IoT Protocols

Qian Zhang
Agenda
01
Fog Computing Architecture for IoT

02
Protocols of IoT (ZigBee, IEEE 802.11ah, …)

03
Long range wide area network for IoT

04
Energy-efficient WiFi for IoT
Fog Computing: A Platform
for IoT and Analytics
Cloud Computing
only cloud is not the optimal solution to handle this massive explosion
Fog Computing
• Fog computing is making use of
decentralized servers in
between network core and
network edge for data
processing and to serve the
immediate requirements of the
end systems.
• Fog computing is non-trivial
extension of Cloud computing
paradigm to the edge of the
network.
 Need for fog computing

• Why can’t do all in cloud?

– Cloud computing frees the enterprise and the end user
from many details.
– This bliss becomes a problem for latency-sensitive
applications.

• Why can’t do all in end systems?

– Physical constraints: energy, space, etc.,
Illustrative Use Cases to Drive Fog computing

• Use Case 1: A smart Traffic Light System (STLS)

• Use Case 2: Wind Farms

To abstract the major requirements to propose an

architecture that addresses a vast majority of the IoT
requirements.
Use Case 1: A Smart Traffic Light System(STLS)
System Outline:
• STLS calls for deployment of a STL at each intersection.
• The STL is equipped with sensors that
1. Measure the distance and speed of approaching vehicles from every
direction.
2. Detect presence of pedestrians/other vehicles crossing the street.
- Issues “Slow down” warnings to vehicles at risk to crossing in red and
even modifies its own cycle to prevent collisions.
 STLS: System outline continued..

• STLS has 3 major goals:

1. Accidents prevention
2. Maintenance of steady flow of traffic (green waves along the main
roads)
3. Collection of relevant data to evaluate and improve the system

Note:
Goal (1) requires real-time reaction, (2) near-real time, and (3) relates
to the collection and analysis of global data over long periods.
 Key requirements driven by STLS
1. Local Subsystem latency:- Reaction time needed is in the order of <
10 milliseconds.
2. Middleware orchestration platform:- Middleware to handle a # of
critical software components. A. Decision maker(DM), B. message
bus.
3. Networking infrastructure:- Fog nodes belongs to a family of
modular compute and storage devices.
4. Interplay with the cloud:- Data must be injected into a Data center/
cloud for deep analysis to identify patterns in traffic, city pollutants.
 STLS Key requirements, cont’d.
5. Consistency of a highly distributed system:- Need to be
Consistent between the different aggregator points.
6. Multi-tenancy:- It must provide strict service guarantees all the
time.
7. Multiplicity of providers:- May extend beyond the borders of a
single controlling authority. Orchestration of consistent policies
involving multiple agencies is a challenge unique to Fog
Computing.
 Use case 2: Wind Farm

Brings up requirements shared by a number of Internet of

Everything (IoE) deployments:
1. Interplay between real time analytics and batch analytics.
2. Tight interaction between sensors and actuators, in closed
control loops.
3. Wide geographical deployment of a large system consistent
of a number of autonomous yet coordinated modules –
which gives rise to the need of an orchestration platform.
 System outline:
There are 4 typical regions:

1. Region1: Wind speed is very low(say, 6m/sec), not so economical to

run the turbine.
2. Region2: Normal operating condition(winds between 6-12m/sec), so
maximum conversion of wind power into electrical power.
3. Region3: Winds exceed 12 m/sec, power is limited to avoid exceeding
safe electrical and mechanical loads.
4. Region4: Very high wind speeds above 25 m/sec, here turbine is
powered down to avoid excessive operating loads.
Key requirements driven by Wind Farm
1. Network Infrastructure: An efficient communication network between
sub-systems, system and the internet (cloud)
2. Global controller: gathering data, building the global state, determining
the policy.
3. Middle Orchestration platform: A middleware that mediates between
sub-systems and the cloud.
4. Data analytics: (1) requires real-time reaction, (2) near-real time, and (3)
relates to the collection and analysis of global data over long periods.
 Key attributes of Fog computing

