Understanding 5G radio

specifications and measurement

techniques.
Anritsu Spectrum Analyser Roadshow - 2017
 Introduction to 5G NR.
5G NR evolution from LTE.

CP-OFDMA
Scalable Numerology

Modulation types: BPSK/QPSK/16QAM/256QAM

FDD vs TDD
Mini-slot and self contained sub-frame

Carrier Aggregation.

Spatial Multiplexing MIMO

3D MIMO (Massive MIMO)

Slide Title
2 Copyright© ANRITSU
Air interface overview

Slide Title
3 Copyright© ANRITSU
Candidate frequency bands for 5G.

Slide Title
4 Copyright© Anritsu Corporation
 5G = LTE enhancements + Next Generation Radio Technology

HIGHER

Next Generation Radio Technology

5G
• Prioritize Performance to serve
New Radio (NR)
different use case
• Not Backward compatible to LTE
Performance

Big Gain
• Supports WRC15 & WRC19 Spectrum

LTE
LTEadvanced Advanced Pro
LTEadvanced Rel 12&13 Rel 14 & 15
Rel 10&11 Macro Assisted Continuous Evolution
LTE Rel 8&9 Small Cell
CA, eICIC, CoMP,
Enhancement
Hetnet
(Phantom Cell)
* LOWER ~2015 ~2020 Year

Slide Title
5 Copyright© ANRITSU
 5G New Radio technology components

300 MHz 3 GHz 30 GHz 300 GHz

New spectrum bands and access methods
Nomadic nodes

Buildings
Bus stop

Lamp posts
nodes

Dense and moving networks MIMO

Massive multi-antenna systems

Context-aware Air interfaces for new

interference and applications and
mobility management reduced signaling
Slide Title
6 Copyright© ANRITSU
Scalable Subcarrier spacing

• 15 KHz to 480 KHz – ranging from LTE macro cell to mmWave

Scalable Cell Bandwidth

• LTE to mmWave ranging from 20 MHz to 400 MHz

Scalable Framework
• Scalable TTI
• Integrated framework to support diverse deployment scenarios & network
topologies
• Dynamic switching between DL & UL based on traffic conditions

Slide Title
7 Copyright© ANRITSU
Physical Layer – general description
• NR supports paired and unpaired spectrum
• FDD operation on a paired spectrum. TDD operation on an unpaired spectrum

• Multiple numerologies are supported, derived by scaling a basic subcarrier spacing.

• The numerology used can be selected independently of the frequency band although it is assumed not to use
a very low subcarrier spacing at very high carrier frequencies.

• Flexible network and UE channel bandwidth is supported.

Key DL concepts
• The downlink transmission is CP-OFDM. QPSK, 16QAM, 64QAM and 256QAM (with the same constellation
mapping as in LTE).
• NR defines physical resource block (PRB) where the number of subcarriers per PRB is the same for all
numerologies.
• Multiplexing different numerologies within a same NR carrier bandwidth is supported for both downlink and uplink.

Key UL concepts
• QPSK, 16QAM, 64QAM and 256QAM (with the same constellation mapping as in LTE).
• CP-OFDM-based waveform is supported.
• DFT-S-OFDM based waveform is supported at least for eMBB uplink for up to 40GHz.
• CP-OFDM waveform can be used for a single-stream and multi-stream (i.e. MIMO) transmissions.
• DFT-S-OFDM based waveform is limited to a single stream transmissions (for link budget limited cases).

Slide Title
8 Copyright© ANRITSU
Scalable Numerology

Slide Title
9 Copyright© ANRITSU
OFDM (Orthogonal Frequency Division)

Parallel data transmission on several orthogonal subcarriers with lower rate

c
f
k3

Maximum of one subcarrier frequency appears exactly at a frequency

where all other subcarriers equal zero
 superposition of frequencies in the same frequency range
Amplitude subcarrier: sin(x)
SI function= x

Slide Title
10 Copyright© Anritsu Corporation
Flexible use of RF bandwidth – scalable numerology.

Efficient use of spectrum in different deployment scenarios.

Less control channel overhead.

Wideband carriers suited to higher data rates (shorter symbol time)

Slide Title
11 Copyright© ANRITSU
Scalable numerology – scaling of subcarrier spacing.

Change of sub-carrier spacing is possible, changing numerology “μ”.

Slide Title
12 Copyright© ANRITSU
Scaleable Numerology

Change of “μ” changes data rate on the sub-carrier, hence symbol (bit) rate.

Table 4.1-1: Supported transmission numerologies.

 f  2  15 [kHz] Cyclic prefix

0 15 Normal
1 30 Normal
2 60 Normal, Extended
3 120 Normal
4 240 Normal
5 480 Normal


Table 4.2.2-1: Number of OFDM symbols per slot, N symb , for numerology  and normal cyclic prefix.

 Slot configuration
0 1
 slots,  slots,   slots,  slots, 
N symb N frame N subframe N symb N frame N subframe
0 14 10 1 7 20 2
1 14 20 2 7 40 4
2 14 40 4 7 80 8
3 14 80 8 - - -
4 14 160 16 - - -
5 14 320 32 - - -

Slide Title
13 Copyright© ANRITSU
 Changing “u”, both subcarrier spacing and symbol rate

Wider carriers have shorter symbol rate (Nyquist sampling theory)

Slide Title
14 Copyright© ANRITSU
Subframe alignment of numerologies
All configurations are aligned on 1mS subframe.
So system can change numerology on each 1mS sub-frame interval.

