 and ITU-R M.2133

Results provided in a)IMT-Advanced evaluation report b) Winner+ Deliverable Mobility Simulation (system and link level) Report ITU-R M.2135 § 7.2, Tables 8-2 and 8-7 Report ITU-R M.2134, § 4.6

a)Yes, See Part II-D.2 and Annex D b) Yes, D4.2 section 6.2 Intra- and inter-frequency handover interruption time

§ 7.3.4, Table 8-2

a)Yes, See Part II-D.2 and Annexes A and B

b)Yes, D3.4 section 3.5

Inter-system handover

Inspection Report ITU-R M.2135 § 7.4.3

b)Yes, D3.4 section 3.6 VoIP capacity Simulation

(system level) Report ITU-R M.2135 § 7.1.3, Tables 8-2, 8-4 and 8-6 Report ITU-R M.2134, § 4.8

a)Yes, See Part II-D.2 and Annex E

b)Yes, D4.2 section 6.3 Deployment

possible in at least one of the identified IMT bands

b)Yes, D3.4 section 3.7

Channel bandwidth scalability

b)Yes, D3.4 section 3.2 Support for a

wide range of services

b)Yes, D3.4 section 3.8 The evaluation criteria are grouped by their evaluation method, which is either inspection, analytical or simulation. Evaluation by inspection only requires evaluation groups to check the candidate proposal if the requirement is addressed and met. Inspection Requirements are:

• Bandwidth: The candidate systems must support scalable bandwidth allocations up to and including 40 MHz. Furthermore, proponents are encouraged to support higher bandwidths of up to 100 MHz. The proponents must demonstrate to support at least three different bandwidths allocation including the minimum and maximum value for the candidate system.

• Inter-system handover: The candidate systems must support inter-system handover between the candidate IMT-Advanced system and least on IMT-2000 system.

• Deployment possible in at least one of the identified IMT bands

• Channel bandwidth scalability

• Support for a wide range of services: Candidate systems must be able to support multiple service classes such as background, streaming, interactive and conversation.

Analytical evaluation involves some calculations to determine whether the candidate reaches the minimum requirement. This group comprises:

• Peak Spectral Efficiency: Gross data rate offered by the physical layer of the candidate technology. This criterion allows to estimate the overhead introduced by the physical layer.

• Control Plane Latency: Allows for estimation of call setup duration. It measured by the transition time, e.g. between the idle and active mode of a user terminal.

(16)

• User Plane Latency: A minimum transmission time for IP-Packets through the radio access network is required.

• Intra- and Inter-frequency handover interruption time: Candidates are required to support seamless handovers between cells of the system. Therefore a minimum handover interruption time is required.

Some aspects of IMT-A candidate systems cannot be investigated analytically and need to be addressed by simulation. These criteria are:

• Cell Spectral Efficiency: IMT-A systems should provide their users with high data rates. The assigned spectrum must be utilized efficiently.

• Cell Edge User Spectral Efficiency: High data rates must be provided to users, while at all times a minimum data rate should be available to cell edge users. Cell spectral efficiency and cell edge user spectral efficiency are to be determined in the same simulation runs.

• Mobility: The candidate system should be able to operate at mobile terminal speeds of up to 350km/h. This is evaluated by link-level simulations.

• VoIP Capacity: IMT-A systems shall not only be able to support high data rates, but also large number of users. The VoIP capacity is used to evaluated maximum load of users - with rather low traffic demand – that can be supported.

The present document reports on the simulation work that has been carried out in WINNER+ to analyse the LTE-Advanced performance with respect to the last group of requirements. Analytical and inspection requirements have been studied by WP3 and the results are reported in deliverable D3.4 [WIN+D3.4]. Table 3.3 provides the numerical values of performance requirements set by ITU for IMT-Advanced candidates.

