SnW: Internet of Things and enabling technologies

Firma convenzione
Politecnico di Milano e Veneranda Fabbrica
del Duomo di Milano
Aula Magna – Rettorato
Mercoledì 27 maggio 2015
Lesson 2: Internet of Things and enabling
technologies
Luca Cerina – luca.cerina@polimi.it
Ver. Updated – 30/11/2017

Luca Cerina, R.A. @ NECSTLab – DEIB
The Cloud

The Internet of Things

The Quantied Self
“Not all important things in life can be measured and not
everything that can be measured is important”
(Quantified Self Institute)

The Hackathon
Device design
Software &
infrastructure

Device design: outline
●
Existent technologies
●
NFC
●
Bluetooth
●
LoRa
●
Bird’s eye view on: Fog Computing
●
Focus on: design for human interaction

NFC/RFID
●
The reader uses the antenna to
constantly scan the area sending
radio waves
●
The transponder send its ID, with an
active signal (active RFID) or
piggybacking the reader waves
(passive RFID)
The frequencies are:
●
LF (125-134KHz) 1-10cm→
●
HF (13.56MHz) 1m (NFC)→
●
UHF (865-960MHz) long range→

NFC/RFID: Example
RC 522 RFID READER
Frequency: 13.56 MHz
Connection protocol: SPI
Support: Read/Write
Usage: Arduino/Microcontrollers

Call for Ideas
What would you design?

Bluetooth Low Energy
Bluetooth LE enable short-burst wireless
connections and uses multiple network topologies:
●
Point-to-Point (1:1) data transfer and fitness→
trackers
●
Broadcast (1:many) beacons, information→
sharing
●
Mesh (HOT! NEW!) (m:m) large scale topology→
for sensor networks, automation, devices chatting

Architecture

BLE Radio
●
Frequency: 2.4 GHz ISM
(Industrial Scientific
Medical) band
●
40 channels with 2 MHz
spacing
●
Range 0 – 100m (BT
5.0)
●
Enabling technology
invented by Hedi Lamarr

BLE Roles
Advertiser broadcasts adv. packets, can be→
connectable or not
Scanner listen to adv. Packets, passively or→
actively (scan request)
Master/central device connected to→
slave/peripherals
Slave/peripheral device connected to Central→
Mesh node advertise and scan at the same time,→
performs drop logic on the packets

BLE Generic Atribute Proile
Provide access to device discovery & connection
control
GAP defines roles:
●
Broadcaster advertiser→
●
Observer scanner→
●
Peripheral always slave→
●
Central always master, never advertise→

BLE GAP Modes
Connectable
●
Can make a connection
●
Not connectable / connectable
Discoverable
●
Can be discovered (advertising)
●
None, limited, general
Bondable
●
Can pair with a device for long term connections
●
Not bondable / bondable

BLE Generic ATTributes (GATT)
●
GATT defines only two roles: Server and Client
●
The attribute is the smallest data element available, it is
defined by a 128 bit Unique Identifier (UUID) or a 16bit
handle.
An attribute can be: Readable, Writable, R/W, None

GATT hierarchy
GATT defines a precise hierarchy
for data and reusability.
Service/Profile:
conceptually related attributes
(e.g. heart rate)
Characteristic:
single value or metadata
container
Descriptor:
single value, which meaning its
defined by the characteristic
metadata.

GATT hierarchy

How to use BLE in your applicaton
Raspberry Pi & co:
→ Bluepy ( Python )
→ BlueZ ( C )
Microcontrollers:
→ Redbear Nano/Duo (redbear.cc)
→ Arduino 101 (arduino.cc)
→ SensiBLE (sensiedge.com)
Smartphone
→ Bluetooth core libraries

LoRaWAN: Long Range Wide Area Network

LoRaWAN: Long Range Wide Area Network
Long range communication link
●
Up to 2.5km in cities
●
Up to 10km in countryside
Frequency (Europe): 868 MHz
Low data rate: 0.3 – 50 kbps

LoRaWAN structure

LoRa end-device
Every network is identified by a 7bit identifier (NwkID, not
unique).
Private networks use 0x0 as ID.
Each device has a 25bit address, static or given by the
network at join-time.
The full DevAddr is 32bit = NwkID + deviceID