Slide Title
15 Copyright© ANRITSU
Massive MIMO

Slide Title
16 Copyright© ANRITSU
Massive MIMO key concepts
Traditional ‘Spatial Multiplexing’ MIMO uses similar
number of Tx and Rx beams, and creates multiple
data streams by using channel estimation/coding.

Using a large number of antenna elements, we can create a narrow beam.

Can be applied to both Tx and Rx antennas.

When using a massive number of Tx elements versus number of required beams,

we can create and steer a number of beams simultaneously. [M >>K]

Slide Title
17 Copyright© ANRITSU
 Massive MIMO Antenna Architecture
• Radio baseband is augmented with many more connections into the antenna array subsystem
• TXRUs (Transmit/Receive Unit) connect into a RDN (Radio Distribution Network)
– RDN connects to the individual antenna elements
• The RDN enables a variety of beamforming techniques
– Digital BF, analog BF, hybrid BF
– Each connection architecture provide different capabilities and has different hardware, power and
complexity trade-offs

TXRU

TXRU Radio Physical

Baseband
Distribution Antenna
Unit
... Network Array

...
TXRU

Antenna Array Subsystem

Slide Title
18 Copyright© ANRITSU
 Possible Radio Distribution Networks

• All-analog beamforming
– Phase-shifting/weighting at RF
– One TXRU per desired beam

• All-digital beamforming
– One TXRU per antenna element
– Complete flexibility of beamforming
– Very expensive

• Hybrid analog/digital beamforming

– Both techniques used
– Supports SU-MIMO, MU-MIMO

IEEE Communications Magazine, Dec 2014 pg 111

Slide Title
19 Copyright© ANRITSU
 Some Possible TXRU / RDN Configurations
• 1-D Sub-array • 1-D Full-connection
– Wideband (analog) elevation steering – Wideband elevation steering
– Digital azimuth BF per subarray – Multiple azimuth beams
– Full spatial resolution of array
TXRU
TXRU BF N = NTXRU = # of columns
TXRU
TXRU N =NTXRU = # of columns
M = # of elements per column TXRU
TXRU
TXRU
TXRU BF M = # of elements per column
MTXRU = # of subarrays
TXRU
TXRU BF & MTXRU = # of array connections

...
TXRU
TXRU
P = # of polarizations
MTXRUNTXRUP = total TXRUs TXRU
TXRU S P = # of polarizations
TXRU
TXRU MTXRUNTXRUP = total TXRUs

• 2-D Sub-array
– Single wideband azimuth & elevation
• 2-D Full-connection
beam per subarray – Wideband elevation steering
– Wideband azimuth steering
TXRU BF N = # of columns
TXRU BF NTXRU = # of horizontal subarrays N = # of columns
M = # of elements per column TXRU BF M = # of elements per column
TXRU &

...
BF MTXRU = # of vertical subarrays P = # of polarizations

...
TXRU BF
P = # of polarizations S Total TXRUs depends on number of beams
TXRU
MTXRUNTXRUP = total TXRUs

Slide Title
20 Copyright© ANRITSU
Beam management concepts
Reference TR 38.80x
TRP sweeps beams in space, Switching TRP beams based UE switches beams based on
UE determines best beam on UE measurements. UE measurements.
pair to use.

• P-1: is used to enable UE measurement on • P-2: is used to enable UE • P-3: is used to enable UE
different TRP Tx beams to support selection measurement on different measurement on the same
of TRP Tx beams and UE Rx beams. TRP Tx beams to possibly TRP Tx beam to change
• For beamforming at TRP, it typically change inter/intra-TRP Tx UE Rx beam in the case
includes a intra/inter-TRP Tx beam sweep beam(s), from a possibly UE uses beamforming.
from a set of different beams. smaller set of beams for beam
• For beamforming at UE, it typically includes refinement than in P-1. Note
a UE Rx beam sweep from a set of different that P-2 can be a special case
beams. of P-1.

Slide Title
21 Copyright© ANRITSU
Industry schedule

Slide Title
22 Copyright© ANRITSU
 High-level schedule and milestones for 5G standards
2015 2016 2017 2018 2019 2020 2021 2022
Report IMT feasibility above 6 GHz
Recommendation vision of IMT beyond 2020
Approval of resolutions
WRC-15
Technical performance requirements
Report M.[IMT-2020. TECH PERF REQ]
Evaluation criteria and method
Report M.[IMT-2020. EVAL]
Requirements, evaluation criteria and submission template
Report M.[IMT-2020. SUBMISSION]
Circular letter “IMT-2020”
IMT-2020 workshop WRC-19
Proposals
Initial technology submission deadline
Evaluation
Outcome and decision
Specifications
Detailed specification submission deadline

RAN workshop
Channel model SI
Requirements SI
Rel-14
5G pre-release
Rel-15
5G ‘early drop’
5G first release
Rel-16
3GPP Release schedule shown is indicative 5G second release

First 5G networks, trials and Initial commercial launch with Commercial launch with
limited commercial use cases. a subset of features full feature set

Slide Title
23 Copyright© ANRITSU
5G NR radio testing and test solutions.

