EE 359: Wireless Communications About me

Professor Andrea Goldsmith

 Engineering prof dad, cartoonist mom

 Undergrad at UC Berkeley 1982-1986
 Worked in Silicon Valley 1986-1989
 Fell in love with wireless.
 Ph.D. from UCB: 1989-1994
 Summers at AT&T Bell Labs
 Taught at Caltech 1994-1999
Next-Gen Cellular/WiFi  Moved to Stanford in 1999
Smart Homes/Spaces
 Lots of stuff in addition to research, teaching
Autonomous Cars
Smart Cities  Founded 2 wireless companies: Quantenna (QTNA’06) and Plume WiFi’10
Body-Area Networks Much work around diversity in STEM (in academia, industry, and IEEE)
Internet of Things  Best Results: Daniel (22) and Nicole (20)
All this and more …

1 2

Course Information*
Outline People
 Instructor: Andrea Goldsmith, Pack 371, andrea@ee,
 Course Basics OHs: TTh immediately after class and by appt.
 Course Syllabus  TA: Tom Dean ([email protected])
 Wireless History  Discussion section: Wed 4-5 pm (hopefully taped)
 OHs: Wed 5-6pm, Th 4-5pm, Fri 11-12pm (tentative). Emails
 The Wireless Vision received during OHs will be responded to during or just
after. Email questions are ideally via Piazza.
 Technical Challenges  Piazza: https://piazza.com/stanford/win2020/ee359/home.
All are registered, will use to poll on OH/discussion times
 Current/Next-Gen Wireless Systems
 Class Administrator: Dash Corbett, email:
 Spectrum Regulation and Standards [email protected], 365 Packard, 723-2681.
 Emerging Wireless Systems (Optional Lecture) Homework dropoff: Fri by 4 pm.
*See web or handout for more details
3 4

1
 Course Information Course Information
Nuts and Bolts Policies
 Prerequisites: EE279 or equivalent (Digital Communications)  Grading: Two Options
 No Project (3 units): HW – 25%, 2 Exams – 35%, 40%
 Textbook: Wireless Communications (by me), draft 2nd Ed.
 Project (4 units): HWs- 20%, Exams - 25%, 30%, Project - 25%
 Available as reader at bookstore or on website
 Raffle for $100 Amazon gift card for typos/mistakes/suggestions  HWs: assigned Thu, due following Fri 4pm (starts next week)
 Supplemental texts at Engineering Library.
 Homeworks lose 33% credit after 4pm Fri, lowest HW dropped
 Class Homepage: www.stanford.edu/class/ee359  Up to 3 students can collaborate and turn in one HW writeup
 All announcements, handouts, homeworks, etc. posted to website  Collaboration means all collaborators work out all problems together
 “Lectures” link continuously updates topics, handouts, and reading  Unpermitted collaboration or aid (e.g. solns for the book or from prior
 Calendar will show any changes to class/OH/discussion times years) is an honor code violation and will be dealt with strictly.
 Extra credit: up to 2 “design your own” HW problems; course eval
 Class Mailing List: ee359-win1920-students@lists (automatic
for on-campus registered students).  Exams:
 Guest list ee359-win1920-guest@lists for SCPD and auditors: send Tom  Midterm week of 2/17 (It will be scheduled outside class time; the
email to sign up. duration is 2 hours.) Final on 3/17 from 3:30-6:30pm (pizza after)
 Sending mail to ee359-win1920-staff@lists reaches me and Tom.  Exams must be taken at scheduled time (with very few exceptions)

5 6

Course Information
Projects Course Syllabus
 The term project (for students electing to do a project) is a
research project related to any topic in wireless
 Overview of Wireless Communications
 Path Loss, Shadowing, and Fading Models
 Two people may collaborate if you convince me the sum of
the parts is greater than each individually  Capacity of Wireless Channels
 A 1 page proposal is due 2/7 at midnight.
 Digital Modulation and its Performance
 5-10 hours of work typical for proposal  Adaptive Modulation
 Must create project website and post proposal there (submit web link)
 Preliminary proposals can be submitted for early feedback  Diversity
 The project is due by midnight on 3/14 (on website)  MIMO Systems
 20-40 hours of work after proposal is typical for a project  Multicarrier Systems: OFDM and other waveforms
 Suggested topics in project handout  Multiuser and Cellular Systems
 Anything related to wireless or application of wireless techniques ok.