The Use Cases that were discussed brings up a # of attributes

that differentiate Fog computing platform from the Cloud.
– Applications that require very low and predictable latency. (STLS,
SCV)
– Geo-distributed applications (pipeline monitoring, STLS)
– Fast mobile applications (Smart connected vehicle, rail)
– Large-scale distributed control systems (STLS, smart grid)
– IoT also brings Big Data with a twist: rather than high volume, the
number of data sources distributed geographically
Geo-distribution: A new Dimension of Big Data
• 3 Dimensions: Volume, Velocity and Variety.
• IoT use cases: STLS, Connected Rail, pipeline monitoring are naturally
distributed.
• This suggests to add a 4th dimension: geo-distribution.
• Since challenge is to manage number of sensors (and actuators) that
are naturally distributed as a coherent whole.
• Call for “moving the processing to the data”
• A distributed intelligent platform at the Edge (Fog computing) that
manages distributed compute, networking, and storage resources.
The Edge (Fog) and the core (Fog) interplay:
Many uses of same data
Fog Software Architecture
• Fog nodes are
heterogeneous in nature
and deployed in variety
of environments
including core, edge,
access networks and
endpoints

• Fog architecture should

facilitate seamless
resource management
across diverse set of
platforms
 Conclusion

• We looked at Fog computing and key aspects of it

• How fog complements and extends cloud computing
• We looked at use cases that motivated the need for fog
• Seen a high-level description of Fog’s architecture
Agenda
01
Fog Computing Architecture for IoT

02
Protocols of IoT (ZigBee, IEEE 802.11ah, …)

03
Long range wide area network for IoT

04
Energy-efficient WiFi for IoT
IoT Ecosystem
Protocols for IoT
1. Bluetooth
• Started with Ericsson's Bluetooth Project in 1994 for radio-communication between
cell phones over short distances
• Named after Danish king Herald Blatand (AD 940-981) who was fond of blueberries
• Intel, IBM, Nokia, Toshiba, and Ericsson formed Bluetooth SIG in May 1998
• Version 1.0A of the specification came out in late 1999
• IEEE 802.15.1 approved in early 2002 is based on Bluetooth. Later versions handled by
Bluetooth SIG directly
• Key Features:
• Lower Power: 10 mA in standby, 50 mA while transmitting
• Cheap: $5 per device
• Small: 9 mm2 single chips
History
Bluetooth Versions
• Bluetooth 1.1: IEEE 802.15.1-2002
• Bluetooth 1.2: IEEE 802.15.1-2005. Completed Nov 2003. Extended SCO, Higher
variable rate retransmission for SCO + Adaptive frequency hopping (avoid
frequencies with interference)
• Bluetooth 2.0 + Enhanced Data Rate (EDR) (Nov 2004): 3 Mbps using DPSK. For
video applications. Reduced power due to reduced duty cycle
• Bluetooth 2.1 + EDR (July 2007): Secure Simple Pairing to speed up pairing
• Bluetooth 3.0+ High Speed (HS) (April 2009): 24 Mbps using WiFi PHY + Bluetooth
PHY for lower rates
• Bluetooth 4.0 (June 2010): Low energy. Smaller devices requiring longer battery
life (several years). New incompatible PHY. Bluetooth Smart or BLE
• Bluetooth 4.1: 4.0 + Core Specification Amendments (CSA) 1, 2, 3, 4
• Bluetooth 4.2 (Dec 2014): Larger packets, security/privacy, IPv6 profile
Naming for Bluetooth 4.x
Bluetooth Smart
• Low Energy: 1% to 50% of Bluetooth classic
• For short broadcast: Your body temperature, Heart rate, Wearables, sensors,
automotive, industrial
Not for voice/video, file transfers, …
• Small messages: 1Mbps data rate but throughput not critical
• Battery life: In years from coin cells
• Simple: Star topology. No scatter nets, mesh, …
• Lower cost than Bluetooth classic
• New protocol design based on Nokia’s WiBree technology
Shares the same 2.4GHz radio as Bluetooth
 Dual mode chips
• All new smart phones (iPhone, Android, …) have dual-mode chips
BLE Roles
Topology
BLE Power Status
Bluetooth Smart PHY
• 2.4 GHz. 150 m open field
• Star topology
• 1 Mbps Gaussian Frequency Shift Keying
Better range than Bluetooth classic
• Adaptive Frequency hopping. 40 Channels
with 2 MHz spacing
• 3 channels reserved for advertizing and 37 channels for data
• Advertising channels specially selected to avoid interference
with WiFi channels
Bluetooth Smart MAC
• Two Device Types: “Peripherals” simpler than “central”
• Two PDU Types: Advertising, Data
• Non-Connectable Advertising: Broadcast data in clear
• Discoverable Advertising: Central may request more information. Peripheral
can send data without connection
• General Advertising: Broadcast presence wanting to connect. Central may
request a short connection.
• Directed Advertising: Transmit signed data to a previously connected master
Bluetooth Smart Protocol Stack
Generic Attribute Profile - GATT
GATT Operations
• Central can
• discover UUIDs for all primary services
• Find a service with a given UUID
• Find secondary services for a given primary service
• Discover all characteristics for a given service
• Find characteristics matching a given UUID
• Read all descriptors for a particular characteristic
• Can do read, write, long read, long write values etc.
• Peripheral
• Notify or indicate central of changes
Security
• Encryption (128 bit AES)
• Pairing (Without key, with a shared key, out of band pairing)
• Passive eavesdropping during key exchange (but fixed in
Bluetooth 4.2)
• Many products are building their own security on top of BLE
• Check out Mike Ryan (iSec partners) work on security
Bluetooth Smart Applications
• Proximity: In car, In room 303, In the mall
• Locator: Keys, watches, Animals
• Health devices: Heart rate monitor, physical activities
monitors, thermometer
• Sensors: Temperature, Battery Status, tire pressure
• Remote control: Open/close locks, turn on lights
Use Cases – Physical Security
Use Cases – Home Automation
Use Cases – Geo-fencing/ Positioning
Use Cases - Fun
Development Kits/Boards
Operating System Support
• iOS 8 
• OSX 10.10 
• Android 4.3, 4.4, 5.0 
• Linux 3.4, BlueZ 5.0 
• Windows Phone 8.1 (only central) 
• Windows 8.1 (app mode) 
2. ZigBee Markets
ZigBee Technology-Performance
• Proven excellent in-building coverage
– Inherently robust radio link
– Mesh networking
– Acknowledge oriented protocol
– Now proven in major deployments in Australia, Sweden, & USA
• Proven tolerance to interference
– Trade shows like CES-works when WiFi and Bluetooth fail
– Montage Hotels and MGM City Center deployments
– Products which implement multiple radio technologies
• Proven coexistence
– Many multi-radio products and multi-radio deployments
• Proven scalability
– City Center at 70,000 plus radios
– Montage Hotels at 4000 plus radios per property
ZigBee Platform Interoperability