Slide Title
24 Copyright© ANRITSU
Anritsu 5G product portfolio

TRx device RF/L1 Integration Conformance Production

Vector Network Analyzer NR Radio communication

test station Conformance test
system

Production tester
NSA-NR test solution
Power Master
Signal Analyzer

OTA chamber
Spectrum Master

Slide Title
25 Copyright© ANRITSU
 5G R&D solution – Spectrum Master -
Key Feature
 mmWave capabilities for 5G, wireless backhaul, 802.11ad, Ultra portable spectrum
satcom, and more analyzer covering 9kHz to
 Ultraportable form factor enables measurements right at the 110GHz with excellent
device under test dynamic range
 Measure: channel power, adjacent channel power, occupied
bandwidth MS2760A
Spectrum Master
 Patented NLTL technology provides > 100 dB dynamic range
 Dynamic Range 103dB to 70GHz
 70 GHz sweep in under 8 seconds
 110GHz sweep in under 12 seconds
 -132 dBm DANL to 70 GHz
 -125dBm DALN to 110GHz
 Up to 6 traces, 4 trace detectors, and 12 markers

Slide Title
26 Copyright© ANRITSU
 5G R&D solution - Signal Analyzer -
Hardware capability
 Wideband analysis MS2850A
Signal Analyzer
 1GHz BW analysis bandwidth
 Wide dynamic range 9kHz–44.5GHz
 DANL -143 dBm/Hz @28GHz in 1GHz BW (meas.)
 SFDR - 70 dBc over 1GHz BW (nom.)
 Excellent Flatness Basic view Result summary
(Single carrier) (Single carrier)
 Amplitude ± 1.2 dB @28GHz, 1GHz BW (nom.)
 Phase ± 2.4° (4.8° p-p) @28GHz, 1GHz BW (nom.)
 Pure spectrum
 Phase Noise - 111 dBc/Hz @28GHz, 10 kHz offset (meas.)
 Fast analysis
 Built-in 5G ‘multi-carrier’ measurement software
 PCIe / USB 3.0 / Ethernet interface for data transfer
Resource block Time difference
domain view over carriers
Measurement software (Multi-carrier) (Multi-carrier)
 Remote controllable built-in application
 Single-shot multi-carrier simultaneous measurement
 Optimized settings and results for specification
 EVM and power for RS, xPDCCH, xPDSCH in V5GTF

CONFIDENTIAL
27 Copyright© ANRITSU
 High Speed Interface (PCIe, USB3.0)
 High-speed interfaces (PCIe, USB3.0) realize stress-free data transfer.
 Measurement speed can be improved with external high spec. PC.
 User can build optimized measurement environment on your PC easily by using Anritsu API.

Anritsu API realizes;

• Low Hardware Dependency
• Optimized Configuration
High-speed • High Expandability
interfaces
Signal Analyzer Apps Anritsu API
PCIe
Anritsu API USB3.0 VISA
1000Base-T (Std.)
Windows Windows
SCPI
I/Q Data

Remote Desktop
Windows Network Remote Control
MS2850A
Slide Title
(*1) API: Application
28 Programming interface Copyright© ANRITSU
High-speed transfer of captured data
• Data size of 1GHz BW digitized data can be large, using 32GB capture memory.
• High-speed interfaces (PCIe, USB3.0) provide high speed data transfer to external PC.

Transfer time comparison

(SPAN=1GHz,100ms IQ data)
43.5 sec
Stress-free data transfer
MS2850A
Signal Analyzer
PCIe
USB3.0
100 times
Faster

2.99 sec
0.42 sec

PCIe USB 3.0 1000Base-T

Slide Title
29 Copyright© ANRITSU
Testing of 5G NR

3GPP test items and specifications.

Specific measurement issues:
Phase flatness for massive MIMO.
Dynamic range and level settings for wide band EVM measurements.

Slide Title
30 Copyright© ANRITSU
 New methodology is required for 5G NR test
Adoption of millimeter-wave and beamforming is making OTA test more important than ever. Industry requires OTA-based
new methodology to replace conventional test and measurement approaches.

 Modulation
 Expand to 400MHz BW per carrier compared with 20MHz for LTE
 Achieve toward 8 Component carriers in mmWave, 2 Component carriers in Sub 6GHz
 Both Downlink and Uplink adopt based on OFDM modulation
 Expand radio of transmission bandwidth and channel width larger than 90%
 Adopt single carrier (DFT-s-OFDM) focus on coverage for UL
 Data throughput
 Initial target is 5Gbps by average, 10Gbps by peak rate
 Achieve beamforming by using massive MIMO
 End-to-End test by high data throughput and low latency
 Connection with DUT
 OTA environment is required for mmWave testing
 Evaluation for beamforming/searching/tracking function test
 Signalling test over the air by mobility/beam switching
 RF performance test over the Air
 Frequency accuracy, Max/Min Power, EVM, Spurious Emission RRM, Demodulation, Blocking
 Channel model: Geometry-based stochastic channel mode, three-dimensional channel model, large antenna array, large bandwidth
 Antenna characterization and calibration
 TRP, EIRP, EIS, Directivity, beam width
 Phase/gain calibration for array antenna element

Slide Title
31 Copyright© ANRITSU
Examples of 5G test cases
Mass-production test
Single carrier meas. 100 MHz BW f
- Channel power
- Frequency error
255 MHz BW
- EVM (Standard)

- OBW
Alt. Adj. Adj. Alt.
RF conformance test (L) (L) (U) (U)
+ Out-of-band emissions 100 MHz BW f
- ACP/ACLR
- Spectrum emission mask
510 MHz BW
(w/ Opt-033)

