5G

What to expect and where to start

Mark Cudak
Principal Research Specialist
Technology & Innovation

1 © Nokia 2015
Outline
 5G Overview and Requirements
 Air Interface for 5G
 5G < 6GHz and cmWave (6-30 GHz)
 mmWave (30-100 GHz)
 Massive MIMO
 5G Proof-of-Concept (PoC) and Standards Timeline
 Summary and Next Steps

2 © Nokia 2015
5G will expand the human possibilities of the connect world

Throughput

Gigabytes in a second
3D video – 4K screens
Work and play in the cloud

Smart city cameras Augmented reality

Voice Industry & vehicular automation

Mission critical broadcast

Sensor NW
Self Driving Car

# of Devices; Latency;
Cost; Power Reliability

(Low power) Wide area Crowd Outdoor

Ultra-dense
A trillion devices with different needs GB transferred in an instant Mission-critical wireless control and automation

3 © Nokia 2015
5G will expand the human possibilities of the connect world

>10 Gbps Throughput 100 Mbps

peak data rates avg. goodput
Gigabytes in a second

10-100 3D video – 4K screens 10 000

x more devices Work and play in the cloud x more traffic

Smart city cameras Augmented reality

M2M Voice Industry & vehicular automation Ultra

ultra low cost
Mission critical broadcast reliability
Sensor NW
Self Driving Car

10 years # of Devices; Latency; <1 ms

on battery Cost; Power Reliability latency

(Low power) Wide area Crowd Outdoor

Ultra-dense
A trillion devices with different needs GB transferred in an instant Mission-critical wireless control and automation

4 © Nokia 2015
5G radio access to match the available new and old frequency bands

A new RAT may be motivated by new spectrum allocation (bands above 6GHz), lower latency, or
specific use cases.

LTE-A evolution beyond 3GPP Rel-12 needs to be backwards

compatible, meaning: “Legacy LTE devices must be able to access the
LTE-A will be essential foundation of the integrated system without degradation in performance”
5G system – must continue to evolve in parallel to 5G Backwards compatibility requirement may be relaxed, if specific
needs (e.g. new bands without legacy devices), such as LAA-LTE, are
identified and agreed on

5G below 6 GHz 5G cmWave 5G mmWave

LTE-A evolution

1 GHz 2 GHz 6 GHz 10 GHz 20 GHz 30 GHz 60 GHz

100 GHz
Within WRC2015
Expected to be within WRC2019 scope
scope

5 © Nokia 2015
Why 6-100 GHz?
• 6-100 GHz expected to be in the scope of WRC 2019
• Channel models exist below 6 GHz
- e.g., 3GPP 3D channel model, WINNER
- Question: will these models be consistent with channel models from 6-100
GHz?
• E.g., can a reasonable comparison be made between three simulated
systems: one at 2.6 GHz, one at 10 GHz, and one at 72 GHz?
• Why 100 GHz as the upper limit?
- Plenty of spectrum to exploit below 100 GHz, no need at this moment to go
above 100 GHz
- Technologically it is easier to stay below 100 GHz
- Availability of measurements
6 © Nokia 2015
5G is to enable above 6 GHz access & optimize below 6 GHz access
Expanding the spectrum assets to deliver capacity and experience

Spectrum availability
Antenna Interference
LOS Spectrum technologies conditions
90 GHz
+ ~1 GHz
Low Rank
More noise
mmWave carrier bandwidth MIMO/BF
limited

Different spectrum licensing,

sharing and usage schemes
Ultra broadband
Dynamic TDD  efficient beam (70-90GHz)
30 GHz steering

cmWave + Several
Higher Rank
Strong
10 GHz
~100 MHz interference
Enhanced SC* MIMO & BF
Dynamic TDD handling
3 GHz + Up to 100 MHz
10 cm
< 6GHz High Rank
carrier bandwidth Full coverage is
MIMO &
Wide area diverse spectrum,
beamforming
essential
FDD and TDD
Cell size 300 MHz
1m
LOS/NLOS *) SC = Small Cells