Table 3.3: IMT-Advanced Performance Requirement Values [M.2134]

Minimum technical requirements

Category Required value

Test environment Downlink or uplink

Cell spectral efficiency (bit/s/Hz/cell)

Indoor Downlink 3

Uplink 2.25

Micro cellular Downlink 2.6

Uplink 1.8

Base coverage urban Downlink 2.2

Uplink 1.4

High speed Downlink 1.1

Uplink 0.7

Peak spectral efficiency (bit/s/Hz)

Not applicable Downlink 15

Uplink 6.75

Bandwidth Not applicable Up to and including (MHz)

40 Scalability Support of at least three

band-width values Cell edge user spectral

efficiency (bit/s/Hz)

Indoor Downlink 0.1

Uplink 0.07

Microcellular Downlink 0.075

Uplink 0.05

Base coverage urban Downlink 0.06

Uplink 0.03

High speed Downlink 0.04

Uplink 0.015

Control plane latency (ms) Not applicable Not applicable Less than 100 ms User plane latency (ms) Not applicable Not applicable Less than 10 ms

(17)

Minimum technical requirements

Category Required value

Test environment Downlink or uplink

Mobility classes Indoor Uplink Stationary, pedestrian

Microcellular Uplink Stationary, pedestrian, vehicular up to 30 km/h Base coverage urban Uplink Stationary, pedestrian,

vehicular

High speed Uplink High speed vehicular,

vehicular Mobility

Traffic channel link data rates (bit/s/Hz)

Indoor Uplink 1.0

Microcellular Uplink 0.75

Base coverage urban Uplink 0.55

High speed Uplink 0.25

Intra-frequency hand-over interruption time (ms)

Not applicable Not applicable 27.5

Inter-frequency handover interruption time within a spectrum band (ms)

Not applicable Not applicable 40

Inter-frequency handover interruption time between spectrum bands (ms)

Not applicable Not applicable 60

Inter-system handover Not applicable Not applicable Not applicable Number of supported VoIP

users (active users/ sector/MHz)

Indoor As defined in Report ITU-R M.2134

50 Microcellular As defined in Report

ITU-R M.2134

40 Base coverage urban As defined in Report

ITU-R M.2134

40 High speed As defined in Report

ITU-R M.2134

(18)

4

Channel Model Calibration

The assessment of the different IMT-Advanced proposals, LTE-Advanced included, was made through cross checking the reports prepared by a set of External Evaluation Groups (EEGs) and the self-evaluation made by the proponents. There are fourteen registered EEGs, one of them the European project WINNER+.

Coordination between evaluation groups was strongly recommended by ITU-R to facilitate comparison and consistency of results and to simplify the understanding of differences in evaluation results achieved by the independent evaluation groups. Indeed, the divergence in the results obtained in the evaluation of the same system is a common problem encountered in all forums where researchers coming from different bodies try to provide their contributions to the progress of science and technology. A possibility to overcome this situation is the comparison of different approaches using the same calibration process and benchmark data. In this framework, and in order to simplify the IMT-Advanced assessment, ITU-R has already approved a number of documents describing the evaluation process, requirements and evaluation criteria. In particular, Report [M.2135] contains the detailed simulation assumptions and the evaluation methodologies of IMT-Advanced. The M.2135document represents a significant calibration effort that intends to ensure proper harmonization of the tools used by the EEGs for performance evaluation of the IMT-Advanced technologies.

Report [M.2135] is mainly focused on the definition of the reference scenarios for system level simulations including large and small scale parameters of the channel model. The new stochastic geometric model proposed by the ITU-R is far from being simple to implement. Several implementations were freely offered, but the main problem is that these implementations are not coherent and do not provide the same output statitics. Without a proper calibration of the channel model implementation it is not feasible to build up a consistent evaluation of candidates. This is why the WINNER+ project addressed this channel calibration from the beginning.

This chapter intends to describe the channel model calibration effort made in WINNER+. Besides, this action aimed to share the experience, information and benchmark data with the remaining EEGs in order to foster the required coordination and unification of results.

The calibration data presented in this chapter is also available in Excel format at the WINNER+ IMT-Advanced evaluation web page:

http://projects.celtic-initiative.org/winner+/WINNER+%20Evaluation%20Group.html

4.1

Calibration of large scale parameters

For Large Scale Calibration (LSC) multi-cell system level simulations must be used. Given that this kind of simulators also must be used for evaluating the IMT-Advanced requirements of cell spectral efficiency, cell edge user throughput, VoIP capacity and mobility, this LSC is also useful to make a prompt detection of potential incoherencies among contributors. Some important properties of the system simulations are determined by the environment description in [M.2135], including the propagation and channel models. The metrics used in this calibration are the path gain and the wideband SINR, which are essentially technology independent and hence calibration of these metrics can be performed using just a few additional assumptions compared to what is given in [M.2135].

The path gain is defined as the ‘average’ signal attenuation between a user terminal and its serving base station cell. The measure includes distance attenuation, shadowing and antenna gains (both at the base station and at the user terminal) while the effects from fast fading are excluded. The path gain may hence be defined as the difference between the (average) received power and the (average) transmitted power.