RF design and verification

+ Multi-carrier / Single band meas. 100 MHz BW x 8
f
- Amplitude / Phase flatness
- Time differences
1 GHz BW
(Opt-034)
Slide Title
32 Copyright© ANRITSU
5G NR testing example, TS38.141 Base Station tests

Conducted transmitter characteristics Radiated transmitter characteristics

Base station output power Radiated transmit power

Output power dynamics OTA Base station output power
Transmit ON/OFF power OTA Output power dynamics
- Transmitter transient period OTA Transmit ON/OFF power
Basic limit OTA Transmitted signal quality
non-AAS Minimum requirements OTA Unwanted emissions
AAS Minimum requirements - OTA Occupied bandwidth
Transmitted signal quality - OTA Adjacent Channel Leakage Power Ratio
Unwanted emissions (ACLR)
- Occupied bandwidth - OTA Operating band unwanted emissions
- Adjacent Channel Leakage Power Ratio (ACLR) - OTA Transmitter spurious emissions
- Operating band unwanted emissions OTA Transmitter intermodulation
- Transmitter spurious emissions
Transmitter intermodulation

Conducted receiver characteristics Radiated receiver characteristics

Reference sensitivity level OTA sensitivity

Dynamic range OTA Reference sensitivity level
In-band selectivity and blocking OTA Dynamic range
Out-of-band blocking OTA In-band selectivity and blocking
Receiver spurious emissions OTA Out-of-band blocking
Receiver intermodulation OTA Receiver spurious emissions
In-channel selectivity OTA Receiver intermodulation
OTA In-channel selectivity

Slide Title
33 Copyright© ANRITSU
Solutions for phase alignment measurement in Massive MIMO

Slide Title
34 Copyright© ANRITSU
 Excellent Phase Flatness
 5G uses Massive MIMO technology and some phase shifters exist for adjusting phase
difference between each antenna.
 The resolution of phase shifter is around 10° and smaller phase flatness is required.

Anritsu Competitor 1 Competitor 2

MS2850A
Flatness (Phase) 5˚ p-p 6˚ p-p 8˚ p-p

Measurement example
BTS TX block diagram

Phase Shifter
(Resolution = around 10°)
Slide Title
35 Copyright© ANRITSU
Amplitude and Phase Flatness (measured data)
帯域内位相直線性
Amplitude (Left)
Amplitude
5
Phase (Right)
3
4
3

Phase Error [degree]

2

帯域内周波数特性 [dB]
1
1.2 dB 2 5°p-p
Nominal+ 1 14950.70313

Nominal- 0 28225.70313
0
-1 39500.70313
14950MHz
-1 -2 nominal+
28225MHz

-2
1.2 dB 39500MHz
-3
nominal-

-4
-3 -5
-500 -400 -300 -200 -100 0 100 200 300 400 500 -500 -400 -300 -200 -100 0 100 200 300 400 500
Offset Frequency [MHz] Offset Frequency [MHz]
1 GHz BW

Slide Title
36 Copyright© ANRITSU
Necessity of phase difference measurement between antennas
Beam forming technology is one of the means to improve the energy efficiency of cellular networks.
Beamforming can reduce interference to other UEs by controlling the directivity with multiple
antennas and concentrating energy to a specific UE.

If the phase of each antenna is not aligned, the directivity will be weaken and the beamforming will
not work properly.

Thus, to measure and adjust the phase difference between the antennas is important.

Also, if the phase of the whole signal band is not aligned, the directivity of the
modulation signal will be weaken and the energy efficiency will be degraded.

Therefore, to confirm the phase difference of the whole signal band between the
antennas with good repeatability is important.

phase phase
freq. freq.

Same Difference phase

phase
freq. freq.

When the phase of antennas are aligned When the phase of antennas are NOT aligned

Fig.1 Beam forming technology

Slide Title
37 Copyright© ANRITSU
Methods of phase difference measurement between antennas

There are three methods to measure the phase difference between antennas.

(1) Use Vector Network Analyzer (VNA)

(2) Use Signal Analyzer (Dual Ports SA) and Signal Generator (SG)

(3) Use Signal Analyzer (Single Port SA) and Signal Generator (SG) <= MS2850A

Slide Title
38 Copyright© ANRITSU
(1) Use VNA
• VNA doesn’t have time variation of the phase because the same LO signal
is used for the transmitter and receiver.
• However, Signal Analyzer (SA) is also necessary for modulation analysis.

Merit
 No time variation of the phase
 Simple measurement setup

Demerit
 No function for modulation analysis (Need SA separately)
 To measure Antenna includes Up/Down Converter is difficult

Tx Ant 1

Tx Ant 2 Rx Ant Vector Network

RF In Analyzer (VNA)
...
.

.....
RF Out
Switch
Tx Ant N

Fig.2 Setup example of VNA

Slide Title
39 Copyright© ANRITSU
(2) Use Dual Ports SA and SG
• 2 ports SA doesn’t have time variation of the phase because two signals can be
measured at same time.
• However, it is not suitable for phase difference measurement between antennas
because the measurement setup is very complicated
Merit
 No time variation of the phase
 Modulation analysis capability
 Possible to measure the Antenna including Up/Down Converter
Demerit
 The switch circuit becomes complicate because signals from two Tx Ant simultaneously are necessary.
 To avoid crosstalk between Rx Ant 1 and 2 is difficult. (Probes may be necessary)
 Expensive

RF In
Tx Ant 1

...
(port 1)
. Cross
Rx Ant 1
Signal Switch Dual ports
talk RF In
Generator Tx Ant 2 SA
(port 2)
Ref In RF Out Rx Ant 2

.....
Ref Out
...
.

