© Robert W. Heath Jr. (2015)
Millimeter wave as the future of 5G
Robert W. Heath Jr., PhD, PE
Cullen Trust Endowed Professor
Wireless Networking and Communications Group
Department of Electrical and Computer Engineering
The University of Texas at Austin
www.profheath.org
© Robert W. Heath Jr. (2015)
Heath Group in the WNCG @ UT Austin
12 PhD
students
mmWave communication
and radar for car-to-car!
mmWave precoding!
mmWave wearables!
mmWave for tactical ad!
hoc networks!
mmWave for infrastructure-to-car!
next generation!
mmWave LAN!
mmWave licensed
shared access for 5G!
mmWave 5G performance!
2
© Robert W. Heath Jr. (2015)
Why millimeter wave for 5G?
cellular
300 MHz
30 GHz
u
u
u
UHF (ultra high frequency) spectrum!
note: log scale so even smaller over here
WiFi
3 GHz
300 GHz
Huge amount of spectrum possibly available in mmWave bands
Technology advances make mmWave possible for cheap consumer devices
mmWave research is as old as wireless, e.g. Bose 1895 and Lebedew 1895
3
© Robert W. Heath Jr. (2015)
MmWave is coming for consumers
Standard
Bandwidth
Rates
Approval Date
WirelessHD
2.16 GHz
3.807 Gbps
Jan. 2008
IEEE 802.11ad
2.16 GHz
6.76 Gbps
Dec. 2012
u
Standards developed @ unlicensed 60 GHz band
ª  WirelessHD: Targeting HD video streaming
ª  IEEE 802.11ad: Targeting Gbps WLAN
u
Epson projector**
Compliant products already available
ª  Dell Alienware laptops, Epson projectors, etc.
ª  11ad Chipset available from Wilocity, Tensorcom, Nitero
u
Dell Laptop**
Wilocity’s chipset***
Only single stream MIMO beamforming
ª  Next generation will likely support multi-stream (>20 Gbps)*
Tensorcom’s chipset***
* http://www.ieee802.org/11/Reports/ng60_update.htm ** http://www.wirelesshd.org/consumers/product-listing/ *** http://www.dailytech.com/
4
© Robert W. Heath Jr. (2015)
Spectrum considerations
u
u
u
There is no specific allocation for 5G cellular at millimeter wave yet
Some candidate bands and their bandwidth (many shared with fixed and
mobile satellite, and federal / non-federal users)
28 GHz (LMDS)
1.3 GHz
39 GHz
1.4 GHz
37 / 42 GHz
2.1 GHz
71-­‐76 GHz 81-­‐86 GHz (E-­‐Band)
10 GHz
FCC released a notice of inquiry to start the conversation about mmWave
ª  NOI posses many questions that are being addressed by research at UT
u
Not obvious that exclusive licensing will happen in mmWave
ª  Shared licensed access may be attractive due to reduced co-channel interference
ª  Cognitive radio techniques may allow co-existence with satellite or radar
See UT’s response to comments here http://apps.fcc.gov/ecfs/comment/view?id=60001017585
5
© Robert W. Heath Jr. (2015)
How does mmWave enable ultrafast broadband?
#1 spectrum
more spectrum (10x or more)
larger channels (5-100x)
#2 large arrays & narrow beams
reduced interference
(better SINR)
spectrum reuse (multiple
users share same channel)
6
© Robert W. Heath Jr. (2015)
Role of MIMO for mmWave
millimeter wave band
1.3 GHz
2.1 GHz
28 GHz
37 / 42 GHz
7 GHz
10 GHz
(unlic)
60GHz E-Band
spatial multiplexing & beamforming
mmWave
aperture
isotropic
radiator
TX
RX
sub-6GHz
aperture
Beamforming for antenna gain
several GHz of spectrum
is promising but found in
many separate bands
… to 300 GHz
just beamforming
multiple data streams
Spatial multiplexing for spectral efficiency
Shu Sun, T. Rappapport, R. W. Heath, Jr., A. Nix, and S. Rangan, `` MIMO for Millimeter Wave Wireless Communications: Beamforming, Spatial Multiplexing,
or Both?,'' IEEE Communications Magazine, December 2014.