7 8

2
 Tentative Detailed Syllabus
Lecture # Date Topic Required Reading Class Rescheduling
Introduction
1 1/7 Overview of Wireless Communications Chapter 1

2-3 1/9, 1/14

Wireless Channel Models
Path Loss and Shadowing Models, Millimeter wave
Chapter 2
 No lectures Thu 1/16, Tue 2/4, Thu 2/20 and Thu 3/5.
propagation

4-5 1/17, 1/21 Statistical Fading Models, Narrowband Fading Section 3.1-3.2.3

6 1/23 Wideband Fading Models Section 3.3

 These lectures are tentatively rescheduled as:
Impact of Fading and ISI on Wireless Performance
 Lecture on Thu 1/16 is rescheduled to Fri 1/17 at lunch
7 1/28 Capacity of Wireless Channels Chapter 4

8,9,10 1/30, 2/5, 2/6 Digital Modulation and its Performance

Lec 8: Chapter 5  Lecture on Tue 2/4 is rescheduled to Wed 2/5 at lunch
Lec 9-10: Chapter 6
Flat-Fading Countermeasures  Lecture on Thu 2/20 is rescheduled to Fri 2/21 at lunch
11 2/11 Diversity Chapter 7
 Lecture on Thu 3/5 is rescheduled to Wed 3/4 at lunch
MT Week of 2/17 Midterm (outside class time) Chapters 2 to 7

12-13 2/13-2/18 Adaptive Modulation Chapter 9.1-9.3  Last lecture on 3/12 has an optional component 11:50-12:30
14-15 2/21-2/25 Multiple Input/Output Systems (MIMO)

ISI Countermeasures
Chapter 10, Appendix C
on advanced topics with lunch.
Multicarrier Systems, OFDM, and other multicarrier
16-17 2/27, 3/3 Chapter 12
waveforms
18-19 3/4-3/10 Multiuser and Cellular Systems Topics in Chapters 13-15
Course Summary

Course summary/final review (and optional advanced

20 3/12
topics lecture over lunch)

Final 3/17 3:30-6:30pm Pizza party to follow

9 10

Future Wireless Networks

Wireless History Ubiquitous Communication Among People and Devices

 Ancient Systems: Smoke Signals, Carrier Pigeons, …

 Radio invented in the 1880s by Marconi
 Many sophisticated military radio systems were
developed during and after WW2
 Exponential growth in cellular use since 1988:
approx. 8B worldwide users today
 Ignited the wireless revolution Next-Gen Cellular/WiFi
Smart Homes/Spaces
 Voice, data, and multimedia ubiquitous Autonomous Cars
 Use in 3rd world countries growing rapidly Smart Cities
 Wifi also enjoying tremendous success and growth Body-Area Networks
Internet of Things
 Bluetooth pervasive, satellites also widespread All this and more …

11 12

3
 Challenges Software-Defined (SD) Radio:
 Network/Radio Challenges 5-6G AdHoc Is this the solution to the device challenges?
 Gbps data rates with “no” errors Short-Range
 Energy efficiency
BT
FM/XM A/D
 Scarce/bifurcated spectrum Cellular GPS
A/D
 Reliability and coverage DVB-H

 Heterogeneous networks Apps DSP

Processor WLAN A/D
 Seamless internetwork handoff
Media
Wimax
 Device/SoC Challenges Processor A/D
BT
Radio
 Performance
 Complexity Cellular
GPS
 Wideband antennas and A/Ds span BW of desired signals
 Size, Power, Cost, Energy Cog
 DSP programmed to process desired signal: no specialized HW
 High frequencies/mmWave Mem WiFi
 Multiple Antennas Today, this is not cost, size, or power efficient
 Multiradio Integration CPU mmW
 Coexistance SubNyquist sampling may help with the A/D and DSP requirements