Tx Ant N
Switch
Fig.3 Setup example of Dual ports SA
Slide Title
40 Copyright© ANRITSU
(3) Use Single Port SA and SG (MS2850A)

• The measurement setup with Single port SA is simple and it has modulation analysis capability.
• Time variation occurs for the phase measurement but it can be reduced by devising the
measurement method.

Merit
 Modulation analysis capability
 Simple measurement setup
 Possible to measure the Antenna including Up/Down Converter

Demerit
 The time variation of the phase occurs →Possible to reduce

Tx Ant 1

Signal
Generator Tx Ant 2 Rx Ant Single port
RF Out RF In SA
...
Ref In
.

.....
Ref Out
Switch
Tx Ant N

Fig.4 Setup example of Single ports SA

Slide Title
41 Copyright© ANRITSU
How to reduce the time variation of the phase
Even if 10MHz frequency reference is connected between SA and SG, the phase varies over time. Fig.5
shows measurement result of phase vs. time when MS2850A (SA) and MG3697C (SG) are directly
connected.
The cause of the big phase variation is that the phase noise components are different on SA and SG. The phase
noise near the carrier affects the phase variation significantly.
The bad effect of phase noise can be eliminated by shortening the measurement interval and the phase variation
can be decreased dramatically.

In other words, shortening the switching

interval between antennas can reduce the
effect of phase variation.

The phase variation of MS2850A at every 0.3

ms is σ≦ 1 deg.
phase difference becomes
σ≦ 1 deg. by setting the switching interval to
σ≦ 0.3 ms.

Center Freq. : 28GHz

Span : 1GHz
measured data

Fig.5 The time variation of the phase (MS2850A)

Slide Title
42 Copyright© ANRITSU
Summary of measurement methods

“Single port SA” method (MS2850A) is the best choice when the issue of the time
variation of phase can be resolved.

Table.1 Summary of Measurement method

Best choice
Item VNA Dual port SA Single port SA
Simple measurement setup Good Bad Good
Modulation analysis capability Bad Good Good
Measure DUT including Up / Down Converter Bad Good Good
Price Bad†1 Bad Good
Time variation of phase Good Good Bad→Good†2

†1 SA is required for modulation analysis.

†2 Need to shorten measurement interval sufficiently.

Slide Title
43 Copyright© ANRITSU
Phase difference measurement using MS2850A (1/2)
The procedure for measuring phase difference using MS2850A

1. Output CW signal from SG.

2. Set Capture Length to N × T [ms] and start the capture.
*N : Number of antenna, *T : Switching cycle [ms]
3. Switch the route to Tx Ant 1.
4. Measure from Tx Ant 1 to Tx Ant N by switching the route with T[ms] period.
*Enough short switching time T [ms] to meet acceptable
time variation of phase is necessary.
*Switch Ant1 => Ant2 => Ant1 => Ant3 => Ant1=> Ant4 can realize
better measurement accuracy. Control

Tx Ant 1 MS2850A
MG3697C
Signal Analyzer
Signal Generator

Tx Ant 2 Rx Ant
RF Out RF In
...
Ref In
.

.....
Ref Out
Switch
Tx Ant N

Fig.6 Phase difference measurement using MS2850A

Slide Title
44 Copyright© ANRITSU
Phase difference measurement using MS2850A (2/2)
• The phase difference of each antenna can be calculated by removing the part of the
switching time and comparing the phase of each antenna.

T [s]

...

Switching Phase of Switching Phase of Switching Phase of Switching Phase of

time Tx Ant 1 time Tx Ant 2 time Tx Ant 3 time Tx Ant N

Fig.7 Measurement result image

Slide Title
45 Copyright© ANRITSU
Excellent Phase Flatness Repeatability

In the 5G system, ≧ 100 MHz wideband signals is used. If the phase of the whole signal band is not aligned, the directivity
of the modulation signal will be weaken and the energy efficiency will be degraded.

Phase flatness repeatability of MS2850A is σ=0.06 [deg.](Fig.9) at the same frequency offset position and it is very stable.
Therefore MS2850A realizes high accuracy phase difference measurement between antennas over whole signal band.

Center Freq. : 28.225GHz

Span : 1GHz Center Freq. : 28.225GHz
Temperature : 23 deg. (constant) Span : 1GHz
measured data Temperature : 23 deg. (constant)
measured data

Fig.8 Repeatability of phase flatness Fig.9 Repeatability of phase flatness

(1GHz BW) (each frequency offset)

Slide Title
46 Copyright© ANRITSU
Measurement result of phase time variation
Fig.A.1 shows phase time variation results when the measurement time is 100 ms, 25 ms, 5
ms, 0.5 ms.

Center Freq. : 28GHz

Span : 1GHz
Fig.A.1 Phase time variation result at each measurement time measured data

Slide Title
47 Copyright© ANRITSU
Configuring analyser for wideband EVM measurements

Slide Title
48 Copyright© ANRITSU
 Single Carrier Measurement with high accuracy
 Actual EVM performance of MS2850A is < 1% when measuring 100 MHz single
carrier of V5G waveform including 64QAM xPDSCH in 29 GHz.