7 © Nokia 2015
5G PHY Layer considerations
LTE rel 13 5G Macro optimized 5G E small cells 5G Ultra Dense
SI /WI (sub 6GHz) (cm-wave) (mm-wave)
Spectrum 0.7-3.5GHz (may 0.5-10GHz ? 3-30GHz 30-100GHz
likely extend)
Carrier Bandwidth 1.4-20MHz ~ 5-40MHz ~40-200MHz ~400MHz-2GHz

Duplex FDD/TDD FDD/TDD Dynamic TDD Dynamic TDD

(full duplex FFS)
Transmit power >40dBm/ >40dBm/23dBm <~30dBm /23dBm <~30dBm/23dBm
DL/UL 23dBm
Waveform UL/DL OFDMA/SC-FDMA OFDMA/SC-FDMA * OFDMA/OFDMA SC-TDMA/SC-TDMA

Multiple access Time & frequency Time & frequency Time & (frequency) Time
Multi-antenna SU/MU SU /MU SU/MU SU/MU
technology Beamforming and up Beamforming and Beamforming and Beamforming and
to rank 8 medium rank high rank Low rank
TTI 1ms ? (flexible) ~0.25ms ~0.1ms
8 © Nokia 2015
* Other waveforms for massive MTC is FFS
5G architecture to integrate novel and legacy technologies

Key requirements + Low-latency New QoS paradigm Virtual NW & local

Revolutionary
Services services
1. Multi Service Network
2. Network Flexibility

Embedded security
Operator benefits Innovative use cases Service aware radio Ad-hoc virtual subnets

• Support for future

applications Programmability Integrated Telco Cloud with
control SDN/ NW elasticity
• Per service tailored Evolutionary
network SW
• New services &
business models
Diverse services Traffic steering & Virtualization &
• Quicker service time + service chaining SW-driven network
to market +
9 © Nokia 2015

main | integration
5G below 6 GHz and cmWave

10 © Nokia 2015
 5G components

> 10 Gbps peak rate 10 000 x more traffic

Network energy efficiency by
minimizing common signals

Flexible high-efficiency radio ready for

ultra-dense deployments above 6 GHz
High-rank MIMO
Significant new spectrum above 6 GHz and
Higher bandwidth mmW-optimized radio needed to achieve
Lower overhead ultra-dense networks and 10 000 x traffic

100 Mbps when needed < 1 ms latency

Efficient interference mitigation with < 1 ms latency Radio latency achieved with
enhanced mMIMO/CoMP and advanced short TTI frame structure
receivers < 1 ms E2E latency needs core network
Native HetNet support enhancements

11 © Nokia 2015
5G phase 1 to be initially deployed below 6 GHz due to band availability

< 6GHz spectrum availability

WRC Phase 1 radio
GHz
Potential:
up to 2 GHz
6 GHz 2015: Some additional bands 3…6 GHz unpaired band as initial
<6GHz to be identified – in time deployment target
3 GHz
for 2020 deployments Ready for > 6 GHz unpaired
10 cm today New*
2019: Expected to identify >6GHz bands and unlicensed bands as is
bands – too late for 2020 Easily extensible to paired bands,
Fragmented & deployments also under 3 GHz
mixed

100-200 MHz carrier bandwidth supported

TDD FDD
300MHz
High degree of spectrum flexibility required due to fragmented spectrum
1m
Carrier aggregation / dual connectivity, also with LTE bands

12 © Nokia 2015
Dynamic TDD frame structure with short TTI

0.25 ms TTI is the maximum possible for 1 ms radio latency

New frame structure a must for low latency,
TD-LTE subframe scaling not sufficient Delay 5G TDD LTE-A LTE-A
component req’ment TDD FDD
Dynamic TDD for good traffic adaptability – UE Processing 0.25 ms 1 ms 1.5 ms
every TTI can be dynamically selected to carry Frame Alignment 0.125 ms 1.1-5 ms
UL or DL data TTI duration 0.25 ms 1 ms 1 ms
Subframe of at most 0.25 ms for low latency eNB Processing 0.375ms 1.5 ms 1.5 ms
HARQ Re-Tx (10 0.1 ms 1.0-1.16 0.8 ms
Adaptive bundling of subframes to a TTI for
% x HARQ RTT) ms
coverage flexibility
Total Delay 1 ms 6-10 ms 5 ms

