Part 9 Technique for 5G and Beyond

Definition of 5G

Recommendation ITU-R
M.2083-0, “IMT Vision –
Framework and overall
objectives of the future
development of IMT for
2020 and beyond,” p. 14,
September 2015.

© Po-Ning Chen@ece.nctu IDC9-2

Definition of 5G
o Peak data rate:
n max rate per user under ideal conditions
n 10 Gbps for mobiles, and 20 Gbps under certain conditions
o User experienced data rate:
n 95% rate across the coverage area per
user
n 100 Mbps in urban/suburban areas, and
1 Gbps hotspot
o Spectrum efficiency:
n Throughput per Hz per cell
o Mobility:
n max speed at which seamless handover and QoS is guaranteed
© Po-Ning Chen@ece.nctu IDC9-3
Definition of 5G
o Latency:
n radio contribution to latency between transmit and
receive
o Connection density:
n devices per km2
o Network energy efficiency:
n Include network bits/Joule and
user bits/Joule
o Area traffic capacity:
n Throughput per m2

© Po-Ning Chen@ece.nctu IDC9-4

 Key Applications and Schedule
o eMBB (enhanced Mobile BroadBand)
o URLLC (Ultra-Reliable and Low Latency Communications)
o mMTC (massive Machine Type Communications)

2018 2019 2020 2021

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
Rel-15 Rel-15 Rel-15 Rel-15 Rel-15 Rel-15 early drop: New Radio, non-stand alone
early drop early drop main drop main drop late drop late drop main drop: stand alone
freeze ASN.1 freeze ASN.1 freeze ASN.1

Rel-16 Rel-16
Rel-16
content Rel-16 specification development ASN.1
freeze
definition freeze
Rel-17 Rel-17
Rel-17
content Rel-17 specification development ASN.1
freeze
definition freeze

© Po-Ning Chen@ece.nctu IDC9-5

Key Applications
o eMBB: human centric communication
n Better mobile phones and hot spots, high data rates, high
user density
o URLLC: human and machine communication
n Vehicle-to-vehicle communication, industrial IoT, 3D
gaming
o mMTC: machine centric communication
n Very large number of devices, low data rate, low power,
IoT with long battery life time

© Po-Ning Chen@ece.nctu IDC9-6

Schedule
o No major changes are done after a release is frozen.
o Abstract Syntax Notation One (ASN.1) is the notation used
to specify message formats in the final specification.
o Release 15 has 3 stages: Non-standalone (NSA), Standalone
(SA), and Late Drop.
o Rel-15 early drop versus main drop
n Non-Standalone (NSA): Use 5G RAN + 4G Core
o Help accelerate 5G NR deployment
n Standalone (SA): Full 5G RAN + 5G Core

© Po-Ning Chen@ece.nctu IDC9-7

Spectrum for 5G
o Two Frequency Ranges (FRs) specified in Rel-15
n FR1: 450-6000 MHz (Sub 6-GHz)
n FR2: 24.25-52.6 GHz (mm-Waves)
o Good for high throughput in small cells

© Po-Ning Chen@ece.nctu IDC9-8

 Scalable OFDM in 5G
o 4G specifies a fixed 15 kHz subcarrier spacing (SCS).
o 5G can have many SCSs.
n 2n×15 kHz : 15, 30, 60, 120 kHz for n = 0, 1, 2, 3
o R15 allows 15/30/60 kHz SCS for FR1 and 60/120 kHz for FR2.
15 KHz spacing < 3GHz macro cells Subcarrier spacing (KHz) 2n × 15
Symbol duration (us) 66.67 / 2n
30 KHz spacing > 3GHz small cells Cyclic prefix (us) 4.69 / 2n
Symbol + CP (us) 71.36 / 2n
60 KHz spacing 5 GHz Unlic. Indoor Symbols per slot 14
Slot duration (us) 1000 / 2n

120 KHz spacing mmWave small cells + backhaul

© Po-Ning Chen@ece.nctu IDC9-9

Scalable OFDM in 5G
o Max FFT size 4096
n Max 3300 subcarriers for 50, 100, 200, 400 MHz band

4G (LTE) 5G NR 5G NR
Channel Bandwidth (MHz) 20 20 2n × 50
FFT Size 2048 2048 4096
Number of Subcarriers for 15 KHz Spacing 1200 1200
Number of Subcarriers 3300
for 2n × 15 KHz Spacing, n = 0, 1, 2, 3