77
© Robert W. Heath Jr. (2015)
Observations about antenna arrays
64 to 256
elements
Base station
u
4 to 32
elements
User equipment
Large number of antennas used at the base station and mobile station
ª  Antennas will be small -> no form factor challenges at the base station
u
Directionality of the patterns changes many aspects of system design
ª  Physical layer signal processing
ª  Mobility management (e.g. initial access and handoff)
ª  Interference management
8
© Robert W. Heath Jr. (2015)
Blockages will become more severe
X
line-of-sight
non-line-of-sight
blockage due to people
blockage due to buildings
User
Base station
Handset
X
Blocked by users’ body
hand blocking
self-body blocking
many forms of blockage have yet to be modeled and analyzed
9
© Robert W. Heath Jr. (2015)
Observations about blockage
u
Building blockage
ª  High density of infrastructure required to
cover areas around buildings
u
Use
Reflected
r
path
LOS path
BS
Body blockage and self-body blockage
ª  Need rapid switching between line-ofsight and non-line-of-sight paths
ª  Macro diversity where users associate
with multiple base stations
u
Buildings
Hand blockage
ª  Array diversity on the handset
Blocked BS
Serving BS
X
Coordinating BS
10
© Robert W. Heath Jr. (2015)
Analytical model for mmWave cellular systems
Random building model for LOS/NLOS links !
exponent proportional to building density
LOS:
K=0
Buildings
Associated
Transmitter
LOS path
NLOS Path
Typical Receiver
non-LOS
K>0
Random building model
Exponentially decaying LOS prob.
Simplified model for directional beamforming!
Back lobe gain
Main lobe array gain
Interfering Transmitters
Main lobe beamwidth
T. Bai, R. Vaze, and R. W. Heath, Jr., ``Analysis of Blockage Effects in Urban Cellular Networks”, IEEE Transactions on Wireless Communications, 2014.
T. Bai and R. W. Heath Jr., “Coverage and rate analysis for millimeter wave cellular networks”, IEEE Transactions on Wireless Communications, 2015
Tianyang Bai, Ahmed Alkhateeb, and R. W. Heath, Jr., `` Coverage and Capacity of Millimeter Wave Cellular Networks,'' IEEE Communications Magazine, September 2014.
11
© Robert W. Heath Jr. (2015)
Performance calculations
5% rate
（Mbps）
avg rate
(Mbps)
UHF with 1TX 1RX
1.26
67.53
UHF with 4TX 4RX
13.22
148.95
mmWave with low density and building
blockages
2.95
2579.62
mmWave with high density and building
blockages
2427.2
3716.41
mmWave with high density and
building / body blockages
2106.53
3682.32
scenario
UHF & mmWave with high density and
building blockages
2434.1
3733.3
Maximum rate from two bands
UHF (2 GHz)parameters:
Carrier frequency: 2 GHz
BW: 50 MHz
ISD: 500 m
TX power: 46 dBm
MIMO with ZF receiver
MmWave parameters:
Carrier frequency: 28 GHz
BW: 500 MHz
ISD: 100 m (Dense)
200m (Sparse)
TX power: 30 dBm
BS beamwidth: 10 degree
BS beamforming gain: 20 dB
MS beamwidth: 90 degree
MS beamforming gain: 6 dB
Body blocking loss: 30 dB
Body blocking prob.: 1/6
Building statistics:
LOS range: 200 m
(Austin downtown)
Rate computation:
5 dB gap from Shannon
SINR clipped by 30 dB
all outdoor users!
12
© Robert W. Heath Jr. (2015)
Conclusion
Wireless
backhaul
mmWave BS
Microwave Macro BS
Indoor user
Control signals
Femtocell
Buildings
LOS links
Data center
Multiple-BS access for
fewer handovers and
high rate
mmWave D2D
Non-line-of-sight
(NLOS) link
mmWave will impact every aspect of cellular communication
https://www.youtube.com/watch?v=BQ45FuGpFQ0
[1] Tianyang Bai, Ahmed Alkhateeb, and R. W. Heath, Jr., `` Coverage and Capacity of Millimeter Wave Cellular Networks,''
IEEE Communications Magazine, September 2014.
[2] Ahmed Alkhateeb, Jianhua Mo, N. González Prelcic and R. W. Heath, Jr., `` MIMO Precoding and Combining Solutions
for Millimeter Wave Systems,'' IEEE Communications Magazine, December 2014.
[3] Shu Sun, T. Rappapport, R. W. Heath, Jr., A. Nix, and S. Rangan, `` MIMO for Millimeter Wave Wireless Communications:
Beamforming, Spatial Multiplexing, or Both?,'' IEEE Communications Magazine, December 2014.
[4] T. S. Rappaport, R.W. Heath, Jr. , J. N. Murdock, R. C. Daniels, Millimeter Wave Wireless Communications, Pearson, 2014
13