Frame structure borrowing the best of the TD-LTE special subframe – every TTI can be UL or DL
Subframe 0.25 ms or less
DL CTRL DL DATA GP UL CTRL DL CTRL GP UL DATA UL CTRL

13 © Nokia 2015
OFDM for both UL and DL
Dynamic TDD Natural support for more antennas
Same UL and DL structure enables good IC and larger bandwidth
performance against UL  DL and DL  UL
The spatial channel can be equalized subcarrier-wise
interference
 easy support for MIMO with advanced receivers
 low equalization complexity
D2D

Future-proof for D2D operation

Access/backhaul
Peak rate SNR
Enables access/backhaul convergence, required for OFDM
including in-band 6 dB lower than
SC-FDM
Low MIMO processing complexity important
400 MHz + 4x4 MIMO + 256QAM ≈ 10 Gbps
200 MHz + 8x8 MIMO + 256QAM ≈ 10 Gbps

14 © Nokia 2015
Summary – 5G radio phase 1

Frequency bands Waveform & Frame structure

Support for flexible and wide carrier BW OFDM for both DL and UL
Initial target 3…6 GHz TDD Dynamic TDD
Extendible to > 6 GHz and < 3GHz, FDD Short TTI with bundling

Deployment SE mechanisms Energy efficiency

Applicable to both small cells Interference mitigation No overhead channels
and macro cells Massive MIMO LTE for initial access &
DC with LTE mobility
MU-MIMO

Meets most of the 5G requirements

Radio layer ready for meeting all of the 5G requirements

15 © Nokia 2015
 5G mmWave

16 © Nokia 2015
mmWaves - taking the pressure off the lower frequencies
Expanding wireless communications into the outer limits of radio technology

Huge Natural evolution of small cells Massive antenna arrays

potential to overcome propagation challenges
 Higher frequency, higher pathloss
GHz Available
 Shrinking cells sizes • ≥ 16 element arrays
 mmWave cellular feasible at base station
90-95 2.9 GHz  100-150 meter site-to-site distance
2 + .09 GHz • Beamforming
BW  Dynamic TDD where each slot can at RF for low power consumption
be used for Dl/UL/Backhaul
 Latency < 1msec • Chip-scale array elements
70-85 10 GHz
5 GHz BW • Over-the-air power
Permitting high digital data rates combining provides necessary transmit power
1-2 GHz bandwidth possible • Polarization
38 4 GHz enables 2 stream MIMO
50 MHz BW 10 Gbps with 2 Stream, 16 QAM
28 2 GHz > 100 Mbps cell edge rates result of
150/852 MHz
Technology progress finally makes
BW noise limited system mmWaves practical to use
<6 1 GHz

17 © Nokia 2015
mmWave – propagation and link budget
First step towards deployment of mmWave in ultra dense environments

Channel characterization
at 73 GHz
Measurements in cooperation with
NYU and Aalto University

Delay Outage Penetration Reflections Pathloss Pathloss

spread loss exponent
< 1 ns
LOS conditions,
narrow beam Body loss Oxygen/rain 3–5
quite high not an issue for reflective paths 21 dB
~25ns radius < 200m compared to 5 GHz
RMS delay  steerable  can be used to LOS and NLOS
spread in non-LOS directional antenna Foliage loss establish non-LOS 29 dB very similar to
conditions arrays required severe links compared to 2 GHz 3.5GHz band

18 © Nokia 2015
mmWave – propagation and link budget
Indoor channel vs. outdoor channel at 73 GHz

Indoor PLE STD (dB) Highlights Smaller Slightly larger

RMS delay azimuth angle
LOS (measured) .1.5 1.0 spread indoor spreads
LOS (predicted) 1.5 0.8 vs. outdoor indoor vs.
NLOS (measured) 3.1 9.0
outdoor
NLOS (predicted) 3.1 8.5

Outdoor PLE STD (dB)

Elevation Azimuth angle Full details in
angle spreads distribution: publications
LOS B (measured) 2.0 4.2 and biases uniform (VTC-Fall 2014
LOS B (predicted) 3.5 7.9 monotonically (compared to and ICNC
wrapped Gaussian
decrease with for outdoor) 2015)
NLOS M (measured) 2.0 5.2
distance
NLOS M (predicted) 3.3 7.6