13 14

“Sorry America, your On the Horizon,

airwaves are full*” the Internet of Things

Source: FCC

15 16

4
 Are we at the Shannon
What is the Internet of Things: limit of the Physical Layer?
We are at the Shannon Limit
 Enabling every electronic device to be  “The wireless industry has reached the theoretical limit of
connected to each other and the Internet how fast networks can go” K. Fitcher, Connected Planet

 “We’re 99% of the way” to the “barrier known as Shannon’s

 Includes smartphones, consumer electronics, limit,” D. Warren, GSM Association Sr. Dir. of Tech.
cars, lights, clothes, sensors, medical devices,… Shannon was wrong, there is no limit
 “There is no theoretical maximum to the amount of data
that can be carried by a radio channel” M. Gass, 802.11
 Value in IoT is data processing in the cloud Wireless Networks: The Definitive Guide
 “Effectively unlimited” capacity possible via personal cells
(pcells). S. Perlman, Artemis.
Different requirements than smartphones: low rates/energy consumption

17 18

Current/Next-Gen
What would Shannon say? Wireless Systems
We don’t know the Shannon  Current:
capacity of most wireless channels 

4G Cellular Systems (LTE-Advanced)
6G Wireless LANs/WiFi (802.11ax)
 Time-varying channels.  mmWave massive MIMO systems
 Satellite Systems
 Channels with interference or relays.  Bluetooth
 Cellular systems  Zigbee
 WiGig
 Ad-hoc and sensor networks
 Emerging
 Channels with delay/energy/$$$ constraints.
 5G Cellular and 7G WiFi Systems
 Ad/hoc and Cognitive Radio Networks Much room
Shannon theory provides design insights  Energy-Harvesting Systems For innovation
and system performance upper bounds  Chemical/Molecular

19 20

5
 Cellular Systems:
Spectral Reuse Reuse channels to maximize capacity
Due to its scarcity, spectrum is reused  Geographic region divided into cells
 Freq./timeslots/codes/space reused in different cells (reuse 1 common).
In licensed bands and unlicensed bands  Interference between cells using same channel: interference mitigation key
 Base stations/MTSOs coordinate handoff and control functions
 Shrinking cell size increases capacity, as well as complexity, handoff, …

BS

BASE
STATION
MTSO

Cellular Wifi, BT, UWB,…

Reuse introduces interference

21 22

4G/LTE Cellular 5G Upgrades from 4G

 Much higher data rates than 3G (50-100 Mbps)
 3G systems has 384 Kbps peak rates
 Greater spectral efficiency (bits/s/Hz)
 More bandwidth, adaptive OFDM-MIMO,
reduced interference
 Flexible use of up to 100 MHz of spectrum
 10-20 MHz spectrum allocation common
 Low packet latency (<5ms).
 Reduced cost-per-bit (not clear to customers)
 All IP network

23 24

6
 Wifi Networks
Future Cellular Phones Multimedia Everywhere, Without Wires
Burden for this
Everything performance
wireless is on the backbone network
in one device
San Francisco

BS BS
802.11ac

LTE backbone is the Internet

Internet
Paris
Nth-Gen Phone Nth-Gen
Cellular System Cellular

BS
• Streaming video
• Gbps data rates
Much better performance and reliability than today • High reliability Wireless HDTV
- Gbps rates, low latency, 99% coverage, energy efficiency • Coverage inside and out and Gaming

25 26

Wireless LAN Standards Why does WiFi performance suck?

 802.11b (Old – 1990s) Carrier Sense Multiple Access:
 Standard for 2.4GHz ISM band (80 MHz) if another WiFi signal
 Direct sequence spread spectrum (DSSS) detected, random backoff
 Speeds of 11 Mbps, approx. 500 ft range Many
WLAN Collision Detection: if collision
 802.11a/g (Middle Age– mid-late 1990s) cards detected, resend
 Standard for 5GHz band (300 MHz)/also 2.4GHz have many
 OFDM in 20 MHz with adaptive rate/codes generations
 Speeds of 54 Mbps, approx. 100-200 ft range
 The WiFi standard lacks good mechanisms to mitigate
 802.11n/ac/ax or Wi-Fi 6 (current gen) interference, especially in dense AP deployments
 Standard in 2.4 GHz and 5 GHz band  Multiple access protocol (CSMA/CD) from 1970s
 Adaptive OFDM /MIMO in 20/40/80/160 MHz  Static channel assignment, power levels, and sensing thresholds
 Antennas: 2-4, up to 8  In such deployments WiFi systems exhibit poor spectrum reuse
 Speeds up to 1 Gbps (10 Gbps for ax), approx. 200 ft range and significant contention among APs and clients
 Other advances in packetization, antenna use, multiuser MIMO  Result is low throughput and a poor user experience
 Multiuser MIMO will help each AP, but not interfering APs