[ ]

rx tx

Pathgain=P −P dB _(4.1)

The downlink wideband SINR, sometimes also called the geometry, is the (average) power received from the serving cell in relation to the (average) received power from all other cells plus noise. For a user terminal connected to the base station cell i the geometry (G) is defined as:

, , 0 rx i rx j j i P G P N ∀ ≠ = +

∑

, (4.2)

where Prx,j is the received power from the base station cell j and N0 is the noise power.

In addition to the evaluation principles and assumption in [M.2135] and the channel model clarifications that followed, the next assumptions have been used to derive the path gain and wideband SINR distributions.

(19)

Table 4.1: Large Scale Assumptions Cell Selection 1 dB Handover Margin

Feeder Loss 2 dB

BS antenna tilt InH UMi UMa RMa SMa

N.A. 12 12 6 6

In the following sub-sections calibration results from up to seven WINNER+ partners are presented.

4.1.1 Path gain and wideband SINR for InH

0 10 20 30 40 50 60 70 80 90 100 -100 -90 -80 -70 -60 -50 -40 C .D .F . [% ] Pathgain [dB] Org 1 Org 2 Org 3 Org 4 Org 5 Org 6 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 -10 0 10 20 30 40 50 60 C .D .F . [% ] Wideband SINR [dB] Org 1 Org 2 Org 3 Org 4 Org 5 Org 6 Org 7 Org 8 Average

Figure 4.1: Path gain and wideband SINR distributions in the InH scenario

4.1.2 Path gain and wideband SINR for UMi

0 10 20 30 40 50 60 70 80 90 100 -140 -120 -100 -80 -60 -40 C .D .F . [% ] Pathgain [dB] Org 1 Org 2 Org 4 Org 5 Org 6 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 -10 -5 0 5 10 15 20 C .D .F . [% ] Wideband SINR [dB] Org 1 Org 2 Org 4 Org 5 Org 6 Org 7 Org 8 Average

Figure 4.2: Path gain and wideband SINR distributions in the UMi scenario

4.1.3 Path gain and wideband SINR for UMa

0 10 20 30 40 50 60 70 80 90 100 -140 -130 -120 -110 -100 -90 -80 -70 -60 -50 C .D .F . [% ] Pathgain [dB] Org 1 Org 2 Org 3 Org 4 Org 5 Org 6 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 -10 -5 0 5 10 15 20 C .D .F . [% ] Wideband SINR [dB] Org 1 Org 2 Org 3 Org 4 Org 5 Org 6 Org 7 Org 8 Average

(20)

4.1.4 Path gain and wideband SINR for RMa 0 10 20 30 40 50 60 70 80 90 100 -130 -120 -110 -100 -90 -80 -70 -60 -50 C .D .F . [% ] Pathgain [dB] Org 1 Org 2 Org 3 Org 4 Org 5 Org 6 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 -10 -5 0 5 10 15 20 C .D .F . [% ] Wideband SINR [dB] Org 1 Org 2 Org 3 Org 4 Org 5 Org 6 Org 7 Org 8 Average

Figure 4.4: Path gain and wideband SINR distributions in the UMa scenario

4.1.5 Comparison of calibration results

At a first glance the calibration results of all partners seem to fit quite well. Some really small differences appear in terms of path gain, most of all for the InH case, but these differences are counteracted when considering the wideband SINR distribution. Taking all the results of the given scenarios into account the simulator geometry and large scale parameters are considered calibrated.

4.2

Calibration of small scale parameters

In the small-scale fading characteristics we include the delay spread and the angular spread at the base station and at the user terminal. For simplicity, the small-scale fading calibration is performed using omni-directional antennas at both the base station and the user terminal. If other antenna patterns are assumed, e.g., a directional antenna pattern at the base station, the results will be different. Moreover, the calibrations are performed separately for LoS, NLoS and outdoor-to-indoor (OtoI) propagation conditions. OtoI propagation is relevant only in the UMi scenario. For calibration of the angular spread for LoS propagation channels it is important to account for the correction under Section 3 in [IMT-ADV/3].

Now assume that each propagation channel comprises N clusters and that each cluster comprises M rays. Assume further that the delay of ray m in cluster n is denoted τn,m and that the associated power is denoted pn,m. In case of LoS propagation the LoS ray is here included as a separate cluster for which, according

[M.2135], only the first ray in the cluster has a non-zero power.