Basic view Numeric result summary

(Single carrier) (Single carrier)

Slide Title
49 Copyright© ANRITSU
 Multi-Carrier (wide band) Measurement of EVM
 MX285051A supports multi-carrier (8 carriers)
simultaneous measurement.
 Frequency Error, Transmit Power, EVM and Time
difference of 8 carriers can be observed easily.

Summary view (Multi-carrier) Power vs. RB graph (Multi-carrier)

Slide Title
50 Copyright© ANRITSU
Configuring correct gain/level settings to minimise distortion effects on
EVM measurement for wideband 5G measurements.

•Predominant Cause of EVM

EVM measurement of OFDM signals can be degraded by the following issues with the measuring instrument
performance.

 SNR (Signal to Noise Ratio)

 Distortion
 Phase Noise

• Since the phase noise performance is unique to each instrument, it cannot be adjusted.
• However, the SNR and distortion can be optimized by adjusting measuring instrument settings.
• Generally, the dynamic range is best at the point where the measuring instrument distortion components
and noise floor are at the same level, which is when the EVM performance is best.
• However, since the integrated power of the noise increases in proportion to the signal bandwidth, the
measuring instrument relative dynamic range decreases as the signal band becomes wider.

Slide Title
51 Copyright© ANRITSU
Configuring correct gain/level settings to minimise distortion effects on
EVM measurement for wideband 5G measurements.

•Relationship between Signal Bandwidth and SNR

At a same signal Tx power, the power density drops as the signal band becomes wider.
For example, Table 1 lists the power density, SNR, and EVM due to the SNR when outputting signals under the following conditions for five
signal patterns: CW, W-CDMA, LTE (20 MHz BW), Verizon 5G (V5G, 1 Carrier), and Verizon 5G (V5G, 8 Carriers).
・Average Output Level: –10 dBm
・Average Noise Density: –150 dBm/Hz

For example, when the target EVM value is <1% (= 40 dB),

W-CDMA signal, the width of the 1% area is 64.2 dB – 40 dB ≒ 24 dB,
V5G (1 Carrier) signal, the area of <1% is narrower at 10.5 dB.
V5G (8 Carriers) signal, the 1% area is just 1.4 dB.

The settings for obtaining the optimum SNR for the V5G signal are much more difficult in comparison to the W-CDMA signal.

Communications Tx Bandwidth*1 Power Density SNR [dB] EVM due to SNR

Method [MHz] [dBm/Hz] [%]
CW - –10 140.0 -
W-CDMA 3.84 –75.8 64.2 0.06％
LTE 18 –82.5 57.5 0.13%
V5G (1 Carrier) 90 –89.5 50.5 0.30％
V5G (8 Carriers) 720 –98.6 41.4 0.85%

Slide Title
52 Copyright© ANRITSU
Spectrum Analyser level settings.

RF-
IF-ATT ADC
ATT

MIXER

Pre-AMP

Signal Analyzer MS2850A Simplified Block Diagram

The MS2850A has three kinds of level adjustment mechanisms:

・RF-ATT: Located at RF stage, it adjusts power input to Mixer and Preamplifier. It is effective when Mixer and Preamplifier have distortion.
・IF-ATT: Located at IF stage, it adjusts power of input to Analog Digital Converter (ADC). It is effective when the ADC overflows.
・Pre-AMP: Amplifies weak signals and improves Noise Figure (NF) of overall system. It is effective when wanting to lower the noise floor.

Unlike the RF-ATT and Pre-AMP, which the user can set arbitrarily, the IF-ATT cannot be set by the user.

The IF-ATT is set according to the following equations.

• At Pre-AMP = OFF
• IF-ATT [dB] = Reference Level – RF-ATT + 10; when, 0 dB ≤ IF-ATT ≤ 10dB
• At Pre-AMP = ON
• IF-ATT [dB] = Reference Level – RF-ATT + 30; when, 0 dB ≤ IF-ATT ≤ 10 dB

For example, at Pre-AMP = OFF, when Reference Level = –3 dB and RF-ATT = 4 dB, IF-ATT [dB] = –3 – 4 +10 = 3 dB.

Slide Title
53 Copyright© ANRITSU
Measurement with Pre-AMP = OFF

The “Level Over” message is displayed on-screen when the ADC overloads, (error in measurement data).

MS2850A ADC Overload Level Target Value at RF-ATT = 0 dB, IF-ATT = 0 dB (Signal Analyzer Mode)

SPAN ≤31.25 MHz ≤31.25 MHz ≥50 MHz ≥50 MHz

Pre-AMP Off On Off On
Overload Level –7 dBm –27 dBm 0 dBm –20 dBm
Target Value

ADC Overload Levels at Each Condition

Example 5G signal with an average power of –4 dBm and a crest ratio of 14 dB to the MS2850A.
The peak power of the above signal is +10 dBm (= –4 dBm + 14 dB).
The MS2850A has ADC overload level target value of 0 dBm when the MS2850A is in Signal Analyzer mode with a SPAN setting of
≥50 MHz and RF-ATT/IF-ATT = 0 dB,

The total of the inserted RF-ATT and IF-ATT must be greater than 10 dB (10 dBm – 10 dB = 0 dBm) to prevent ADC overload.