27 28

7
 Self-Organizing Networks for WiFi Satellite Systems
- Channel Selection
SoN  Cover very large areas
Controller - Power Control
- etc.  Different orbit heights
 Orbit height trades off coverage area for latency
 GEO (39000 Km) vs MEO (9000 km) vs LEO (2000 Km)
 SoN-for-WiFi: dynamic self-organization network
 Optimized for one-way transmission
software to manage of WiFi APs.
 Radio (XM, Sirius) and movie (SatTV, DVB/S) broadcasts
 Allows for capacity/coverage/interference mitigation  Most two-way LEO systems went bankrupt in 1990s-2000s
tradeoffs.  LEOs have resurfaced with 4G to bridge digital divide
 Global Positioning System (GPS) ubiquitous
 Also provides network analytics and planning.
 Satellite signals used to pinpoint location
 Popular in cell phones, PDAs, and navigation devices
29 30

Bluetooth IEEE 802.15.4/ZigBee Radios

 Cable replacement RF technology (low cost)  Low-rate low-power low-cost secure radio
 Complementary to WiFi and Bluetooth
 Short range (10m, extendable to 100m)  Frequency bands: 784, 868, 915 MHz, 2.4 GHz
 2.4 GHz band (crowded)  Data rates: 20Kbps, 40Kbps, 250 Kbps
 1 Data (700 Kbps) and 3 voice channels, up  Range: 10-100m line-of-sight
to 3 Mbps  Support for large mesh networking or star clusters
 Support for low latency devices
 Widely supported by telecommunications,
 CSMA-CA channel access
PC, and consumer electronics companies
 Applications: light switches, electricity meters,
 Few applications beyond cable replacement traffic management, and other low-power sensors.
8C32810.61-Cimini-7/98

31 32

8
 Spectrum Regulation Standards
 Spectrum a scarce public resource, hence allocated  Interacting systems require standardization
 Spectral allocation in US controlled by FCC  Companies want their systems adopted as standard
(commercial) or OSM (defense)  Alternatively try for de-facto standards

 FCC auctions spectral blocks for set applications.  Standards determined by TIA/CTIA in US
 Some spectrum set aside for universal use  IEEE standards often adopted
 Process fraught with inefficiencies and conflicts
 Worldwide spectrum controlled by ITU-R
 Worldwide standards determined by ITU-T
 Regulation is a necessary evil.
 In Europe, ETSI is equivalent of IEEE
Innovations in regulation being considered worldwide
in multiple cognitive radio paradigms Standards for current systems summarized in text Appendix D.

33 34

Advanced Topics Lecture: See Backup Slides

Emerging Systems Main Points

 New cellular system architectures  The wireless vision encompasses many exciting applications
 mmWave/massive MIMO communications  Technical challenges transcend all system design layers
 Software-defined network architectures
 5G networks must support higher performance for some
 Ad hoc/mesh wireless networks users, extreme energy efficiency and/or low latency for others
 Cognitive radio networks
 Cloud-based software to dynamically control and optimize
 Wireless sensor networks wireless networks needed (SDWN)
 Energy-constrained radios
 Distributed control networks  Innovative wireless design needed for 5G cellular/WiFi,
mmWave systems, massive MIMO, and IoT connectivity
 Chemical Communications
 Applications of Communications in Health, Bio-  Standards and spectral allocation heavily impact the evolution
of wireless technology
medicine, and Neuroscience
35 36

9
 Rethinking “Cells” in Cellular
How should cellular
Coop Small
MIMO Cell systems be designed for
Relay
- Capacity
Backup Slides: - Coverage
- Energy efficiency
Emerging Systems
DAS