To calculate the delay spread, the average delay τis first calculated according to equation (4.3).

, , 1 1 , 1 1 N M n m n m n m N M n m n m p p τ τ = = = = ⋅ =

∑∑

(4.3)

Then, the root-mean-square (RMS) delay spread (στ) is calculated according to equation (4.4).

2 , , 1 1 , 1 1 ( ) N M n m n m n m N M n m n m p p τ τ τ σ = = = = − ⋅ =

∑∑

(4.4)

For the angular spread we use the circular angular spread (σAS) as defined in Annex A of [25.996], where

the angular spread is the minimum spread over different linear shifts ∆. One small addition is used here, however. Before calculating θn,m,µ(∆) we wrap the quantity µθ(∆) into the interval [-π, π] according to

equation (2.5). This step is not explicitly stated in [25.996].

2 ( ) ( ) ( ) ( ) ( ) ( ) 2 ( ) if if if θ θ θ θ θ θ θ π µ µ π µ µ µ π µ π µ π + ∆ ∆ < −   ∆ = ∆ ∆ ≤  _{∆ −} _{∆ >}  (4.5)

(21)

The RMS delay spread (στ) and the circular angular spread (σAS) at the base station and at the user

terminal are calculated for a large number of radio links and in the calibrations we compare the corresponding distributions. Taking a downlink perspective on the radio channel, the angular spread at the base station and at the user terminal are often denoted angle of departure (AoD) and angle of arrival (AoA), respectively. This notation is also used here below.

4.2.1 Small-scale fading in InH

0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (us) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (µs) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average

Figure 4.5: RMS delay spread for InH NLoS (left plot) and LoS (right plot)

0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOA (degrees) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOA (degrees) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average

Figure 4.6: Circular AoA for InH NLoS (left plot) and LoS (right plot)

0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOD (degrees) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOD (degrees) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average

(22)

4.2.2 Small-scale fading in UMi 0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (us) Org 1 Org 2 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (µs) Org 1 Org 2 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (µs) Org 1 Org 2 Org 4 Org 7 Org 8 Average

Figure 4.8: RMS delay spread for UMi NLoS (upper left plot), LoS (upper right plot) and outdoor-to-indoor (lower plot)

0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOA (degrees) Org 1 Org 2 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOA (degrees) Org 1 Org 2 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOA (degrees) Org 1 Org 2 Org 4 Org 7 Org 8 Average

Figure 4.9: Circular AoA for UMi NLoS (upper left plot), LoS (upper right plot) and outdoor-to-indoor (lower plot)

(23)

0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOD (degrees) Org 1 Org 2 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOD (degrees) Org 1 Org 2 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0 40 80 120 C .D .F . [% ] AOD (degrees) Org 1 Org 2 Org 4 Org 7 Org 8 Average

Figure 4.10: Circular AoD for UMi NLoS (upper left plot), LoS (upper right plot) and outdoor-to-indoor (lower plot)

4.2.3 Small-scale fading in UMa

0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (us) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (µs) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average

Figure 4.11: RMS delay spread for UMa NLoS (left plot) and LoS (right plot)

(24)

Figure 4.13: Circular AoD for UMa NLoS (left plot) and LoS (right plot)

4.2.4 Small-scale fading in RMa

0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (us) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average 0 10 20 30 40 50 60 70 80 90 100 0,001 0,010 0,100 1,000 10,000 C .D .F . [% ] Delay (µs) Org 1 Org 2 Org 3 Org 4 Org 5 Org 7 Org 8 Average

Figure 4.14: RMS delay spread for RMa NLoS (left plot) and LoS (right plot)

Figure 4.15: Circular AoA for RMa NLoS (left plot) and LoS (right plot)

(25)

5

LTE Rel-8 Baseline Configuration & Calibration

The evaluation group WINNER+ has its focus on evaluating the 3GPP IMT-Advanced proposal and as a preparation of the system level evaluations a simulator calibration for a basic LTE Release 8 configuration was performed. The reference LTE Release 8 configuration can be found in [36.814]. The purpose of the calibration is hence to make sure that the different simulation tools produce comparable output. Note that an optimization of the system performance was not in the scope of this activity and it was desired to observe not the best performance figures but well aligned performance figures. A similar calibration activity was performed by 3GPP and is presented in Annex A2.2 of [36.814].

