4. 2019-07-03 | | Page 4
Phase 1: Service Setup
5G on Mid-bands
Dual connectivity with
5G on mid bands
Shared low
bands
Increased coverage
for wide-area &
outside in coverage
5G on High
bands
Increased capacity
and ultra low latency
Shared mid bands
Maximal coverage,
capacity and cell edge
performance
High bands (24 GHz – 40 GHz)
Mid bands (3.5 GHz – 8 GHz)
Mid bands (1 GHz – 2.6 GHz)
Low bands (sub –1 GHz)
Dual connectivity
Spectrum sharing
Carrier aggregation
Performance characteristics
Baseline
Capacity and coverage
Cell edge performance
Capacity/Speed
Latency
2G +3G 4G 5G 4G+5
G
Cell edge performance
Capacity/Speed
Latency
Cell edge performance
Capacity/Speed
Latency
Cell edge performance
Capacity/Speed
Latency
Cell edge performance
Capacity/Speed
Latency
Discussing the spectrum strategy
5. 2019-07-03 | | Page 5
• 4G and 5G in mid-low bands, same coverage area
• Both technologies share the same radio site,
connected to the existing Core network
• Example of use cases: eMBB, FWA in wide areas
• 4G in low bands and 5G in high-bands, different coverage
areas
• 5G radios may be deployed in new site as needed
• Both technologies are connected to the existing Core
network
• Example of use cases: eMBB, FWA in selected areas
• Initial 5G deployments in low bands, benefit from
larger coverage areas
• New 5G radio sites, connected to the new 5G Core
• Example of use cases: eMBB, FWA, Industrial IoT
Non-standalone in low-mid bands
Non-standalone in mid-high bands
Standalone 5G
Phase 1: Service Setup
4G in low-mid bands 5G low-mid bands 5G in high bands
1
2
3
Different deployment strategies
6. 2019-07-03 | | Page 6
Accurate AAS Modeling
Individual beamforming and traffic pattern files are imported into the planning
tool.
Beamforming technology enables the use of a single antenna to address different
problems by configuring the broadcast beams to address different scenarios
e.g. macro, hotspot or high rise.
Antenna VH pattern
(from AAS vendor )
Broadcast patterns
- including beamswitching
(from AAS vendor)
High-rise (HPBW:
H = 20°, V= 30°)
Macro (HPBW:
H = 65°, V= 10°)
Hotspot (HPBW:
H = 65°, V= 30°)
Tilt = A Tilt = B
Beamforming patterns by tilt
Phase 1: Service Setup
7. 2019-07-03 | | Page 7
Based on network observations we can see
how traffic increases on network level
Investigate type of traffic
• Bursty (less good for MU-MIMO)
• Mobility in the cell (less good for MU-
MIMO)
Network segmentation will also show cells
with capacity constrain today
Apply growth will show how many cells that
are exposed for capacity constrains over time
when traffic grow
Phase 2: LTE Assessment
Traffic growth and type 2 years observation
8. 2019-07-03 | | Page 8
•Site list
•Sector configuration
•Power split
CM data
•Air Interface limitations: PRB usage,
Scheduling Entities, PDCCH CCEs
•KPI correlations: Rank and modulation
usage, spectral efficiency, user
throughput, active users, RRC
connections, average CQI…
PM data
•Geolocation
•User spread
CTR data
Phase 2: LTE Assessment
Capacity analysis (II) - Methodology
Traffic map
KPIs correlations
Decision Tree
9. 2019-07-03 | | Page 9
Phase 3: eMBB Nominal design
Link Budget – Radio dimensioning
process
— Usually the objective is to find the cell
range where quality requirements are
met
— e.g. Uplink and downlink bitrates
— Can also set cell range and calculate
bitrates.
— More common with overlay
network designs
— A Link Budget is used for coverage
— The Ring Method is used for Capacity
and throughput profile analysis
Finding the cell range is
automated in RNPT eMBB
using the Cell Range Solver
10. 2019-07-03 | | Page 10
Phase 3: eMBB Nominal design
Problem areas
Identification of issues per pixel to
be addressed including severity:
— Coverage hole
— Congestion hotspot
— Street-level
— Indoor
— High-rise
Candidates
— Which frequency band?
— Which RAT?
— Which solution &
scenario?