ZigBee Compliant
Platform

• Ensures Network interoperability but does not imply application layer

interoperability
• There are multiple Compliant Platforms to choose from
ZigBee Product Interoperability

ZigBee
Compliant
Product

• Products with the same application profiles interoperate end to end

• ZigBee has published a set of Public Application Profiles ensuring
end product interoperability
Basic Network Characteristics
• 65,536 network (client) nodes
• 27 channels over 2 bands
• 250Kbps data rate
• Optimized for timing-critical
applications and power
management
• Full Mesh Networking Support Network coordinator
Full Function node
Reduced Function node

Communications flow
Virtual links
Basic Radio Characteristics

ZigBee technology relies upon

IEEE 802.15.4, which has
excellent performance in low
SNR environments
ZigBee Mesh Networking

Slide Courtesy of
ZigBee Mesh Networking

Slide Courtesy of
ZigBee Mesh Networking

Slide Courtesy of
ZigBee Mesh Networking

Slide Courtesy of
ZigBee Mesh Networking

Slide Courtesy of
ZigBee Stack Architecture
Initiate and join network
Manage network
Application Determine device relationships
Send and receive messages

Application ZDO

App Support (APS) Device management

Security functions Device discovery
SSP
Service discovery
NWK
Network organization
Device binding
Route discovery Medium Access (MAC) Messaging
Message relaying
Physical Radio (PHY)
ZigBee Device Types
• ZigBee Coordinator (ZC)
– One required for each ZB network.
– Initiates network formation.

• ZigBee Router (ZR)

– Participates in multihop routing of messages.