The calibration approach used in the WINNER+ evaluation group was a two-step approach. The first step was the channel model calibration shown in previous chapter, and the second step was the system level simulator calibration (LTE Release 8 configuration - based). Note that harmonization of simulators used a procedure that involved all simulator building blocks (e.g. link-to-system interface, link-adaptation algorithms) to produce output results that are further assessed concerning an alignment of the performance figures originating from different simulation tools.

In the WINNER+ simulator calibration, the focus was on the user throughput distributions and the cell spectral efficiency as well as the cell-edge user throughput. Information about the meaning of these parameters is included in Chapter 3.

Below, in Figures 5.1 to 5.10, the normalized downlink and uplink user throughput distributions are presented for the different deployment scenarios.

5.1

Calibration results in uplink

(26)

Figure 5.2: CDF of Normalized User Throughput in Uplink UMi deployment scenario

(27)

Figure 5.4: CDF of Normalized User Throughput in Uplink RMa deployment scenario

(28)

5.2

Calibration results in downlink

Figure 5.6: CDF of Normalized User Throughput in Downlink InH deployment scenario

(29)

Figure 5.8: CDF of Normalized User Throughput in Downlink UMa deployment scenario

(30)

Figure 5.10: CDF of Normalized User Throughput in Downlink SMa deployment scenario

The LTE Release 8 calibration activity was approached by 7 independent organizations. Downlink calibration in InH, UMi, UMa, RMa, SMa was performed by 5 up to 7 organizations – the number of participants varied depending on test environment. Uplink calibration was approached by 2 to 3 parties. The observed results obtained from different simulation tools are well aligned with each other. There are deviations observed (see section 5.3 for quantitative assessment of calibration) but considering the complexity of the system level simulator and the amount of degrees of freedom (despite a common set of assumptions and models), the obtained alignment is satisfactory. For illustration purposes, a comparision of WINNER+ calibration results with 3GPP calibration results can be found in Figure 5.6 where curves obtained by different WINNER+ organizations are collocated with an average curve obtained by 3GPP contributors.

5.3

Quality of calibration – quantitative assessment

In order to perform quantitative characterization of calibration quality, we introduced the coefficient of variation calculated as:

SE v SE σ c = µ ,

where σSE and µSEare, respectively, standard deviation and mean of the cell or cell-edge spectral

efficiency in the given deployment scenarios. Standard deviation and mean of spectral efficiency for any given deployment scenario were calculated over the results from different partners.

Table 5.1 presents the coefficients of variation. Note that in the uplink, the figures are based on only a few (two or three) simulation results.

Since the coefficient of variation reflects deviations from the mean relative to the mean value itself, this measure is very sensitive for deviations in the low range of spectral efficiency values. That is why coefficients of variation associated with cell-edge spectral efficiency are significantly higher compared to cell spectral efficiency.

Coefficients of cell-edge spectral efficiency variation for all test cases falls into range of 8% – 28% and respectively, coefficients of cell spectral efficiency variation falls into range of 2% – 15%.

Table 5.1: Coefficients of spectral efficiency variation for Release-8 basic configuration simulation results in various deployment scenarios

Deployment scenario InH UMi UMa RMa SMa

Downlink Cell-edge spectral efficiency 20% 28% 20% 16% 10% Cell spectral efficiency 5% 12% 15% 6% 12%

Uplink Cell-edge spectral efficiency 14% 9% 8% 16% 13% Cell spectral efficiency 6% 2% 7% 10% 8%

(31)

6

Simulation Results

The assessment results by simulation for LTE-Advanced are presented and compared to the ITU-R requirements in this chapter. First the Cell and Cell Edge spectral efficiencies are addressed for the different test environments for uplink, downlink, FDD RIT and TDD RIT. Next the performance evaluation results related to mobility are shown. The traffic channel link data rates and mobility classes are considered. Finally, the VoIP capacity estimation results are presented. The simulation assumptions and detailed results for these assessments are also described.

