talking with things: what's hot in low-power long-range iot connectivity
TRANSCRIPT
Talking with Things
ECHELON MALAYSIA - 2017, PENANG, APR 13
What's Hot in Low-Power Long-Range IoT Connectivity
Disclaimer
English is not my mother tongue. bear with me
What’s the heck an Indonesian doing here in Penang :) ?
Well… I might have some stories/experiences to share with you
Internet of Things Connectivity
First coined in 2009 by Kevin Ashton, RFID pioneer and cofounder of the Auto-ID Center at the Massachusetts Institute of Technology (MIT)
Internet of Things
“Network of physical objects with embedded electronics, software, connectivity, and people to enable connectivity to exchange data, for intelligent applications and services„
Internet of Things
Things Connectivity People & Processes
Data Data
Internet of Things
A bit of story in 2003around IoT connectivity
My final year project in 2003 was kind of IoT
Design and Implementation of Home Lighting Control System and Home Monitoring System
using Mobile Phone over Internet
Circa 2003, original archive
Lamps
Web Cam
Modem for SMS & GPRS
Home Server*
GPRS GPRS
My final project Architecture
“The Thing”
*Obviously it’s not based on single-board computer (SBC). SBC is non-existent commercially back then
GPRS 2.5G, 40 - 80 kbps in theory
Mobile app Home Server app
Achieved 12 secs/frame NOT 12 frame/secs (fps) :)
My final project Video streaming feature
Circa 2003, original archive
J2ME app, on Nokia 3650 J2SE-based
That’s what I had in 2003for wide area IoT connectivity
CONNECTIVITY is one of the biggest challenges to
creating a true IoT
…yet, it always fascinates me!
Fast forward to now
oh hi…
Before continuing…
Andri Yadi
fb.com/andri.yadi | a at dycode dot com
http://andriyadi.me | http://dycodex.com
makestro.comCEO
ambassador
150+ Speakings
30+ talks about IoT for the past 2 years
Proudly coder for 19 years
ASM, QBasic, Pascal, c, C++, Java, PHP, Bash, C#, Visual Basic, HTML, JavaScript, Python, Objective-C, Swift
.NET, Qt/QML, Java ME/EE/SE, Android SDK, iOS SDK, Node.js
ARM MBED, ESP8266, ESP-IDF, Arduino
Cellular IoT (CIoT) connectivity we NOW have
Fast, efficient Up to 10 Mbps for 4G LTE
Ubiquitous coverage
Reliable & secure
Not designed for IoT in mind
High power consumption
Relatively expensive: modules, data plan
Provisioning, manageability
Advantages Considerations
Cellular IoT (CIoT) connectivity we WILL have
EC-GSM-IoT
LTE-M / eMTC
NB-IoT
Low data throughput
Low power
Low device & deployment cost
Extended coverage
Technologies Common Traits
3
Evolution of IoT Connectivity in 3GPP/GSMA5 MHz200 kHz 1.4 MHz 5/10/15/20 MHz
Other influences
GSM LTE Cat-1+
Delay Tolerant Access
Cat-0
Cat-M1
Cat-NB1EC-GSM
UMTS
� GSM is the original wide-area M2M wireless connectivity technology. EC-GSMenhances it to keep it competitive.
� UMTS did not see any significant push towards a low-power variant.
� LTE-M (Cat-M1) is a concession to the low-power/low-throughput device within mainstream LTE.
� NB-IoT (Cat-NB1), a new RAN technology, is the official LPWAN contestant from the 3GPP/GSMA stable
Evolution of IoT Connectivity in 3GPP/GSMA
LPWA Competing technologies
Communication Technologies - Overview
Mbps
Kbps
bps
10 m 100 m 1 km 10 km
Baud rate
Range
Wi-Fi / BT
Short Range LPWAN
ST Confidential
Cellular
-M
-NB-IOT
5G
850/1900 MHz900/1800 MHz
Sub-GHz
2.4 GHz
WIFI/BT
Short Range LPWAN
Cellular
IoT Connectivity: Range vs Speed
03 | WHITEPAPER
Last mile connectivity options
A device, machine, appliance or vehicle can beconnected in many different ways. It can use awide area network to relay data directly to aserver or it can use a short-range technology,such as WiFi or Bluetooth, to connect to a meshnetwork, hub or gateway. Wide area networkstend to be less complex than mesh networks asthe endpoints can be connected directly to agateway or base station, rather than relying on arelay system to transmit messages.
There are many different wide area and short-range technologies available, but they can bedivided into two categories: those that aredesigned for high throughput rates and,therefore, have relatively high powerconsumption and those that are designed tominimise energy consumption and, therefore,provide relatively low throughput. Conventionalcellular technologies and WiFi tend to offer highthroughput rates, while new low power wide area(LPWA) technologies and Bluetooth are lowenergy technologies (see graphic).
Low power wide area options
Designed specifically to support IoT, LPWAtechnologies are optimised for use in low costdevices that need to transmit small amounts ofdata. The objective is typically to provide low costand energy-efficient connectivity for a largenumbers of devices in a small geographic area.Such devices include sensors that can monitor cityinfrastructure, environmental conditions, mobileassets and supply chains, as well as energy andwater meters. Typically these connected devicestransmit regular updates, such as a temperaturereading, or configurable event triggers.
Broadly, there are two kinds of LPWAtechnologies: Standardised technologies that usethe licensed spectrum belonging to telecomsoperators and proprietary technologies thatoperate in unlicensed spectrum.
Proprietary LPWA technologies inunlicensed spectrum
A number of LPWA technologies, such as Sigfoxand LoRa, have already been deployed to connectsensors and other devices to the Internet ofThings. As they use unlicensed spectrum andhaven’t gone through a standardisation process,these proprietary LPWA technologies have cometo the market quickly.
