Internet of Things

Module 1: Introduction and Course Overview

UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN
Background:
How the Internet Works
 1. Internet Architecture
• How is the Internet built?

Background: 2. Networking Routing

How the Internet

• How does the Internet
figure out paths to
destinations?

Works 3. Network Devices

• What does the inside of a
router look like?
 How Can Two Hosts Communicate?
Clock

Data
1 0 1 0 0 1 1 1 0 0 1
Manchester
(as per IEEE 802.3)

• Encode information on modulated “Carrier signal”

• Phase, frequency, and amplitude modulation, and combinations thereof
• Ethernet: self-clocking Manchester coding ensures one transition per clock
• Technologies: copper, optical, wireless
5
How Can Many Hosts Communicate?

• Naïve approach: full mesh

• Problem: not very scalable
• >25B devices connected in 2019
6
How Can Many Hosts Communicate?

• Better approach: Multiplex traffic with routers

• Goals: make network robust to failures and attack, maintain spare capacity, reduce
operational costs
• New challenges: What topology to use? How to find paths? How to identify destinations?
7
How Can Many Hosts Communicate?

8
 How Can Many
Robert’s local
DNS server
Hosts Communicate?
.com authoritative
DNS sever
Robert Routing Table at B
Prefix IF Hops
10.1.8.7 4.0.0.0/8 D 1
23.2.0.1 IP address
Dest=4.5.16.2 A
4.9.0.1
10.1.0.0/16 B
23.2.0.0/24
4.18.5.1
Prefix
C D
10.1.0.1 81.2.0.0/24 4.0.0.0/8
Routing Table at A
Prefix IF Hops
B 2 81.2.0.1 4.5.16.2
youtube.com
4.0.0.0/8

Routing Table at C
Prefix IF Hops
4.0.0.0/8 D 1
youtube.com’s authoritative
DNS server
• Hosts assigned topology-dependent addresses
• Routers advertise address blocks (“prefixes”)
• Routers compute “shortest” paths to prefixes
• Map IP addresses to names with DNS
9
 What is a Protocol? Can we meet next
week sometime?

Sure, I’m free

Tuesday
afternoon and
How about
Thursday
Tuesday 5pm?
evening.

I’d prefer earlier.

Ok!
Tuesday 2pm
ok?

Ok, great!
See you then!
What is a Protocol?
• Sequence of communications used to conduct some activity in a
distributed system

• Protocols are widely used in networks

• Figure out how fast to send data, discover paths to destinations,
replicate data, encode data into transmittable patterns, etc.

• Protocols often organized into “suites” or “stacks”

• Handle collection of activities associated with particular environment
• Examples: TCP/IP (Internet), Infiniband (Data Center), Bluetooth (IoT)
Networks Have Protocols To….

Compute paths through networks Routing protocols

Figure out how fast to send data Transport protocols

Encrypting messages so others can’t
Encryption protocols
read them
Figure out who has an address Address resolution protocols
Figure out what kinds of things the
Service discovery protocols
network can do
 The TCP/IP Protocol Stack
Layer 7 Application IMAP, HTTP,
H.323, RDP, SSH,
Layer 6 Presentation Application BitTorrent, Bitcoin
JSON/XML
Layer 5 Session
TCP, UDP, QUIC,
Layer 4 Transport Transport SCTP
BGP, OSPF, IS-IS
Layer 3 Networking Networking IPv4, IPv6
BFD, ICMP

Layer 2 Datalink Ethernet, Wireless

Datalink &
Ethernet (WiFi, ARP, DHCP
Physical 802.11)
Layer 1 Physical
OSI Model TCP/IP Model Data Control
(1984, International Standards (1973-1989, Cerf/Kahn/DARPA) Protocols Protocols
Organization)
Protocol Encapsulation
Data Application