6.1

Cell and cell edge spectral efficiency

6.1.1 Summary of results

In this subsection, the requirements and simulation results for the Cell spectral efficiency are summarized for the FDD RIT in Table 6.1 and for TDD RIT in Table 6.2. Simulation results show that the requirements are achieved for all environments by using different MIMO features of the LTE advanced. The same conclusion is available for Tables 6.3 and 6.4 where the Cell Edge Spectral Efficiency assessment results are summarized for FDD and TDD RIT. The assumptions used to derive simulations and some more simulation results for several organizations are described in the subsequent subsections

Table 6.1: Cell spectral efficiency results for FDD RIT

Test environment

Link direction Indoor Microcellular Base coverage urban

High speed

Requirements DL (bit/s/ Hz/cell) 3 2.6 2.2 1.1

UL (bit/s/ Hz/cell) 2.25 1.80 1.4 0.7

Organization 1 DL (bit/s/ Hz/cell) L = 3 / 2 / 1* 4.12 / 4.55 / 4.83 (4x2 SU-MIMO) 2.87 / 3.16 / 3.46 (4x2 MU-MIMO) 2.37 / 2.62 / 2.86 (4x2 MU-MIMO) 1.96 / 2.14 / 2.27 (4x2 SU-MIMO) 3.47 / 3.82 / 4.18 (4x2 MU-MIMO) UL (bit/s/ Hz/cell) 3.46 (1x4 SIMO) 6.06 (1x4 MU-MIMO) 5.58 (2x4 SU-MIMO) 2.23 (1x4 SIMO) 2.41 (2x4 BF) 2.59 (2x4 SU-MIMO) 1.77 (1x4 SIMO) 2.94 (2x4 BF) 1.97 (2x4 SU-MIMO) 2.08 (1x4 SIMO) 2.34 (2x4 BF) 2.38 (2x4 SU-MIMO) Organization 2 DL (bit/s/ Hz/cell) 4.24

(4x2 SU-MIMO) 2.88 (4x2 MU-MIMO) 2.43 (4x2 MU-MIMO) 2.77 (4x2 MU-MIMO)

Organization 3 DL (bit/s/ Hz/cell) 1.87

(4x2 SU-MIMO) Organization 4 DL (bit/s/ Hz/cell) 3.76

(4x2 SU-MIMO 2.90 (4x2 MU-MIMO 2.33 (4x2 MU-MIMO 3.20 (4x2 MU-MIMO UL (bit/s/ Hz/cell) 2.86 (1x4 SIMOA) 1.90 (1x4 SIMOA) 1.43 (1x4 SIMOA) 1.62 (1x4 SIMOA) Mean cell spectral efficiency for FDD RIT ** DL (bit/s/ Hz/cell) 4.10 (4x2 SU-MIMO) 2.88 (4x2 MU-MIMO) 2.38 (4x2 MU-MIMO) 1.92 (4x2 SU-MIMO) 3.15 (4x2 MU-MIMO) UL (bit/s/ Hz/cell) 3.16 (1x4 SIMO) 6.06 (1x4 MU-MIMO) 5.58 (2x4 SU-MIMO) 2.07 (1x4 SIMO) 2.41 (2x4 BF) 2.59 (2x4 SU-MIMO) 1.60 (1x4 SIMO) 2.94 (2x4 BF) 1.97 (2x4 SU-MIMO) 1.85 (1x4 SIMO) 2.34 (2x4 BF) 2.38 (2x4 SU-MIMO)

(32)

* See Table 6.6. L = 3 values were taken for mean calculation.

** Mean value of all contributing organizations for the given antenna configurations. Note that different assumptions were made in the underlying simulations, so that the mean value does not represent the performance of one particular system setup. Values in bold (maximum values) are taken as main results.

Table 6.2: Cell spectral efficiency results for TDD RIT

High speed

Requirements DL (bit/s/ Hz/cell) 3 2.6 2.2 1.1

UL (bit/s/ Hz/cell) 2.25 1.80 1.4 0.7

Organization 1 DL (bit/s/ Hz/cell) L = 3 / 2 / 1* 3.93 / 4.22 / 4.50 (4x2 SU-MIMO) 2.76 / 2.95 / 3.25 (4x2 MU-MIMO) 2.27 / 2.44 / 2.68 (4x2 MU-MIMO) 1.92 / 2.06 / 2.20 (4x2 SU-MIMO) 3.45 / 3.69 / 4.06 (4x2 MU-MIMO) UL (bit/s/ Hz/cell) 3.24 (1x4 SIMO) 5.67 (1x4 MU-MIMO) 5.15 (2x4 SU-MIMO) 2.03 (1x4 SIMO) 2.20 (2x4 BF) 2.35 (2x4 SU-MIMO) 1.66 (1x4 SIMO) 1.79 (2x4 BF) 1.81 (2x4 SU-MIMO) 1.93 (1x4 SIMO) 2.17 (2x4 BF) 2.19 (2x4 SU-MIMO) Organization 5 DL (bit/s/ Hz/cell) 3.91