• ZigBee End Device (ZED)

– Does not allow association or routing.
– Enables very low cost solutions
ZigBee Network Topologies

Mesh

Star

ZigBee Coordinator
ZigBee Router
Cluster Tree ZigBee End Device
ZigBee Public Profiles
• Home Automation (HA)
• Smart Energy (SE)
• Commercial Building Automation (CBA)
• ZigBee Health Care (ZHC)
• Telecom Applications (TA)

• ZigBee RF4CE Remote Control

• +Future profiles proposed by member companies…

ZigBee Home Automation: for Home Control

Set-top-box
TV/Display Remote access

Closures

Lighting

Heating/cooling
Switches
Security

ZigBee Home Area Network (HAN)

Smart Energy & Home Automation
Urgent demand for Smart Energy + compatibility with mainstream
Home Automation systems enables customer choice
3. WirelessHART
• The HART (Highway Addressable Remote Transducer Protocol)
communication protocol is designed to add diagnostic information to
process devices compatible with legacy 2-20mA analog instrumentation
• The overall performance has been designed to satisfy process automation
needs. It is able to work on distances up to 1500m
• WerelessHART is an extension of HART, its functions include
• Implements an RF self-healing mesh network
• Allows for network-wide time synchronization
• Enhances the publish/subscribe messaging
• Adds network and transport layers
• Adds a fast pipe for time critical traffic and ciphering
Overview
• WirelessHART targets
sensors and actuators,
rotating equipment such as
kiln dryers, environmental
health and safety
applications such as safety
showers, condition
monitoring, and flexible
manufacturing in which a
portion of the plant can be
reconfigured for specific
products.
WirelessHART
• WirelessHART main characteristics
• Low power consumption and low-cost devices
• Data rate of 250 kbps per channel in 2.4GHz ISM band with 15 channels
• Based on IEEE 802.15.4-2006 PHY layer
• Based on a proprietary data link layer with TDMA and CSMA/CA
• Supporting channel hopping and channel blacklisting
• Network layer implementing self-healing mesh network
• Application layer fully compatible with HART
WirelessHART
Comparison between HART, wirelessHART and ZigBee
The Network Architecture
• Each wirelessHART network includes four main elements
• Field devices. They include wirelessHART process transmitters and
wireless adapters
• Gateway. Gateway bridges the wirelessHART network with wired
infrastructures
• Network manager (only one). It is responsible for network configuration,
communication among devices, management of routing messages and
monitor network conditions
• Security manager. Security manager deals with security and encryption,
setting up session keys and their periodic change
• Handhold devices for maintaining purposes are optional
The Network Architecture
Example wirelessHART network
4. Z-Wave
• Z-Wave is a low-power MAC protocol designed for home automation and
has been used for IoT communication, especially for smart home and
small commercial domains
• It covers about 30-meter point-to-point communication and is suitable for
small messages in IoT applications, like light control, energy control,
wearable healthcare control and others
• It uses CSMA/CA for collision detection and ACK messages for reliable
transmission
• It follows a master/slave architecture in which the master control the
slaves, send them commands, and handling scheduling of the whole
network
Z-Wave Vs. Zigbee: What do they have in common?
• Both technologies are mesh networks
• Each node in the system acts as both a wireless data source and a repeater.
Information from a single sensor node hops from node to node until the
transmission reaches the gateway
• Both technologies use the IEEE 802.15.4 low-rate personal area network
(LR-PAN) protocol
• for the unified physical layer (OSI layer 1), structuring packets, and creating
MAC (Medium Access Control) schemes
• Both are widely used in local area sensor data networks
• like in security systems, urban smart grid controllers, HVAC control
systems, home automation, and lighting controls
Z-Wave Vs. Zigbee: How are they different?
• Z-wave has a tightly controlled product ecosystem that caters to the smart home and
smart building space, whereas Zigbee can be used for a number of applications
• There’s no expectation that two Zigbee devices are interoperable unless the interoperability
is preplanned. A Z-Wave application, on the other hand, will almost always integrate with
another Z-Wave device
• Zigbee uses the global standard 2.4GHz ISM frequency band, whereas Z-Wave uses
the 915 MHz ISM band (in the U.S.) and the 868 MHz RFID band (in Europe).
• 2.4 GHz band can be subject to intense interference from WiFi and Bluetooth systems,
whereas the sub-GHz bands Z-Wave uses do not face the same interference issues
• Lots of providers make Zigbee radios, but Z-Wave uses a proprietary radio system from
Sigma designs
• Z-Wave uses frequency-shift keyed modulation (FSK), whereas Zigbee modulation is
carried out through direct sequence spread spectrum (DSSS)
5. IEEE 802.11ah sub 1GHz WLAN for IoT
• Defines operation of license- What lies beneath Wi-Fi HaLow
exempt (ISM) IEEE 802.11 wireless
networks in frequency bands
below 1 GHz
• excluding the TV White
Space bands (802.11af)
• IEEE 802.11 WLAN user
experience for fixed, outdoor,
point to multi point applications
IEEE 802.11ah: scope
• Defines an OFDM PHY operating in the license-exempt bands below
1 GHz
• and enhancements to the IEEE 802.11 MAC to support this PHY, and to provide
mechanisms that enable coexistence with other systems in the bands (e.g. IEEE
802.15.4 P802.15.4g)
• The PHY is meant to optimize the rate vs. range performance of the
specific channelization in a given band
• transmission range up to 1 km
• data rates > 100 kbit/s
• The MAC is designed to support thousands of connected devices
IEEE 802.11ah: use cases
• Use Case 1 : Sensors and meters
• Smart Grid -meter to pole
• Environmental monitoring
• Industrial process sensors
• Healthcare
• Home/Building automation
• Smart city
• Use Case 2 : Backhaul sensor and meter data
• Backhaul aggregation of sensor networks
• Long point-to-point wireless links
• Use Case 3 : Extended range Wi-Fi
• Outdoor extended range hotspot
• Outdoor Wi-Fi for cellular traffic offloading
IEEE 802.11ah: PHY (1)
• Advantages of transmitting in sub 1 GHz:
• Spectrum characteristics
• good propagation and penetration
• large coverage area and one-hop reach
• license-exempt, light licensing
• Reliability:
• less congested frequency band
• high sensitivity and link margin
• available diversity –(frequency, time, space)
• Battery operation
• long battery life
• short data transmissions
IEEE 802.11ah: PHY (2)
• Channelization:
• Configurable bandwidth (channel bonding) of: 1, 2, 4, 8 and 16MHz