Transport
TCP Flags
Port Numbers

Destination Sender IP
IP Address Address Networking

Destination Sender MAC

MAC Address Address Datalink &
Physical

• Each layer of protocol stack encapsulates data passed to it

• Each forwarding layer inspects data only at that encapsulation layer
• Switching only looks at Ethernet header, Routing only looks at IP header, etc.
• Terminology: “Layer-3 switch”, “Layer-4 load balancer”, “Layer-7 load balancer”
 How Can Many Hosts Communicate?
Routing Table at B
Prefix IF Hops
10.1.8.7 4.0.0.0/8 D 1
23.2.0.1 IP address
A
4.9.0.1
10.1.0.0/16 B
23.2.0.0/24
4.18.5.1
Prefix
C D
10.1.0.1 81.2.0.0/24 4.0.0.0/8
Routing Table at A
Prefix IF Hops
4.0.0.0/8 B 2 81.2.0.1 4.5.16.2

Routing Table at C
Prefix IF Hops
4.0.0.0/8 D 1

• Hosts assigned topology-dependent addresses

• Routers advertise address blocks (“prefixes”)
• Routers compute “shortest” paths to prefixes
• Map IP addresses to names with DNS
17
Scenario: Sending a Letter

A B

603 S Wright St. 308 E Green St.

Name: B
Address: 308 E Green St.
Champaign, IL
61820

201 N Goodwin Ave

Scenario: Address Allocation

61801
A B
61822

603 S Wright St. 308 E Green St.

Name: B
Address: 308 E Green St.
Champaign, IL
61820
61820
Scenario: Access Control

C
“Inspect mail to
308 E Green St.” D
308 E Green St.
Name: C

603 S Wright St

Name: C
Address: 308 E Green St.
Champaign, IL
201 N Goodwin Ave 61820
Internet Addressing:
Different Layers Use Different Addresses
URLs (e.g., http://www.cs.Illinois.edu/index.html)
• Identifies “resources” – files, content
Application
Domain names (e.g., illinois.edu)
• Identifies groups of computers under single administrative control

Port numbers (e.g., 143 for IMAP, 22 for SSH)

Transport • Identifies application running on a computer

Networking IP Addresses (e.g., 18.220.149.166)

• Identifies location where a computer’s interface attaches to the Internet

Datalink & MAC Addresses (e.g., 9C-EF-D5-FE-27-72)

Physical • Identifies specific computer interface

 All these addresses are used for end-to-end communication

 Internet Addressing:
MAC Address vs IP Address
MAC Address IP Address
Datalink layer Network layer

Hierarchically-assigned, location-dependent
Flat (location-independent) identifier
identifier
Like a social security number
Like a postal address
Usually hard-coded, requires no
Needs to be manually configured, assigned by
configuration
DHCP
Portable; can stay the same as the host moves Not portable; must be changed if host changes
networks
Used to get packet to destination on same
Used to get packet to destination IP subnet
LAN

Example: 9C-EF-D5-FE-27-72 Example: 18.220.149.166

Can We Use TCP/IP for IoT?
Yes

But, IoT introduces additional challenges:

• Very tight power/compute constraints
• Need to work closely with wireless
• Need to address applications, not just interfaces

Also, creating new protocols can help lock-in and market control
• Bad for innovation but good for security
Common IoT Protocols
Different IoT Protocols
for Different Environments
Data Rate
Wireless LAN (WLAN) Low-Power Wide-Area Network (LPWAN)

High
(Gbps)
Low Power Wireless LAN (LPWLAN)

Medium
(Mbps)
Low Power Wireless Personal Area
Networks (LPWPAN)

Low
(Kbps)

Communication Range
Short Medium Long
Zigbee Protocol Stack
Application Layer (APL)
Application Zigbee Device Application
Framework Object Support Sub-layer
Defined in the
ZigBee Standard
Network Layer (NWK)
Security Message Routing Network
Management Broker Management Management

Medium Access Control (MAC)

Security Message Routing Network
Management Broker Management Management
Defined in the
IEEE 802.15.4 Standard
Physical Layer (PHY)
2.4 GHz 915 MHz 868 MHz
(250 kbps) (40 kbps) (20 kbps)
 Bluetooth Low Energy Protocol Stack

Application

Application
Layer Generic Attribute Generic Access Host
Functions Protocol (GATT) Profile (GAP)

Attribute
Protocol (ATT)

Security
Manager
Network Layer
Functions Logical Link Control and Adaptation
Layer Protocol (L2CAP)

Link Layer (LL) Controller

Low Energy Physical Layer (LE PHY)

Is the Internet Just One Network?