19 © Nokia 2015
Air-Interface Design: Options 20 ms superframe

 Air-Interface for mmwave

37 38 39 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 0 1 2

 Different Options
 OFDM/ZT-SOFDM/NCP-SC TDD Frame 500 µs

 TDD (Variable DL/UL traffic, Simpler Transceiver) 0 1 2 3 4

 Frame Size = 500 µs

 Slot Size = 100 µs
 Downlink/Uplink Interval : Variable
 Characteristics of ELA @ mmWave Control Data

 Few users per AP, no need for FDM

 RF beamforming: avoid multiple users from sharing the TDM Slot 100 µs

same Tx/Rx beam -> loss of beamforming gain

 Reduce PAPR LTE 802.11ad B4G-MMW

 Example MA technique (Null CP Single Carrier) Frequency Band < 6 GHz 60 GHz 70 GHz
Supported Bandwidths TBD 2160 MHz 2000 MHz
 Null portion enables RF beam switching in the CP Maximum QAM 64 16 64 64
without destroying the CP property Modulation OFDM SC OFDM NullCP-SC
 BW = 2 GHz Channel Spacing (B) 20 MHz 2.16 GHz 2.16 GHz 2 GHz
FFT Size 2048 512 512 1024
 Data Block Size = 1024
Subcarrier Spacing 15 kHz 4.2 MHz 5.1 MHz 1.5 MHz
 Pilot Block Size = 256 Sampling Frequency 3.072 MHz 1.76 GHz 2.46 GHz 1.54 GHz
-Modulation Tsampling 32.6 ns 5.68 ps 406 fs 651 fs
−π/2-BPSK, π/4-QPSK, 16 QAM, 64QAM Tsymbol 66.7 μs 245 ns 198 ns 666.7 ns
Tguard 4.7 μs 36.4 ns 52 ns 10.4 ns
 Huge Throughput and Cell Edge gains T 71.4 μs 291 ns 250 ns 666.7 ns

20 © Nokia 2015
mmWave – 5G requirements can be met even in challenging environments

Performance in outdoor
environments
Enabled through
• flexible backhaul
• RFIC/antenna integration
75 AP/km2 150 AP/km2 187 AP/km2
AP density
2.1 Gbps 4.1 Gbps 5.1 Gbps
Average UE Throughput Average UE Throughput Average UE Throughput

<1 Mbps 222 Mbps 552 Mbps

Edge Throughput Edge Throughput Edge Throughput
Network capacity

16.4% 3.2% 1%
Outage Probability Outage Probability Outage Probability
Multi-connectivity

21 © Nokia 2015
 Summary
 mmWave Technology can meet the 5G requirements of peak / edge data rates
and latency
 Well suited for Ultra dense deployments
 Outdoor and Indoor Channel Models based on measurements and ray tracing
 Air Interface Design for 5G mmWave
 Dynamic TDD
 Simple low PAPR design
 Per user based control channels with low overhead
 System level Performance for outdoor and indoor deployments
 Meets the 5G peak and edge data rate requirements

22 © Nokia 2015
5G Massive MIMO

23 © Nokia 2015
 What is “Massive MIMO”?

• Massive MIMO is the extension of traditional MIMO technology to (M-1,0) (M-1,1) (M-1,N-1)

antenna arrays having a large number of controllable antennas

(1,0) (1,1) (1,N-1)

• MIMO = Multiple Input Multiple Output = any transmission scheme

involving multiple transmit and multiple receive antennas
- Encompasses all implementations:
(0,0) (0,1) (0,N-1)
• e.g.: RF/Baseband/Hybrid
- Encompasses all TX/RX processing methodologies:
• e.g., Diversity, Beamforming/precoding, Spatial multiplexing, SU & MU,
joint/coordinated transmission/reception, etc.
• Massive  Large number: >> 8
• Controllable antennas: antennas (whether physical or otherwise)
whose signals are adaptable by the PHY layer (e.g., via gain/phase
control)
24 © Nokia 2015
MIMO and massive MIMO will be one core technology in 5G