• Inherited from IEEE 802.11ac (adapted to S1G)

• OFDM
• MIMO + MU-MIMO
• PHY rates ranging from 150kbps to 347Mbps
IEEE 802.11ah: PHY (3)
Expected throughput vs. coverage
IEEE 802.11ah: MAC
• Need to reduce overhead: low data rates + short frames (typical in
some use cases)
• Short MAC headers and Beacons
• Implicit acknowledgement (no ACK needed)

• Need to support thousands of associated devices (increases coverage

 increases reachable STAs)
• Thousands of STAs  huge collision probability!
• Restricted Access Window (RAW): regular RAW
• Divide STAs into groups (AID)
• Split channel access into time slots
• Assign slots to groups (AP indicates RAW allocation and slot assignments in its
Beacons)
• Different backoff rules apply during RAW (due to different contention conditions)
Multihop Relay Operation
• Extend (root) AP coverage
• STAs will require lower tx power
• STAs may use faster MCS (less tx time)
IEEE 802.11ah: Summary
6. LTE-A
• Long-Term Evolution Advanced (LTE-A) is
a set of standards designed to fit M2M • LTE-A is a scalable, lower-cost protocol compared
communication and IoT applications in to other cellular protocols
cellular networks • LTE-A uses OFDMA (Orthogonal Frequency
Division Multiple Access) as a MAC layer access
technology, which divides the frequency into
multiple bands and each one can be used
separately
• The architecture of LTE-A consists of a core
network (CN), a radio access network (RAN), and
the mobile nodes
• The CN is responsible for controlling mobile devices
and to keep track of their IPs
• RAN is responsible for establishing the control and
data planes and handling the wireless connectivity
and radio-access control
7. LoRaWAN
• LoRaWAN is a newly arising wireless technology designed for
low-power WAN networks with low cost, mobility, security, and
bi- directional communication for IoT applications
• It is a low-power consumption optimized protocol designed for
scalable wireless networks with millions of devices
• It supports redundant operation, location free, low cost, low
power and energy harvesting technologies to support the
future needs of IoT while enabling mobility and ease of use
features
Agenda
01
Fog Computing Architecture for IoT

02
Protocols of IoT (ZigBee, IEEE 802.11ah, …)

03
Long range wide area network for IoT

04
Energy-efficient WiFi for IoT
Short range vs. long-range IoT
IoT-connectivity technologies
Multiple standards, different attributes
LPWA requirements
Low Power Wide Area wireless connects low bandwidth, low power
devices and provides long-range coverage
LPWA requirements
The most critical factors in a LPWAN are:
 Network architecture
 Communication range
 Battery lifetime or low power
 Robustness to interference
 Network capacity (maximum number of nodes in a network)
 Network security
 One-way vs. two-way communication
 Variety of applications served
IoT -the connectivity pyramid
Low-Power Wide-Area Networks
25 mW transmission power