© Po-Ning Chen@ece.nctu IDC9-10

Frame Structure

frame (10 ms)

subframe
0 1 2 3 4 5 6 7 8 9
1 ms
The number of slots per subframe depends on
slot .. slot the subcarrier spacing.

symbol Subcarrier Slot Slots/subframe

0 1 .. 12 13 spacing (KHz) Duration (us)
or 10 11 15 1000 1
30 500 2
60 250 4
120 125 8

© Po-Ning Chen@ece.nctu IDC9-11

Physical Resource Block (PRB): Example
1 slot = 0.25 ms

Frequency
o 4G specifies PRB = 12
subcarriers × 6 or 7
symbols
720 KHz SCS = 60 KHz
o 5G specifies PRB = 12
subcarriers × 12 or 14
symbols 1 slot = 0.5 ms

Resource Element 360 KHz SCS = 30 KHz

1 slot = 1 ms
180 KHz SCS = 15 KHz
Time
© Po-Ning Chen@ece.nctu IDC9-12
Non-Orthogonal Multiple Access
o The development of NOMA in 3GPP
n Release-12 Network-Assisted Interference Cancellation and
Suppression (NAICA) in 2014
n Release-13 Downlink Multiuser Superposition Transmission
(MUST) in 2015
o Focus on downlink
o Power-domain NOMA
n Release-15 NOMA for NR in 2018
o Focus on uplink
o Power-domain NOMA, interleave division multiple access
(IDMA), multi-user shared access (MUSA), sparse code
multiple access (SCMA)…

© Po-Ning Chen@ece.nctu IDC9-13

o
Non-Orthogonal Multiple Access
o Massive IoT technologies for 3GPP Rel-16
n Can be either scheduled or grant-free
n Increase device density and network efficiency

Power Power
OMA NOMA

Frequency Frequency

Successive Interference
Cancellation (SIC) or joint
maximum-likelihood (JML)

© Po-Ning Chen@ece.nctu IDC9-14

NOMA for Multiple Access Channel

OMATDMA:

OMAFDMA: (See Slide IDC6-60.)

Rate of User 2

NOMA

OMAFDMA
OMATDMA

NOMA:
Rate of User 1

© Po-Ning Chen@ece.nctu IDC9-15

NOMA for Broadcast Channel

Rate of User 2 NOMA

OMATDMA

Rate of User 1

© Po-Ning Chen@ece.nctu IDC9-16

Channel Coding in 5G
o (Quasi-cyclic) multi-edge low-density parity-check code
(LDPC) for data
n Less complex than 4G turbo codes
n Good for high data rates
o Polar code for control
n Contrary to 4G tail-biting convolutional codes (TBCC)
n Use CRC for joint detection and decoding

© Po-Ning Chen@ece.nctu IDC9-17

5G Trials
o Many operators have announced 5G trials
o Verizon, SK Telecom, Korea Telecom, NTT DoCoMo,
AT&T, China Mobile, ...
o Most are using sub-6GHz spectrum

© Po-Ning Chen@ece.nctu IDC9-18

Non-Terrestrial Transmissions towards 6G
o Use of satellites with 5G
n Not for high throughput but for continuity of coverage.
n PHY retransmission procedure shall be more delay tolerant.
n Impact of propagation
delays should be studied. LEO satellite
o Low-earth orbit (LEO)
500-1000 km
o Geo-stationary earch
orbit (GEO) 35786 km
n Handover and paging

© Po-Ning Chen@ece.nctu IDC9-19

Vision of 6G
5G 6G
Traffic Capacity 10 Mbps/m2 1-10 Gbps/m3
Data rate DL 20 Gbps 1 Tbps
Data rate UL 10 Gbps 1 Tbps
Uniform user experience 50 Mbps 10 Gbps
2D everywhere 3D everywhere
Latency (radio interface) 1 ms 0.1 ms
Reliability (frame error rate) 10-5 10-9
Localization precision 10 cm on 2D 1 cm on 3D
[1] D. Klaus and B. Hendrik, “6G vision and requirements: Is there any need for beyond
5G?,” IEEE Veh. Technol. Mag., Sept. 2018.
[2] E. C. Strinati, S. Barbarossa, J. L. Gonzalez-Jimenez, D. Ktenas, N. Cassiau, L. Maret and
C. Dehos, “6G: The next frontier: From holographic messaging to artificial intelligence using
subterahertz and visible light communication,” IEEE Veh. Technol. Mag., Sept. 2019.