Higher the band, smaller the antenna array - or

alternatively same size fits more antennas
Spectrum availability The antenna size is inversely proportional to the frequency
64-antenna
LOS band, and this gives the opportunity to use more antennas
array size
90 GHz
With very high frequency bands (mmW, 30
mmWave Low Rank
2.7cm2 GHz all the way to 100 GHz) the antennas
Ultra MIMO/BF will be used more to focus the transmitted
@73GHz
broadband  efficient beam energy towards the receiver to overcome
30 GHz steering increased pathloss as due to physics of radio
propagation and too many parallel MIMO
cmWave 64cm2 streams to one user is not required due to
10 GHz Enhanced Higher Rank
@15GHz large bandwidths available at these bands.
Small Cell MIMO & BF

3 GHz Hybrid /RF(digital & analog)

10 cm beamforming architecture can be used to
< 6GHz 1176cm2 High Rank MIMO reduce the transmitter cost and energy
Wide area @3.5GHz & beamforming consumption when using massive
number of antennas
Cell size 300 MHz
1m
LOS/NLOS
Different frequency ranges require different IC
technologies, and we are deeply involved in
25 © Nokia 2015 developing these technologies together with our
8x8 patch antenna technology vendors as well as academia.
(64 antennas)
 Massive MIMO for 4G and 5G Systems
Major Performance Boost across all Spectrum ranges and Cell size
Our approach Operator benefits
• Applicable for both Macro and Small Cells
• Massive MIMO provides high gain adaptive beam-forming • Cell edge gain +100%
with antenna arrays
• Spectral efficiency gain +80%
• >> 16 antenna ports (e.g. 16, 32, 64, 256 antenna ports)
• Coverage gain to compensate the path loss on high
• Massive MIMO with large arrays becomes practical because bands making cm and mm waves more practical
the antenna size is small at high spectrum
• Massive MIMO known also as 3D Nokia innovation examples
MIMO and full dimension MIMO • mmWave (70 GHz) PoC system
• Currently a study item in 3GPP for with DoCoMo
LTE-A • 3D MIMO leader in 3GPP
• Phased Array Architecture vs. Band • Leader in Channel modeling &
of Operation propagation measurements
• Baseband (1 transceiver/ant,~< 6GHz) 2x2 RFIC Dies

• Hybrid (N Ant/B RF chains, ~6- 30 GHz)

• RF (1 transceiver/RF beam, >30 GHz)
• Chip-scale array elements for
compact implementation at high Carrier plate onto which multiple RFIC die are bonded
frequency band
26 © Nokia 2015
Trends for MIMO/BF in 4G and 5G as BW Increases

< 6 GHz/low BW 6-55 GHz/moderate BW >55 GHz/high BW

Bandwidth Limited Huge Bandwidths
Interference Limited Noise Limited
Emphasis on Spectral Emphasis on Gain
Efficiency
Per-antenna channel Per-beam channel
knowledge knowledge
Baseband Architectures Hybrid / RF Architectures
Small Scale Arrays: Large Scale Arrays are
SU-MIMO sufficient required with an initial
Large Scale Arrays: emphasis on SU-MIMO
high-order MU-MIMO
27 © Nokia 2015
5G Proof-of-Concept (PoC) and Standards

28 © Nokia 2015
Nokia 5G mmWave beam tracking demonstrator

Mobile device
First 5G demos
CEATEC 2014
Access point

70 GHz band
1 GHz bandwidth

Lens antenna with 3˚ beam

64-beam width
switching

29 © Nokia 2015
 mmWave PoC System @ 2GHz BW supporting 10 Gbps Peak rate
New platform designed by NI to meet Nokia’s 5G specification

Parameters Value
74 GHz IF Baseband
Receiver Downconverter Receiver

Operating Frequency ~74 GHz

Processing Data

74 GHz IF Baseband Bandwidth 2 GHz

Receiver Downconverter Receiver

IF
Analog Digital Peak Rate ~10 Gbps
Baseband Baseband

74 GHz
Transmitter
IF
Upconverter
Baseband
Transmitter
Modulation Null Cyclic-Prefix Single
Carrier
Processing Data R=0.9, 16 QAM
2x2 MIMO
74 GHz IF Baseband
Transmitter Upconverter Transmitter Antenna Horn Antenna