A
B

C D

30
How Does Internet Routing Work?
Internet routing works on two levels:

• Each AS runs an intra-domain routing protocol internally

• Establishes routes to internal prefixes and between routers
• Example protocols: OSPF, IS-IS

• Each AS runs an inter-domain routing protocol on links to

neighboring ASes
• Establishes routes to external destinations
• Border Gateway Protocol (BGP)
 Intra- vs. Inter-domain Routing
dest

source Sprint

AT&T
BGP session

• Run “Interior Gateway Protocol” (IGP) within ISPs

• OSPF, IS-IS, RIP
• Use “Border Gateway Protocol” (BGP) to connect ISPs
• To reduce costs, peer at exchange points (AMS-IX, MAE-EAST)
“Source: XO Communications / http://tinyurl.com/y45z65pz”
L2 Switching vs L3 Routing:
Routing Proactively Builds State
Dest Nexthop Dest Nexthop
Control 17.0.0.0/24 fe3/0 17.0.0.0/24 fe0/1
Messages

0D:73:CA:F3
17.3.6.2 Dest Nexthop

17.0.0.0/24 fe0/1

F5:5A:21:03
Dest Nexthop
29.51.130.9
17.0.0.0/24 fe7/1
Hey everybody –
I own Dest Nexthop
17.3.6.0/24! 17.0.0.0/24 fe0/0
L2 Switching vs L3 Routing:
Switching Relies on Broadcast
Dest Nexthop
Data 0D:73:CA:F3 fe0/1
Messages

0D:73:CA:F3
17.3.6.2

F5:5A:21:03
29.51.130.9
Here is a data
Say… I received a data packet with src MAC
packet I want to
address 0D:73:CA:F3 on fe0/1…
send to
Next time I get a packet destined to 0D:73:CA:F3
F5:5A:21:03!
I know what direction they are…
Virtualizing Ethernet with VLANs
Guest Corporate Corporate Credit Card
• Divide up hosts into logical printers printer Server
groups called VLANs VLAN
• Like virtual machines, but for
LANs (creates “virtual
networks”)
• VLANs isolate traffic at layer 2 Corporate
B1
• Each VLAN corresponds to IP workstatio
subnet, single broadcast n
domain
• Ethernet packet headers have
VLAN tag
• Bridges forward packet only Guest Guest
on subnets on corresponding workstations
VLAN VLAN
Delivery Models
A

B C
• Unicast
D • Broadcast
E
• Multicast
I
• Anycast
H
F G J
Source
Delivery Models
A

B C • Unicast
• One source, one
D destination
E
• Widely used (web,
I cloud, streaming;
many protocols)
H
F G J
Delivery Models
A

B C • Broadcast
• One source, all
D destinations
E
• Used to
I disseminate control
information,
H perform service
F G J discovery
Delivery Models
A

B C • Multicast
• One source, several
D (prespecified)
destinations
E • Used within some
I ISP infrastructures
for content
H delivery, overlay
F G J networks
Delivery Models
A

B C • Anycast
• One source, route
D to “best”
destination
E
I • Used in DNS,
content
H distribution, service
F G J selection
Multicast: Source-Specific Trees
A
• Each source is the
B C root of its own tree

D • One tree per source

E • Tree consists of
I
shortest paths to
H
each receiver
F G J

Member of Sender to
multicast group
UIUC, Spring 2010
F multicast group
Multicast: Source-Specific Trees
A
• Each source is the
B C root of its own tree

D • One tree per source

E • Tree consists of
I
shortest paths to
H each receiver
F G JJ

Member of Sender to
multicast group
UIUC, Spring 2010
F multicast group
Multicast: Shared Tree
A
• One tree used by all
B C members of a group

• Rooted at “rendezvous
D
RP point” (RP)
E
I • Less state to maintain,
but hard to pick a tree
H that’s “good” for
F G JJ
everybody
45
Multicast: Shared Tree
• Ideally, find a “Steiner
A tree” minimum-weighted
tree connecting only the
B C
multicast members
– Unfortunately, this is
D NP-hard
RP