Low-Power Wide-Area Networks

20 years on simple battery
 15-50 km rural outdoor

Low-Power Wide-Area Networks

2-3 km urban indoor
 No scheduling
No routing
ALOHA

Low-Power Wide-Area Networks

Device-initiated com
Huge densities
Low throughput
 250 kHz or less
Narrow-band
Low-Power Wide-Area Networks
 Collisions
Data-over-NAS In-band
Duty cycling
Acknowledgements Guard-bands
License free In licensed spectrum
250 kHz or less
Narrow-band
Low-Power Wide-Area Networks
 Collisions
Data-over-NAS In-band
Duty cycling
Acknowledgements Guard-bands
License free In licensed spectrum
No scheduling
250 kHz or less
Narrow-band
25 mW transmission power 15-50 km rural outdoor
No routing
ALOHA

Low-Power Wide-Area Networks

20 years on simple battery 2-3 km urban indoor Device-initiated com
Huge densities
Low throughput

100 bps 12 byte payload 140 messages 4 messages

(50 kbps max) (50 byte payload) uplink downlink
LPWA market opportunity
Cellular LPWA example applications
Real use cases being deployed now [NB-IoT]
Fixed measurements points, 3km distance from TV tower LoRa IoT station
WHERE DOES LPWAN FIT?
 WHAT IS LoRaWAN™?
LoRaWAN™ defines the communication protocol and system architecture for the network while
the LoRa physical layer enables the long-range communication link.
LoRa
What is it?
• LoRa technology was originally developed by a French company, Cycleo
(founded in 2009 as an IP and design solution provider), a patented spread
spectrum wireless modulation technology that was acquired by SemTech in
2012 for $5 million
• In April 2013, SemTech released the SX1272 chip, which was equipped with
LoRa technology
– At that time, FSK modulated European smart meter transceivers were used, with a
maximum transmission distance of 1 to 2 kilometers
– LoRa operated under the same conditions, and the transmission distance could be
more
LoRa Technology
Two major components

End device: ED
Base Station: BS
 LoRa modulation
• The use of signals with high bandwidth-
time product (BT>1) should make the
radio signals resistant against in band
and out of band interferences
• The use of sufficiently broadband chirps
should help to fight against heavy
multipath fading characteristic for
indoor propagation and urban
environments
LoRaWAN network protocol

• LoRaWAN network protocol is optimized specifically for

energy limited EDs
• LPWAN typically has star topology and consists of BSs relaying
data messages between the EDs and an application server
• The BSs can be connected to the central server via backbone
internet protocol (IP) based link, and the wireless
communication based on LoRa or GFSK modulation is used to
move the data between EDs and the BSs
Network Architecture
Strong Ecosystem Enables Customized Deployment
Network Options
Network Capacity: Adaptive Data Rate
Network Capacity: Single Data Rate
LoRa End Node
Three classes of EDs
LoRaWAN Performance
Data rates settings and frames characteristics
LoRaWAN Performance
Maximum throughput per LoRaWAN channel and ED
COMPARING LPWAN TECHNOLOGY OPTIONS
Conclusion
• LoRaWAN technology, like any other, has its own strengths and
weaknesses
– The high coverage and satisfactory scalability under low uplink traffic
– The most critical drawbacks are low reliability and potentially poor
performance in terms of downlink traffic
• LoRa can be effectively utilized for the moderately dense networks of
very low traffic devices which do not impose strict latency or
reliability requirements
Agenda
01
Fog Computing Architecture for IoT

02
Protocols of IoT (ZigBee, IEEE 802.11ah, …)

03
Long range wide area network for IoT

04
Energy-efficient WiFi for IoT
Wi-Fi: a New Contender of IoT
• Some low-power protocols do not currently enjoy ubiquitous
access to the Internet

IoT Wi-Fi (home) Vision

Low Power

Wide deployments

Long Range

Compatibility with Internet

Wi-Fi: a New Contender of IoT
• Need to support all
traffic demands
• Symmetrical design
is very inefficient
 Lower Sample Rate
in Receiver
to Made Energy Efficient