10 Gbps peak rate using a prototype of NI’s mmWave platform- demonstrated at 5G Brooklyn
summit
30 © Nokia 2015
 Prototype of NI’s mmWave Platform at Brooklyn 5G

mmWave Realtime Software

NI PXIe Platform and mmWave RF Prototype

31 © Nokia 2015
ITU-R and 3GPP requirement work focuses on defining what is ‘Full 5G’
Initial commercial deployment requirements a subset

5G requirements define the system taking us past 2030

First deployments need only the subset

Phase 1 Phase 2
Driven by the commercial timeline (NGMN) Driven by the ITU-R submission schedule
• Commercial system ready in 2020 • Specification ready for submission in
• Standards ready end of 2018 2019

First specification and deployment phase 3GPP SRIT submission to ITU-R must fulfill
does not need to meet all the 5G all the 5G requirements defined by ITU-R
requirements defined by ITU-R and 3GPP and 3GPP

32 © Nokia 2015
3GPP timeline and 5G phasing
Phase 1 for 2020 deployment, Phase 2 for 2022/2023 and final ITU-R submission

Phase 1 specifications
should be completed
in 2018

P1 P2 Phase 2 specifications
should be completed
in 2019

How to map the 5G

timing and phasing to
3GPP releases?

33 © Nokia 2015
 Phase 1 specifications Phase 2 specifications
3GPP timelines should be completed should be completed
Longer or shorter 3GPP release cycles in 2018 in 2019

P1 P2
2015 2016 2017 2018 2019 2020

Rel-13 Rel-14 Rel-15

Rel-13 Rel-14 Rel-15 Rel-16

5G standard definition is LTE-A evolution track 5G Phase 1 and Phase 2

not the sole driver of the should not be availability needs to be
3GPP release cycle jeopardized matched with the LTE-A
release needs

34 © Nokia 2015
 Summary and Next Steps

35 © Nokia 2015
The Nokia way for the 5G Marathon
“If you want to go fast, go alone but if you need to go far, go together”

Outside in 5G • Collaborative research

e.g. 5G PPP, 863 5G
• Customer collaborations
• Drive regulatory and
industry work e.g. ITU-R

Collaboration
• University collaborations
e.g. NYU, TUD, Aalto etc.
Oulu University/CWC
cmWave channel modelling and measurements,
cmWave algorithms, spectrum sharing
Aalborg University Prof. Matti Latva-aho 1 PhD + joint project
Wide range of 5G radio research topics

Prof. Preben Mogensen 5 PhDs Tampere University of Technology

cmWave concepts and algorithms
University of CA, Santa Barbara

• Holistic systems research,

IC technologies for large antenna arrays at
Prof. Mikko Valkama 1 PhD
mmWave band New York University
Prof. James Buckwalter 2 PhDs mmWave channel modelling and measurements
Aalto University
Prof. Ted Rappaport 2 PhDs 5G radio system research on HetNet and mobility

Prof. Olav Tirkkonen 1 PhD

prototyping & development

Bristol University
Antenna and RF propagation modeling for HetNet Poznan University of Technology
Investigations of communication options and related
Prof. Andy Nix 1 PhDs 5G MAC/RRM design aspects for vehicular safety
Prof. K. Wesołowski 2 PhDs
University of Texas Universities in Beijing
5G modelling, D2D University of Kaiserslautern 5G radio research
5G network architecture

• Leverage One Nokia e.g.

Prof. Jeff Andrews 2 PhDs Technical University of Dresden
Prof. Hans Schotten 1 PhD Integrated Self-Organization Techniques for Uplink
using Force Field Network efficiency & quality

Purdue University Prof. Gerhard Fettweis 2 PhDs

Inside out 5G
mmWave access and backhaul
Technical University of Munich
Prof. David Love Analysis of Cooperating Schemes and Massive
1 PhDs
MIMO in Local Area Scenarios

Technologies
Prof. Gerhard Kramer 2 PhDs

36 © Nokia 2015
http://networks.nokia.com/innovation/5g
Q&A
37 © Nokia 2015