E
I
• Instead, use heuristics
– E.g., find a minimum
H
F G JJ spanning tree (much
easier)
46
Example Applications
• So many applications, scenarios, use cases for IoT
• Seems hard to digest it all

• But beneath it all are commonalities

• Common architectures, protocols, designs

• This lecture: some walkthroughs to give you a taste

• Example applications and solutions for IoT
Environmental Monitoring
• Earth is very important to humans
• Air we breathe, food we eat, comes
from earth
• 90% of human diseases (and
medicines) come from wildlife

• Important we understand
environment
• Global warming reducing arable land,
honeybees disappearing, pollution
kills millions of people
Environmental Monitoring: Wildlife
• Earth is facing its 6th major
extinction event
• 10,000 species go extinct every year
• # species halved in last 40 years
• Comparable to “Snowball Earth” and
the asteroid that wiped out the
dinosaurs
• Threats: Escalating poaching, human
encroachment, climate change,
disease
• Understanding the problem can
help us solve it
Environmental Monitoring: Wildlife
• Animal monitoring an essential part of almost all conservation
efforts
• If they are ill, injured, caught in a trap, we can find and help them
• Big changes in migration patterns, population density
• Elephants, Whales, Tigers, Macaws, etc. on the verge of extinction

• IoT: Check up on animals continually, rather than once a day or

by spotting
• Enables studies of new dynamics: “interactions” and social habits,
movement patterns; rapid response to poaching events
 Environmental Monitoring: Wildlife:
How it’s Done

• Animal outfitted with collar containing sensor array, storage, and

networking
• GPS sensor, accelerometer/gyroscope/magnetometer, biometric sensors,
flash memory, wireless transceivers, and CPU
• Considerations: Weight limit (e.g., 3-5 pounds), lifetime (e.g., 1 year with
no human intervention)
Wildlife Tracking Architecture
• No pervasive
Storage/
Logging

infrastructure such as
antenna towers Directional
antenna
• Plane flyovers – listen
for pings from collars RF
Range
• Peer-peer Tracking
collar Cell
communication to Tower
replicate info across
collars
• E.g., use gossip
opportunistically during
encounters
Environmental Monitoring: Wildlife:
Sample Findings (Zebra Monitoring)

Juang et al., 2002

• Movement patterns: grazing, graze-walking, occasionally fast-moving

• Zebras tend not to sleep deeply
• Seek out water about once a day, drink relatively quickly
Other Kinds of Environmental Monitoring

• Forestry monitoring
• Illegal logging, land-use/species changes, health, fire prediction
• Study physical characteristics: tree height, diameter at base, stem
density, canopy/foliage density, discolorations, water content
• Challenges: no fixed infrastructure, large multipath effects (long-wave,
relays, satellite)
Other Kinds of Environmental Monitoring

• Marine Animal monitoring

• Sea turtles, coral reefs, mercury exposure in fish
• Migration patterns, pollution exposure, breeding/nesting patterns
• Challenges: underwater communication (sonar/long-wave)
Smart Homes
• Controls lighting, climate, entertainment systems, and
appliances
• Wireless speakers, thermostats, home security systems, domestic
robots, smoke/CO detectors, energy brokers, lighting, door locks,
refrigerators, laundry machines, flood/water detectors
• Example Applications
• Automating chores: watering lawns,
• Turning on/off lights as you move between rooms
• Automatic doorbell based on presence detection
• Automatically adjusting thermostat based on learned occupancy
patterns
Smart Home Architecture
IoT Protocols (Zigbee, IoT Hub
Zwave, BTE, WeMo, Thread,

• Controlled devices etc).

connected to
“gateway/hub”
802.11
(Wifi)

• Currently, few accepted

Hue SmartBulb

industry standards Wifi

Access
(to DOCSIS
• Companies hide
Point
Headend)

documentation to
prevent independent
Ethernet
DOCSIS
(wired)
development Cable Modem

• Poor release/patching
802.11
SmartFridge
(Wifi)

practices lead to security

issues
• Estimated 87% of devices
vulnerable Laptop
Smart Lock
 Smart Homes:
Elderly Care Monitoring
• 65+ years USA population 35M in 2000,
expected to double by 2030
• Poor health conditions; require help in
times of need