Ranking
Weighting each of the candidate locations & solutions
against technical and business criteria (candidates
with higher impact are ranked first):
— Coverage improvement (% traffic covered)
— Capacity (cell capacity increase, MU-MIMO utilization
%)
— Performance (SINR / throughput)
— Cost (specific location & solution)
11. 2019-07-03 | | Page 11
Phase 4: eMBB Site selection
Final selection
Choosing the final group of candidates to be deployed, from the total list that
is generated by ranking solutions according to their performance impact
(higher impact first), makes it possible to balance between performance &
business criteria
Hotspot 1
NR M
-MIMO
NR 4T4R
High rise 1
NR M
-MIM
O Cost
Performance
LTE sm
all cell
Macro 1
NR M-MIMO
Hotspot 2
NR M
-MIMO
High rise 3
NR M-MIMO
Hotspot 3
NR M
-MIMO
13. Ericsson Internal | 2018-02-21
Updates in NR scheduling
— Based on Use cases : URLLC, eMBB
— Prioritized SR
— Flexible HARQ timing
— Configurable parameters(K0,K1,K2,K3) can handle low latency
— Scheduling for Low latency support
— Front-loaded reference signals and control signaling – not using time interleaving
across OFDM symbols device can start decoding without buffering
— Mini-slot transmission(K0)
— Transmission over a fraction of a slot
— HARQ acknowledgement, approximately one slot(even less depending on UE
capability)(K1)
— UL grant to PUSCH(K2)
14. Ericsson Internal | 2018-02-21
Updates in NR scheduling
Beamforming is heavily used in NR, Beam centric design and multi-antenna
transmission
In High frequency band, it extends coverage
Lower-frequency band, full dimensional MIMO
Aiming for beamforming gain(SNR) and spatial multiplexing gain
Channels and signals have been designed for beamforming, SSB, RACH, no
CRS
Beam Failure Recovery
Prioritized Random Access
Beam Failures, Handover
Power ramping, Back-off Faster
SNR
Cap SNR
C
SNR
C
C = log2(1 + SNR)
(bits/s/hz)
<
Increase capacity using
spatial multiplexing gain
15. Ericsson Internal | 2018-02-21
Updates in NR scheduling
Large transport block size performance degradation in retransmission in LTE
Code Block Group(CBG)based transmission in NR provides finer granularity, various QoS
supporting(preemption)
Asynchronous and adaptive UL HARQ
Scheduling of Reference signals
Lean design
On-demand reference signals for time/frequency offset tracking, beam fine tuning
No CRS, beamformed, Channel becomes concrete mMIMO
BWP(Bandwidth Part)
To support UE that doesn’t support full-BW
16. Ericsson Internal | 2018-02-21
Scheduling functions
Priority handling between UEs
Priority handling between logical channels
Channel dependent dynamic scheduling to combat frequency selective fading, distance
dependent pathloss and random interference variations and to maximize spectral efficiency
Inputs
BSR, SR, CQI, RI, PMI, UL SINR, BLER, Orthogonal factors for MU-MIMO, age of packets, UE
throughput, QoS requirements for DRBs, Doppler spread
Decisions
Resource allocations(Frequency, time, beam(space)), MCS, Transmission mode(including SU
or MU MIMO), Tx Power
18. Ericsson Internal | 2018-02-21
Background
Capacity of Ericsson’s NR product is increasing for meeting customer demands.
18.Q4:
Support 5 connected users
Schedule 1 UE’s DL data transmission and 1 UL data transmission
Each user’s DL data transmission contains 1 radio data bearer
19.Q2:
Support 50 connected users
Schedule up to 2 user’ DL data transmission and/or up to 2 UL data transmission
Each user DL data transmission contains up to 2 radio data bearers
19. Ericsson Internal | 2018-02-21
Feature Overview
This feature enables the following functionality:
Allow maximum 50 Users to connect an NR gNB
Scheduler schedules up to 2 user’ DL data transmission and/or up to 2 UL data transmission
Each user DL data transmission contains up to 2 radio data bearers
Resource fair algorithms is used for UE scheduling priority decision
SU MIMO for resource allocations
No configuration related parameter to be added/removed/modified
No performance management parameter to be added/removed/modified
#5:Talk points:
In general terms, different frequencies determine the coverage area…not necessarily the technology.
In this slide we try to illustrate 3 typical scenarios expected to be somewhat common:
1) when both 4G and 5G radios are deployed in the same frequency, meaning that both will have very similar coverage area,
2) when 4G is deployed low frequencies, while 5G is deployed in high frequencies. This will make 5G coverage smaller than 4G coverage for example
3) when 5G standalone is deployed in low bands for example, with no related 4G coverage (i.e. at least not integrated with 5G network).
These 3 cases can help us to draw some conclusions described in the slides in terms of deployments examples:
1) using non-standalone 5G, same coverage area as 4G, where all the sites can be physically collocated.
2) non-standalone 5G, different coverage area, where 4G and 5G sites will be majority collocated as case 1, but also few 5G radios would be needed to be deployed in between 4G sites, as smaller coverage leads to holes that sometimes need to be covered for specific reasons.
3) standalone 5G, which is not integrated with 4G, will require dedicated deployment