© Po-Ning Chen@ece.nctu IDC9-20

Vision of 6G
o Sub-THz communications
o In < 10 meters of line-of-sight communication range, one Tera
bps is theoretically possible by a single link with hundreds GHz
of carrier frequency.
o Visible light communication (VLC)
o VLC over short range (few meters) links is perhaps a
technology option that can provide optical-fiber comparable
performance.
o Compared to RF band communication, VLC offers ultra-high
bandwidth (Tera Hz), zero electromagnetic interference, free
unlicensed abundant spectrum, and very high frequency reuse.

© Po-Ning Chen@ece.nctu IDC9-21

Example of Polar Codes
for the Binary Erasure Channel (BEC)
o A new channel code invented by E. Arikan in 2008.
o The idea behind polar codes is channel polarization, which
transforms n uses of BEC(!) into extremal polarized channels;
i.e., channels which are either perfect (noiseless) or completely
noisy.
o It is shown that as n goes to infinity, the number of unpolarized
channels converges to zero and the fraction of perfect channels
converges to I(X;Y) = 1 - ! under a uniform input, which is the
capacity of the BEC.
o A polar code can be obtained by sending information bits directly
through those perfect channels and sending known bits (usually
called frozen bits) through the completely noisy channels.
© Po-Ning Chen@ece.nctu IDC9-22
Example of Polar Codes
for the Binary Erasure Channel (BEC)
o Specific property of BEC
n When Y = 0 or 1, X is exactly equal to Y.
n When Y = E is erased, nothing is known about X.

© Po-Ning Chen@ece.nctu IDC9-23

Example of Polar Codes
for the Binary Erasure Channel (BEC)

o We start with the simplest case (often named basic

transformation) of n=2.
o Under uniformly distributed X1 and X2, we have

© Po-Ning Chen@ece.nctu IDC9-24

Example of Polar Codes
for the Binary Erasure Channel (BEC)

o Now consider the following linear modulo-2 operation:

where U1 and U2 represent uniformly distributed independent

message bits.
© Po-Ning Chen@ece.nctu IDC9-25
Example of Polar Codes
for the Binary Erasure Channel (BEC)

o The decoder performs successive cancellation decoding.

n It first decodes U1 from the received (Y1, Y2).
n Then decode U2 based on (Y1, Y2) and the previously
decoded U1 (assuming the decoding is done correctly).

© Po-Ning Chen@ece.nctu IDC9-26

Example of Polar Codes
for the Binary Erasure Channel (BEC)

o This will create respectively two new channels; namely the

worse channel and the better channel given by:

o The names of these channels will be justified shortly.

© Po-Ning Chen@ece.nctu IDC9-27
o Given output E for a BEC, the receiver knows nothing
about the input.
o Thus, is a BEC with erasure probability

© Po-Ning Chen@ece.nctu IDC9-28

o is a BEC with erasure probability .

Thus, let U1 be the frozen bit and U2 be the info bit.

One can transform the system to a BEC(! " )
with code rate 1/2 bits/channel usage.

© Po-Ning Chen@ece.nctu IDC9-29

Overall Capacity Remains the Same

© Po-Ning Chen@ece.nctu IDC9-30

© Po-Ning Chen@ece.nctu IDC9-31
Polar Coding System
o The process of using multiple basic transformations to get
X1, …, Xn from U1, …, Un , where Ui's are i.i.d. uniform
message random variables is called channel combining.

o The process of using Y1, …, Yn and U1, …, Ui-1 to obtain Ui

for i in {1, …, n} is called channel splitting.

o Altogether is called channel polarization.

© Po-Ning Chen@ece.nctu IDC9-32

Consider n=8 BECs
with erasure
probability 0.5

© Po-Ning Chen@ece.nctu IDC9-33

Final Notes
o A key reason for the prevalence of polar coding after its
invention is that it has an explicit low-complexity
construction structure while being capable of achieving
channel capacity as code length approaches infinity.
o More importantly, polar codes do not exhibit the error floor
behavior, which Turbo and (to a lesser extent) LDPC codes
are prone to.
o Due to their attractive properties, polar codes were adopted
in 2016 by the 3rd Generation Partnership Project (3GPP) as
error correcting codes for the control channel of the 5G
mobile communication standard.
© Po-Ning Chen@ece.nctu IDC9-34