Ref. 0 dBm 0 dBm 0 dBm 0 dBm 0 dBm 0 dBm

Level
RF-ATT 10 dB 8 dB 6 dB 4 dB 2 dB 0 dB
IF-ATT* 0 dB 2 dB 4 dB 6 dB 8 dB 10 dB

RF-ATT and IF-ATT Combinations

* IF-ATT is determined by the Reference Level and RF-ATT, and cannot be set by the user.

Slide Title
54 Copyright© ANRITSU
 Measurement with Pre-AMP = OFF
RF-ATT = 10 dB, IF-ATT = 0 dB RF-ATT = 0 dB, IF-ATT = 10 dB

SNR (CC7) EVM (CC7)

RF-ATT = 10 dB, IF-ATT = 0 dB –31.39 dB 2.98%
RF-ATT = 4 dB, IF-ATT = 6 dB –33.11 dB 2.36%
RF-ATT = 0 dB, IF-ATT = 10 dB –33.13 dB 2.35%

SNR and EVM Measurement Results at Each ATT Setting

 SNR and EVM are both improved as the RF-ATT value becomes smaller and the IF-ATT value becomes larger.
• NF of the front end has a large impact on the NF of the overall system.
 Inserting a 10 dB IF-ATT improves the NF of the overall system more than inserting a 10 dB RF-ATT.
• The EVM is improved by the better SNR.
 Setting RF-ATT = 0 dB degrades the VSWR due to Mixer and Pre-AMP matching to the device under test.
• Recommend setting RF-ATT = 4 dB.

Slide Title
55 Copyright© ANRITSU
Measurement with Pre-AMP = ON (Example of SNR dominance)
• Lower signal levels, e.g. a 5G signal with an average signal level of –19 dBm and a crest ratio of 14 dB.
• The peak power is –5 dBm (= –19 dBm + 14 dB).
• When the MS2850A with a SPAN setting of ≥50 Hz, RF-ATT/IF-ATT = 0 dB, the ADC overloads at 0dBm target.
• Measurement is possible without ADC overload when Reference Level = –10 dBm, RF-ATT = 0 dB (IF-ATT = 0 dB).
• However, lowering the average signal level by –20 dBm has an impact on the noise floor and degrades the SNR.
• When RF-ATT = 4 dB, since the NF of the overall system becomes 4 dB worse, the SNR is also degraded by 4 dB.
RF-ATT = 0 dB, IF-ATT = 0 dB RF-ATT = 4 dB, IF-ATT = 0 dB

SNR (CC7) EVM (CC7)

RF-ATT = 0 dB, IF-ATT = 0 dB –25.73 dB 5.65%
RF-ATT = 4 dB, IF-ATT = 0 dB –22.78 dB 8.02%

SNR and EVM Measurement Results at each ATT Setting

 SNR is degraded by the noise-floor effect as the signal level decreases, setting Pre-AMP = ON can improve the SNR.
 Pre-AMP is located immediately downstream of the RF-ATT, setting it to ON improves the NF of the entire system and
lowers the noise floor.

Slide Title
56 Copyright© ANRITSU
 Measurement with Pre-AMP = ON (Example of SNR dominance)

RF-ATT = 6 dB, IF-ATT = 9 dB RF-ATT = 14 dB, IF-ATT = 1 dB

SNR (CC7) EVM (CC7)

ADC overload at –19 dBm when Pre-AMP = On.
RF-ATT = 14 dB, IF-ATT = 1 dB –27.46 dB 4.32%
The total of RF-ATT and IF-ATT must be at least 15 dB (–5dBm – (–20dBm)).
RF-ATT = 10 dB, IF-ATT = 5 dB –28.88 dB 3.68%
RF-ATT = 6 dB, IF-ATT = 9 dB –29.60 dB 3.35%

Example of RF-ATT and IF-ATT Combinations

Ref. Level –15 dBm –15 dBm –15 dBm –15 dBm –15 dBm
IF-ATT[dB] = Reference Level – RF-ATT + 30; when, 0 dB ≤ IF-ATT ≤ 10 dB RF-ATT 14 dB 12 dB 10 dB 8 dB 6 dB
IF-ATT*5 1 dB 3 dB 5 dB 7 dB 9 dB

• Pre-AMP = ON
 Setting Pre-AMP from OFF to ON improves the SNR and the EVM improves from 5.65% to 3.35%.
 Also when IF-ATT is larger than RF-ATT, the overall system NF is improved and the EVM becomes better.

Slide Title
57 Copyright© ANRITSU
Measurement with Pre-AMP = ON (Example of Distortion dominance)
SNR EVM
Single Carrier signal with an average signal level of –19 dBm and a crest ratio of 14 dB. RF-ATT = 14 dB, IF-ATT = 1 dB 41.51 dB 1.07%
SNR and EVM Measurement Results at Each ATT Setting RF-ATT = 10 dB, IF-ATT = 5 dB 43.00 dB 1.02%
RF-ATT = 6 dB, IF-ATT = 9 dB 43.19 dB 1.14%

• SNR and EVM show an improvement when RF-ATT is smaller and IF-ATT is larger with Pre-AMP = OFF or Pre-AMP = ON and 8 Carrier,
• EVM optimum conditions are different at Pre-AMP = ON and Single Carrier due to the effect of measuring-instrument distortion.
• At the same output level, a Single Carrier signal has a power density of 8 times compared to an 8 Carrier signal.
• EVM deteriorates as the effect of Pre-AMP distortion increases.