• Falls are extremely dangerous

• 1 in 3 adults over 50 dies within 12
months of suffering a hip fracture

• Deteriorating memory can lead to behavior

changes and other lifestyle difficulties
• Forgetting to exercise, eat, report for
doctor appointments
• Difficulty evacuating in emergencies
 Smart Homes:
Elderly Care Monitoring
Activities of Daily Living monitoring
• Watch occupant, ensure they are
following daily routines (eating,
drinking, exercise)
• Wellness Determination: learn
wellness profile of occupant; detect if
occupant is ill, suffering behavioral
changes, or otherwise in need of help
• Environmental monitoring: sufficient
food, air quality, temperature
• Partnership: question answering,
robotic caretakers, robotic pets
Smart Buildings

• Commercial and industrial buildings

• Common protocols: BACnet (traditional), Zigbee, Broadband

over Power Lines (IEEE 1901), Wifi (IEEE 802.11)
Smart Buildings: Applications
• Energy management
• Manages interplay between internal energy producers/consumers and
intelligent purchase of energy from the grid
• Occupancy detection adjusts lighting and temperature to personalized
settings
• Morning Warmup – bring building to setpoint just in time for occupancy
• Dim lighting or reduce cooling to respond to grid demand response
incentives
• Networking between smart buildings
Smart Buildings
• Navigation: Guide people to destinations via personal
assistants, route optimization, improve discretion, balance traffic
• Security:
• Identify people at point of entry (facial recognition, key, etc)
• Access control to rooms, route people away from sensitive areas
• Lighting and displays help organize orderly and fast evacuations
• Pressure, humidity, CO/CO2/refrigerant/biological/chemical sensors –
detect if ventilation systems failed or become infected with
contaminants
• Temperature alarms: chilled/hot water supply, supply air, valve
indicators, current sensors (slipping fan belts, clogging strainers at
pumps, etc)
 Smart Building Architecture
Intra-building
Switches
(Ethernet/IP
Routing)

Per-floor
distribution
switches
(Ethernet)

Building
Management Lighting Alarm Security
System (BMS) Controller Panel Headend

Variable Air CCTV Communication

Volume (VAV) DALI Pull hubs,
interface Station Controllers,
and Devices
Building Operations Center Fan Coil Unit Fingerprint/
(BACnet, Zigbee,
(FCU) Access Card
Smoke/ BPL, Wifi, etc.)
Occupancy
Therm (Power over
sensor
Sensor Ethernet: PoE)
Heat Pump Door
Controller

Horn/ Microwave/
Chilled Beam Lights IR/Laser Sensor
Strobe

Energy and Lighting Fire Alarm Access

Power metering Control System Control
Smart City

• Uses sensors and actuators to manage resources and assets

more efficiently
• Power/water/gas supply networks, waste management, police/fire,
transportation systems, schools/libraries/hospitals
Smart City Applications: Traffic Control

• Use AI to manage and route • Track criminals, detect

unsafe/illegal behavior
pedestrian and car traffic
• Balance traffic flow, detour
efficiently around hotspots, smooth flow
• Study: time spent at lights • Prioritize emergency response,
reduced by 40%, travel times assist evacuations
reduced by 25% • Electronic alerts to drivers, road
conditions and emergencies
Smart City Applications: Smart Grid

Power Plant Distribution Grid Smart Meters Smart Appliances

• Electrical grid which leverages IoT • Intelligent appliances in homes “negotiate”

with grid
to improve efficiency and resource • Congestion/load pricing
management • Prediction of workloads to drive power
• Electronic power conditioning, control station ramp-up/provisioning
of power distribution and usage • Distributed storage (car batteries),
distributed generation improve efficiency
 Smart City: Network Architecture
Emergency Plan Command

Applications
border E-Governance
routers firewall IPS
Fiduciary/Accounting
IP Core
3rd Party/Citizen Apps
Internet

IPS
firewall

Service Data Center/Cloud

Domains
(computing, data, analysis)

Electrical power Traffic Control Critical services Medical services Operation Center
services (smart grid, (monitoring, (fire, police, (hospital,
smart metering) parking) security) paramedics,
medivac)