- Low latency
 Traditional cellular design “interference-limited”
 MIMO/multiuser detection can remove interference
 Cooperating BSs form a MIMO array: what is a cell?
 Relays change cell shape and boundaries
 Distributed antennas move BS towards cell boundary
 Small cells create a cell within a cell
 Mobile cooperation via relays, virtual MIMO, network coding.
37 38

mmWave Massive MIMO Software-Defined Network Architectures

Video Security Cloud Computing
Vehicular
Networks
M2M App layer Health
Dozens of devices
10s of GHz of Spectrum

Freq. Power Self QoS CS

ICIC
Hundreds
Allocation Control Healing Opt. Threshold
of antennas
Network Optimization

 mmWaves have large non-monotonic path loss UNIFIED CONTROL PLANE

 Channel model poorly understood
 For asymptotically large arrays with channel state information, no
attenuation, fading, interference or noise HW layer
Distributed Antennas
 mmWave antennas are small: perfect for massive MIMO
 Bottlenecks: channel estimation and system complexity WiFi Cellular mmWave
… Ad-Hoc
Networks
 Non-coherent design holds significant promise

39 40

10
 Ad-Hoc Networks Cognitive Radios
CRTx CRRx
IP
NCR
NCR CR CR NCRRx
NCRTx

MIMO Cognitive Underlay Cognitive Overlay

 Peer-to-peer communications  Cognitive radios support new users in existing
crowded spectrum without degrading licensed users
 No backbone infrastructure or centralized control
 Utilize advanced communication and DSP techniques
 Routing can be multihop.  Coupled with novel spectrum allocation policies
 Topology is dynamic.
 Fully connected with different link SINRs  Multiple paradigms
 Open questions  (MIMO) Underlay (interference below a threshold)
 Fundamental capacity region  Interweave finds/uses unused time/freq/space slots
 Resource allocation (power, rate, spectrum, etc.)  Overlay (overhears/relays primary message while
 Routing cancelling interference it causes to cognitive receiver)

41 42

Wireless Sensor Networks

Data Collection and Distributed Control Energy-Constrained Radios
• Smart homes/buildings  Transmit energy minimized by sending bits slowly
• Smart structures
• Search and rescue  Leads to increased circuit energy consumption
• Homeland security
• Event detection
• Battlefield surveillance  Short-range networks must consider both transmit
and processing/circuit energy.
 Sophisticated encoding/decoding not always energy-
efficient.
 MIMO techniques not necessarily energy-efficient
 Long transmission times not necessarily optimal
 Multihop routing not necessarily optimal
 Energy (transmit and processing) is the driving constraint
 Sub-Nyquist sampling can decrease energy and is
 Data flows to centralized location (joint compression) sometimes optimal!
 Low per-node rates but tens to thousands of nodes
 Intelligence is in the network rather than in the devices

43 44

11
 Where should energy come from? Distributed Control over Wireless
Automated Vehicles
• Batteries and traditional charging mechanisms - Cars
• Well-understood devices and systems - Airplanes/UAVs
- Insect flyers
• Wireless-power transfer
• Poorly understood, especially at large distances and
with high efficiency

• Communication with Energy Harvesting Radios

Interdisciplinary design approach
• Intermittent and random energy arrivals
• Control requires fast, accurate, and reliable feedback.
• Communication becomes energy-dependent
• Wireless networks introduce delay and loss
• Can combine information and energy transmission • Need reliable networks and robust controllers
• New principles for radio and network design needed. • Mostly open problems : Many design challenges

45 46

Applications in Health,
Chemical Communications Biomedicine and Neuroscience
Neuroscience
-Nerve network
Body-Area (re)configuration
Networks -EEG/ECoG signal
processing
- Signal processing/control
for deep brain stimulation
- SP/Comm applied to
bioscience
Recovery from Nerve Damage
 Can be developed for both macro (>cm) and ECoG Epileptic Seizure Localization
micro (<mm) scale communications EEG

 Greenfield area of research:

 Need new modulation schemes, channel ECoG
impairment mitigation, multiple acces, etc.

47 48

12