(4x2 SU-MIMO 2.74 (4x2 MU-MIMO 2.34 (4x2 MU- MIMO 1.73 (4x2 SU-MIMO UL (bit/s/ Hz/cell) 2.80 (1x4 SIMOA) 1.90 (1x4 MU-MIMO) 1.41 (1x4 SIMOA) 1.55 (1x4 SIMOA) Organization 7 DL (bit/s/ Hz/cell) 4.92

4x2 MU-MIMO 2.75 4x2 MU-MIMO 1.88 4x2 MU-MIMO Mean cell spectral efficiency for TDD RIT ** DL (bit/s/ Hz/cell) 3.92 (4x2 SU-MIMO) 4.92 (4x2 MU-MIMO) 2.75 (4x2 MU-MIMO) 2.31 (4x2 MU-MIMO) 1.83 (4x2 SU-MIMO) 2.67 (4x2 MU-MIMO) UL (bit/s/ Hz/cell) 3.02 (1x4 SIMO) 5.67 (1x4 MU-MIMO) 5.15 (2x4 SU-MIMO) 2.03 (1x4 SIMO) 1.90 (1x4 MU-MIMO) 2.20 (2x4 BF) 2.35 (2x4 SU-MIMO) 1.54 (1x4 SIMO) 1.79 (2x4 BF) 1.81 (2x4 SU-MIMO) 1.74 (1x4 SIMO) 2.17 (2x4 BF) 2.19 (2x4 SU-MIMO) * See Table 6.6. L = 3 values were taken for mean calculation.

** Mean value of all contributing organizations for the given antenna configurations. Note that different assumptions were made in the underlying simulations, so that the mean value does not represent the performance of one particular system setup. Values in bold (maximum values) are taken as main results.

(33)

Table 6.3: Cell edge spectral efficiency results for FDD RIT

High speed

Requirements DL (bit/s/ Hz/cell) 0.1 0.075 0.06 0.04

UL (bit/s/ Hz/cell) 0.07 0.05 0.03 0.015

Organization 1 DL (bit/s/ Hz/cell) L = 3 / 2 / 1* 0.180/0.199/0.210 (4x2 SU-MIMO) 0.094/0.104/0.114 (4x2 MU-MIMO) 0.076/0.083/0.091 (4x2 MU-MIMO) 0.076/0.083/0.088 (4x2 SU-MIMO) 0.112/0.123/0.135 (4x2 MU-MIMO) UL (bit/s/ Hz/cell) 0.262 (1x4 SIMO) 0.442 (1x4 MU-MIMO) 0.288 (2x4 SU-MIMO) 0.109 (1x4 SIMO) 0.124 (2x4 BF) 0.127 (2x4 SU-MIMO) 0.085 (1x4 SIMO) 0.092 (2x4 BF) 0.091 (2x4 SU-MIMO) 0.100 (1x4 SIMO) 0.118 (2x4 BF) 0.117 (2x4 SU-MIMO) Organization 2 DL (bit/s/ Hz/cell) 0.160

(4x2 SU-MIMO) 0.085 (4x2 MU-MIMO) 0.061 (4x2 MU-MIMO) 0.103 (4x2 MU-MIMO)

Organization 3 DL (bit/s/ Hz/cell) 0.054

(4x2 SU-MIMO 0.087 (4x2 MU-MIMO 0.063 (4x2 MU-MIMO 0.057 (4x2 MU-MIMO C) UL (bit/s/ Hz/cell) 0.170 (1x4 SIMOA) 0.055 (1x4 SIMOA) 0.060 (1x4 SIMOA) 0.072 (1x4 SIMOA)

Mean cell edge spectral efficiency for FDD RIT ** DL (bit/s/ Hz/cell) 0.173 (4x2 SU-MIMO) 0.089 (4x2 MU-MIMO) 0.067 (4x2 MU-MIMO) 0.065 (4x2 SU-MIMO) 0.091 (4x2 MU-MIMO) UL (bit/s/ Hz/cell) 0.216 (1x4 SIMO) 0.442 (1x4 MU-MIMO) 0.288 (2x4 SU-MIMO) 0.082 (1x4 SIMO) 0.124 (2x4 BF) 0.127 (2x4 SU-MIMO) 0.073 (1x4 SIMO) 0.092 (2x4 BF) 0.091 (2x4 SU-MIMO) 0.086 (1x4 SIMO) 0.118 (2x4 BF) 0.117 (2x4 SU-MIMO) * See Table 6.6. L = 3 values were taken for mean calculation.