The waveforms at RF-ATT = 6 dB and 10 dB are shown below where the red ovals clearly indicate the Pre-AMP signal distortion when RF-ATT = 6 dB.
The Pre-AMP distortion characteristics change with frequency, bandwidth, and crest ratio, so the RF-ATT should be adjusted while observing the waveform.

ATT = 6 dB ATT = 10 dB

 Setting Pre-AMP to ON improves the SNR and the EVM improves from 5.65% to 3.35%.
 When IF-ATT is larger than RF-ATT, the overall system NF is improved and the EVM becomes better.
Slide Title
58 Copyright© ANRITSU
Summary

When measuring the EVM of wideband signals as used in 5G; more accurate EVM measurements can be
obtained by measuring in accordance with the following notes.
 SNR optimization is difficult with wideband modulation signals and requires careful adjustment.
 An ATT must be inserted to prevent the modulation waveform peak power exceeding the ADC overflow
level (high crest factor of the waveform)..
 When the SNR effect is greater than the distortion effect, inserting a larger IF-ATT improves the SNR, but
we recommend inserting an RF-ATT of about 4 dB due to the degraded VSWR.
 In particular, since EVM is degraded by the effect of distortion when Pre-AMP = On, it is necessary to
adjust the ATT to optimize the measurement accuracy.

Slide Title
59 Copyright© ANRITSU
TR 38.811
Study on NR to support non-terrestrial networks

TR 38.812
Study on Non-Orthogonal Multiple Access (NOMA) for NR

TR 38.813
New frequency range for NR (3.3-4.2 GHz)

TR 38.814
New frequency range for NR (4.4-4.99 GHz)

TR 38.815
New frequency range for NR (24.25-29.5 GHz)

TR 38.816
Study on CU-DU lower layer split for NR

TR 38.817-01
General aspects for UE RF for NR

TR 38.817-02
General aspects for BS RF for NR

TR 38.818
General aspects for RRM and demodulation for NR

TR 38.874
NR; Study on integrated access and backhaul

TR 38.889
Study on NR-based access to unlicensed spectrum

TR 38.900
Study on channel model for frequency spectrum above 6 GHz

TR 38.901
Study on channel model for frequencies from 0.5 to 100 GHz

TR 38.912
Study on new radio access technology

TR 38.913
Study on scenarios and requirements for next generation access technologies

Slide Title
61 Copyright© ANRITSU
Configuring for wideband EVM measurements.

Slide Title
62 Copyright© ANRITSU
Wide Dynamic Range (1GHz BW)
 MS2850A has 9 dB better dynamic range than Competitor for 1GHz BW
measurement.
MS2850A Competitor 1
Over Flow Level @28GHz 0 dBm -14 dBm
DANL @28GHz -142 dBm/Hz -147 dBm/Hz

Dynamic Range 142 dB 133 dB

(Over Flow Level – DANL)

High
Over flow level

Measured level MS2850A

9 dB better Competitor

Low
DANL

63 Copyright© ANRITSU CORPORATION

Dynamic Range and DANL
MS2850A would have advantage of 8 dB and over than Competitor in dynamic
range because advantage in maximum input level avoiding overflow is much higher
than disadvantage in noise floor.
Anritsu MS2850A Competitor 1
Over Flow Level@28GHz 0 dBm -14 dBm
Noise floor @28GHz -142 dBm/Hz -147 dBm/Hz (Nominal)
Dynamic Range 142 dB 133 dB

Measured level
Specification in configuration without Preamplifier would High
be -141 dBm/Hz at 28 GHz band. Over flow level
(from preliminary datasheet of MS2850A)

MS2850A
Wider dynamic range Competitor

Low
DANL

Slide Title
64 Copyright© ANRITSU
Best SFDR (Spurious Free Dynamic Range)
 MS2850A has best SFDR with 1GHz BW.
 Signal analyzer has ADC (Analog to Digital Converter) inside.
 The ADC generates spurious but the level of MS2850A is quite low.
Anritsu Competitor 1 Competitor 2
MS2850A
SFDR - 70 dBc -56 dBc -65 dBc

Spurious generated by ADC

Competitor has 14dB higher

spurious than MS2850A.

65 Copyright© ANRITSU CORPORATION

SFDR (Spurious Free Dynamic Range)
When you consider dynamic range, not only DANL, take care SFDR (Spurious Free Dynamic
Range) when you choose a wideband signal analyzer having ADC (Analog to Digital
Converter) inside because it can't avoid spurious comes from the processing theoretically.
Anritsu MS2850A has excellent SFDR performance in the current analyzers.

For your reference, SFDR of Keysight UXA is

-56 dBc in 1 GHz analysis bandwidth (same
condition as the left picture)

Slide Title
66 Copyright© ANRITSU
Noise Floor Reduction (NFR) *SPA mode only
 Noise Floor Reduction realizes 7 – 11 dB better noise floor by
subtracting noise inside Spectrum Analyzer.
 Able to measure very small signal accurately.
 Useful for Spurious measurement.
Better noise floor Accurate measurement

-81 dBm

-89 dBm

Input signal level: NFR Off : -81 dBm

7 – 11 dB better NFR On : -89 dBm
-89 dBm
Slide Title
67 Copyright© ANRITSU