LoRa end-device
Every device should comply to a specific “application” and
use a 64bit IEEE AppEUI value.
To increase security, each end-device uses two 128bit AES
symmetric keys:
●
NwkSKey protection from network artefacts→
●
AppSKey protection on application layer traffic→

LoRa dynamic join
End-device could also avoid pre-provisioning and perform
a join request to the network. In this case the device needs:
●
A 128 bit unique appKey for encryption
●
A valid 64 bit AppEUI
●
A device ID which should be a valid EUI called devEUI
If the join is successful, the server answers with a join
accept and provide DevAddr, NwkSKey and AppSKey
through the downlink

LoRa devices
Nodes
→ The things UNO
→ Arduino MKR WAN 1300
→ Adafruit Feather LoRa
Gateways (~200-250 )€
→ Multitech Conduit
→ Lorrier LR2
→ The things gateway

How to connect
Free, community based, LoRa network

The rise of Edge/Fog computng
https://nordicapis.com/what-is-fog-computing/

State of the art: challenges
31
To represent the real breakthrough in Cloud and
IoT technologies, the Fog needs to be:
Reliable and
scalable [1]
●
Deployment plan of
nodes
●
Orchestration models
●
Nodes interaction
control
[1] Z. Wen, R. Yang, P. Garraghan, T. Lin, J. Xu, and M. Rovatsos, “Fog orchestration for
internet of things services,” 2017.

Secure [2]
●
Resistance to attacks
●
Trusted authentication
systems
●
Detection of rogue
nodes
[2] A. Alrawais, A. Alhothaily, C. Hu, and X. Cheng, “Fog computing for the internet of
things: Security and privacy issues,”, 2017.

Manageable [3]
●
Workflow policies
●
Data migration
●
Scheduling at runtime
[3] Z. Hao, E. Novak, S. Yi, and Q. Li, “Challenges and software architecture for fog
computing,” 2017.

FPGAs in the Fog
The OpenFog consortium technical
reference includes hardware
accelerators such as Field
Programmable Gate Arrays (FPGA)
to:
“Provide supplementary computational throughput” and
“satisfy both latency and power constraints as it relates to
a given scenario”
Minimal power consumption
and high performance
Extreme re-programmability
Faster time-to-market and
lower costs

Proposed architecture

Interacton layer
Sensors produces data, Actuators consume it in form of commands, and
both communicate with the Mesh Layer above them.
The Human-in-the-loop can consume and produce data, and act
upon it (e.g. application interaction or physical outcomes).
All the elements communicate through wireless protocols (BLE 4.2 in the
current deployment)

Fog node and cloud
The Fog Node acts as a central point of communication between the
Mesh and the Cloud. It comprises an ARM CPU and an FPGA accelerator
(Xilinx Zynq 7000 Architecture)
The FPGA circuit can be reprogrammed at runtime to handle specific
tasks, like data aggregation from sensors towards the Cloud, or to directly
act on physical actuators without the intervention of Cloud computing.

Design for human interacton
One basic principle:
KISS

Them?

Of course not

“Interaction design is about shaping digital things for
people’s use” (Jonas Lowgren)
“An interaction designer must become an expert in
how human beings relate to each other, and to the
world, and to the changing nature of technology” (Jon
Kolko)

Design thinking process
https://www.interaction-design.org/literature/article/5-stages-in-the-design-thinking-process

Empathise
CHALLENGE ASSUMPTIONS

Deine
https://www.interaction-design.org/literature/article/stage-2-in-the-design-thinking-process-
define-the-problem-and-interpret-the-results

Ideate
Brainstorm, Brainwrite, Worst Possible Idea, SCAMPER

Prototype
INEXPENSIVE, scaled down versions of the product/feature to be analyzed

Test
Pay close attention to the concepts of COGNITIVE FRICTION and OVERLOAD

Last take away
www.shutterstock.com
User engagement / Actionable items
No direct meaning in the numbers
Abandon rate up to 33% after 6 months
Jacket1234 @Deviantart
Data quality
Need for consistency and validation
Mean heart rate Insights with low valence→
Kasey McMahon
Technology
To integrate wearables with multi-task smart
ambient devices

SnW: Internet of Things and enabling technologies

More Related Content

What's hot (20)

Similar to SnW: Internet of Things and enabling technologies (20)

More from NECST Lab @ Politecnico di Milano (20)

Recently uploaded (20)

SnW: Internet of Things and enabling technologies