Enable Low-Energy By
Pushing Decoding Burdens
to the AP Side
Energy-Efficient
WiFi Support

W. Wang, Y. Chen, L. Wang, and Q. Zhang, “From Rateless to Sampleless: Wi-Fi

Connectivity Made Energy Efficient”, IEEE Infocom 2017.
 Wi-Fi Has A Power Problem
Need to support all
traffic demands

Use the same sampling rates for packet receiving

Heavy traffic Light traffic

High end Energy constrained
If Receivers Can Flexibly Select Sampling Rates
20MHz

A power reduction of 36%(30%) can be achieve

d by selecting ½ sampling rate [1]

20MHz 10MHz 5MHz

[1] Zhang et al. "E-MiLi: Energy-Minimizing Idle Listening in Wireless Networks." MobiCom 2011.
 The Challenge
Flexible Bandwidth
Need to modify
AP’s PHY
• Modifying existing infrastru
cture is costly

• Not compatible with legacy d

evices
Flexible Rates
(if AP adjusts bandwidth according to rx’s sampling rate)
 The Challenge
Fixed Bandwidth
• Rx’s rate < Nyquist rate

• Sparse recovery not work:

Not sparsity in today’s PHY

Cannot decode
Flexible Rates
(if AP uses legacy ADC rate and rx flexibly selects the sampling rate)
Idea: From Rateless to Sampleless
Rateless codes
AP uses highest modulation Gradually add redundancy in extra
schemes that Rx may not be able to transmissions until the packets can
decode under current SNR be decoded

Sampleless Wi-Fi
AP uses legacy bandwidth for Gradually add redundancy in extra
transmission, while Rx uses down- transmissions until the packets can
scaled sampling rates for reception be decoded
 Legacy Transmission

20MHz 20MHz
 Down-Sampled Receiver

20MHz 10MHz (rx is downsampled

by 2)

Freq. Alising
Cannot decode
 Sampleless Wi-Fi

20MHz 10MHz

send the packet twice

Decode over
multiple transmissions
Design Challenge: Adding Constellation Diversity
Rateless codes add redundancy at Tx

Modulation 1 010101… Modulation 2

Correlated
1st Transmission 2nd Transmission

Not compatible
Decode over multipleto legacy AP:
transmissions
Need PHY modifications
Solution: Exploiting Time-Shift Effect
010101…

Legacy Modulation
1st Transmission 2nd Transmission

Add 1redundancyTime-shift
Time-shift at Rx: 2
Compatible to legacy
Decode over multiple tr AP
ansmissions
Solution: Exploiting Time-Shift Effect

Time-Domain Freq-Domain
𝑥(𝑡) 𝑋 𝑓

𝑥(𝑡 − 𝜏) 𝑋 𝑓 𝑒 −𝑗𝜃
 Solution: Exploiting Time-Shift Effect
one example constellation map generated (QPSK with ½ sampling rate)

Confusion points: 3, 9, 12 Winner: 3

With more transmissions, the distance in constellation diagram

increases and it becomes easier to separate neighboring points
 Sampleless Wi-Fi: Reception Pipeline
1
Modulation M
1st Received Signal Map
Upper Layers
Distance D1
Delay FFT 011...
Matrix
Decoding
Pipeline
2nd Received Signal
010110...
Distance D2 Joint
Delay FFT
Matrix Demodulator
Modulation
Map combines multiple packets
M2
based on the maximum
likelihood (ML) algorithm
 Implementation & Evaluation
• Implemented reception pipeline in USRP N210
– In a 10m x 10m office Sender
Receiver

• Trace-driven evaluation for energy saving

– Collected iPhone 5s traces
 BER under Various SNRs
1.00E-01
1.00E-04
BER 1.00E-07
1.00E-10
4 8 12 16 20 24 28 32

Down-sampling’s negative effect on decoding is completely

SNR
eli
802.11-BPSK
minated by SamplelessENFOLD-BPSK
Wi-Fi
Sampleless-BPSK 802.11-QPSK
• Enfold [2]: state-of-the-art downclocking technique
• 802.11 Power Saving Mode (PSM)
*2+ F. Lu et al., “Enfold: downclocking ofdm in wifi,” in Proc. ACM MobiCom, 2014.
 BER under Various SNRs
1.00E+00