** Mean value of all contributing organizations for the given antenna configurations. Note that different assumptions

were made in the underlying simulations, so that the mean value does not represent the performance of one particular system setup. Values in bold (following the antenna configuration selected in Table 6.1) are taken as main results.

(34)

Table 6.4: Cell edge spectral efficiency results for TDD RIT

High speed

Requirements DL (bit/s/ Hz/cell) 0.1 0.075 0.06 0.04

UL (bit/s/ Hz/cell) 0.07 0.05 0.03 0.015

Organization 1 DL (bit/s/ Hz/cell) L = 3 / 2 / 1* 0.168/0.181/0.193 (4x2 SU-MIMO) 0.088/0.095/0.104 (4x2 MU-MIMO) 0.069/0.074/0.082 (4x2 MU-MIMO) 0.072/0.077/0.082 (4x2 SU-MIMO) 0.110/0.117/0.129 (4x2 MU-MIMO) UL (bit/s/ Hz/cell) 0.246 (1x4 SIMO) 0.413 (1x4 MU-MIMO) 0.267 (2x4 SU-MIMO) 0.099 (1x4 SIMO) 0.111 (2x4 BF) 0.114 (2x4 SU-MIMO) 0.079 (1x4 SIMO) 0.085 (2x4 BF) 0.084 (2x4 SU-MIMO) 0.092 (1x4 SIMO) 0.108 (2x4 BF) 0.106 (2x4 SU-MIMO) Organization 5 DL (bit/s/ Hz/cell) 0.17

(4x2 SU-MIMO 0.08 (4x2 MU-MIMO 0.065 (4x2 MU-MIMO 0.04 (4x2 SU-MIMO C) UL (bit/s/ Hz/cell) 0.160 (1x4 SIMOA) 0.055 (1x4 MU-MIMO) 0.048 (1x

D4.2 Final conclusions on end-to-end performance and sensitivity analysis

D4.2 Final conclusions on end-to-end performance

Table of Contents

Channel Model Calibration ... 18

LTE Rel-8 Baseline Configuration & Calibration ... 25

Sensitivity Analysis ... 47

Authors

Executive Summary

List of Acronyms and Abbreviations

Introduction

LTE-Advanced Features

Focus on LTE-Advanced features for IMT-Advanced assessment

The WINNER+ Evaluation Group approach

IMT-Advanced Evaluation Scenarios & Performance Criteria

Table 3.1: Deployment Scenarios

Performance criteria

Table 3.2: List of IMT-Advanced Performance Criteria [M.2134] Characteristic for evaluation Method Evaluation methodology / configurations Related Section in Report ITU-R M.2135 Related section of Reports ITU-R M.2134 and ITU-R M.2133

Characteristic for evaluation Method Evaluation methodology / configurations Related Section in Report ITU-R M.2135 Related section of Reports ITU-R M.2134 and ITU-R M.2133

Table 3.3: IMT-Advanced Performance Requirement Values [M.2134]

Minimum technical requirements

ITU-R M.2134

Channel Model Calibration

Calibration of large scale parameters

4.1.1 Path gain and wideband SINR for InH

4.1.3 Path gain and wideband SINR for UMa

Calibration of small scale parameters

4.2.1 Small-scale fading in InH

Figure 4.9: Circular AoA for UMi NLoS (upper left plot), LoS (upper right plot) and outdoor-to-indoor (lower plot)

Figure 4.13: Circular AoD for UMa NLoS (left plot) and LoS (right plot)

LTE Rel-8 Baseline Configuration & Calibration

Calibration results in uplink

Figure 5.2: CDF of Normalized User Throughput in Uplink UMi deployment scenario

Calibration results in downlink

Figure 5.6: CDF of Normalized User Throughput in Downlink InH deployment scenario

Quality of calibration – quantitative assessment

Table 5.1: Coefficients of spectral efficiency variation for Release-8 basic configuration simulation results in various deployment scenarios

Simulation Results

6.1.1 Summary of results

High speed

Table 6.2: Cell spectral efficiency results for TDD RIT

Table 6.3: Cell edge spectral efficiency results for FDD RIT

** Mean value of all contributing organizations for the given antenna configurations. Note that different assumptions

High speed