1.00E-02

BER 1.00E-04
802.11-QAM64
1.00E-06 ENFOLD-QAM64
Sampleless-QAM64
1.00E-08 802.11-QAM16
ENFOLD-QAM16
1.00E-10
1 2 3 4 5 6 7 8 SNR
 Energy Saving for Various Apps
1.5
88.7%
1 77.6%
0.5
0

PSM Enfold_1/2 Sampless_1/2

Sampleless Wi-Fi consumes 77.8%-88.7% energy at ½ Nyquist
rate.
Estimate the SNR based on the data rate
 Energy Saving for Various Apps
2
1.5
94%
1
66.8%
0.5
0

PSM Enfold_1/4 Sampless_1/4

Sampleless Wi-Fi consumes 66.8%-94% energy at ¼ Nyquist
rate.
 Quick Summary

• Sampleless Wi-Fi provides reliable communications between

legacy APs and low-power devices with various sampling rates

• It leverages the wisdom of rateless codes for under-sampled

packets decoding

• It creates constellation diversity at Rx using the time-shift effect

 Lower Sample Rate
in Receiver
to Made Energy Efficient

Enable Low-Energy By
Pushing Decoding Burdens
to the AP Side
Energy-Efficient
WiFi Support

W. Wang, S. He, L. Yang, Q. Zhang, and T. Jiang, “Wi-Fi Teeter-Totter: Overclock

ing OFDM for Internet of Things”, IEEE Infocom 2018.
The Spectrum Efficiency for Massive IoT
Connectivity

Wi-Fi for IoT devices need a mature and spectrum efficient

multiplexing access technology -- OFDM
 Recent Research
• E-MiLi (ACM MobiCom, 2011)
– Downclocking receiver’s clock rate during idle listening

• Enfold (ACM MobiCom, 2014); Sampleless Wi-Fi(IEEE INFOCOM, 2016)

– Downclocking for OFDM-based Wi-Fi by leveraging the gap between
modulation and SNR

AP  IoT: downclocked OFDM transmission

IoT AP: standard Wi-Fi OFDM transmission
 Symmetrical Design
Medium power

Limited energy
Symmetrical design is very
inefficient
 Idea: Transceiver Asymmetry
Medium power Low power

Transmit at the Receive the signal with

lowest power overclocking
 Overclocking Opportunities

1x: 64 samples
1 2 31 32 63 64

2x: 128 samples

1 2 3 4 63 64 65 66 125 126 127 128

 Overclocking Opportunities
1x: 64 samples Conventional signal
1 2 31 32 63 64

2x: 128 samples Time-shifted signal

1 2 3 4 63 64 65 66 125 126 127 128

𝜏
Interpolated samples
 Phase Rotation of Shifted Signal
Time-Domain Freq-Domain
𝑥(𝑡) 𝑋 𝑓
Time shift FFT Phase rotation

FFT
𝑥(𝑡 + 𝝉) 𝑋 𝑓 𝑒 𝑗𝜽
2𝜋𝒇𝜏
𝜽=
𝑁

Time-shifted samples result in different phase rotations at different subcarriers

 Phase Rotation in Real World

• The phase shifts across all subcarriers in a real Wi-Fi packet when
received at eight-fold clock
 Joint Decoding
…… OFDM symbol OFDM symbol ……

FFT FFT
Phase
compensation
Channel Channel
equalizer equalizer
Constellation
Demapper

01001010……
 Implementation

• Implemented on GNUradio/USRP
platform
• Operates on a 2 MHz or 1 MHz channel
Transmitter
with 52 subcarriers are carried data
values
• Evaluated BPSK, QPSK, 16QAM, and
64QAM modulations

Receiver
 Evaluation – Sync Error

The average synchronization error at 8× clock rate is merely

13% at the standard clock rate at low SNR.
Evaluation – Modulation Scheme

For all modulations scheme, T-Fi outperforms the

standard 802.11 receiver
Evaluation – Wireless Environment

T-Fi achieves stable performance gain over a wide range of

wireless environment
 Quick Summary
• We introduce an asymmetric transceiver paradigm for IoT that
pushes power burden to the AP side

• We propose a reception pipeline to decode legacy packets at

lower SNRs than the conventional transceivers

• We implement the T-Fi system, and the evaluation confirms

the benefits of T-Fi in real environments
End of This Chapter