However, these technologies are vulnerable tointerference from other radio signals transmittedusing the same blocks of unlicensed spectrum.Moreover, they are only supported by a smallnumber of vendors and typically don’t supportroaming across international borders.
Cellular
Ran
ge
Battery Life
LONG
SHORT LONG
Local network(WiFi, ZigBee, Z-Wave)
Personal network(Bluetooth)
Low-Power Wide-Area Network(LPWAN: Sigfox, LoRa, Dash7)
Source: Alexander Vanwynsberghe, Blog article 'Long-range radios will change how the Internet of Things communicates'
Source: Alexander Vanwynsberghe, Blog article 'Long-range radios will change how the Internet of Things communicates'
IoT Connectivity: Range vs Power
What is LPWA
Low Power, Wide Area Networks
Low data throughput = High sensitivity = Long range
Relatively low cost
Multiple Access = One-to-Many Architecture
Using licensed or unlicensed spectrum
LPWA: Typicals
License-free Spectrum
EC-
-m
Licensed Spectrum
LPWA: Technologies
LPWAN - Comparisons
Sources: WAVIoT, NWave, Weightless SIG, and EE Journal
LPWA: Comparison
Disclaimer
Obviously, I can not go deep into each technologies
30 minutes won’t be enough
CostModules, deployment, operational cost
Usage Model / LicensingSIGFOX – Required to utilize their public network LoRa – Proprietary physical layer but open MAC
Regional RegulatoryAllowed frequency for ISM bandIn Europe, duty-cycle is 1% for end-devices
Upstream/Downstream SIGFOX – nearly entirely upstream LoRaWAN – has 3 classes supporting different balances of upstream & downstream
Hardware & Network Availability Is it available NOW?
LPWA: Selection Factors
LoRa?
For that… we’ll focus on…
Not this woman :)
Wireless modulation technology
Physical (PHY) layer for long range communications
Operates in the license-free ISM bands all around the world
• 433, 868, 915 Mhz
• Regulated (power, duty-cycle, bandwidth) E.g: EU: 0.1% or 1% per sub-band duty-cycle limitation (per hour)
Sensitivity: -142 dBm
Link budget (EU): 156 dB
What is LoRa
ISM Regulation ISM worldwide regulation 7
Output Power vs Duty Cycle
Countries Frequency band review Max. output power
EU 868 MHz 14 dBm
USA 915 MHz 20 dBm
Korea 900 MHz14 dBm
Japan 920 MHzMalaysia 862 to 875 MHz
20 dBm
Philippines 868 MHzVietnam 920 to 925 MHz
India 865 to 867 MHzSingapore 922 MHzThailand 920 to 925 MHz
Indonesia 922 MHzANZ 915 to 928 MHz
Taiwan 920 to 925 MHz
China 470 to 510 MHz 17 dBm
919 to 923MHz
Communications protocol & architecture utilizing the LoRa physical layer
Data rates are from 300bps to 5.5kbps Has 2 high-speed channels at 11kbps and 50kbps (using FSK modulation)
It supports • secure bi-directional communication, • mobility
What is LoRaWAN
LoRa/LoRaWAN Architecture
ASSET TRACKING
A real-world use case, that we did…
(Planned) thousand of assets to track
Deployed in country-side: no cellular coverage, hard to reach once deployed
Battery should last at least 3 years
Trigger alert if asset is in-move and track its movement
Requirement
Low power MCU: Microchip/Atmel ATSAMD21 Sufficient clock, flash, RAM, peripherals Interrupt: RTC, external -> useful for wake-up
Brain
SensorIMU sensor: gyroscope, accelerometer. Will wake up MCU upon significant movement Obviously, need GPS module to track location
BatteryLithium-thionyl chloride cells (Li-SoCl2) to reduce self-discharging rate 19Ah enough for 3+ years
Solution: Hardware-side
Most deployment areas are not covered by any cellular services
Cellular
Wi-FiNearby “civilization” is 5 km away. Not having clean LoS to use directional antenna Power consumption consideration
LPWACan be an option, but which one? Should be: Private networkEasy and low-cost enough to implement NOW
Options: Connectivity
3 - 5km LoRaWAN Gateway
NetworkServer
ApplicationServer
Cellular (3G/4G)
Backend
Tracker Node
Solution: Architecture
Makestro Cloud
On-site
System only wake up upon: timer interrupt and external interrupt (significant motion & displacement)
During sleeping:Turn off unneeded MCU peripheralsTurn off/make sleep GPS and radioKeep IMU sensor alive with the lowest update frequency
Only transmit data: By timer (depends on OTA configuration) Upon alert/interrupt
Circuit design optimisation: reduce components, pull-ups, etc
Low power technique
Tracker Node
GPS Module
Backup/RTC
Power
GPS Antenna
LoRa Antenna
Tracker Node
ATMEL SAMD21
IMU SensorLoRa
Module
Battery Sensor
Get One at: https://shop.makestro.com/product/tracker/
LoRaWAN Gateway
LoRaWAN Node
Portable: battery -powered, GPS for accurate timing
CLOSING
LPWA is NOW!
(arguably) a more accessible option, NOW!
How to start?
Cytron’s Arduino & Shields: GPS, LoRa,
LiPo battery
Use development board + modules!
Use development board + modules!
DycodeX’s ESPectro + LoRa + Alora Kit http://makestro.com/espectro
COMING SOON!
SIM868GPRS + GPS
Ultimate IoT connectivity board!
ESP32 MCU WIFi +
Bluetooth
SIM53603G + GPS
Options
Raspberry Pi-compatible pinout Compatible with many Raspberry Pi hats
Ultimate IoT connectivity board!
LoRa/LoRaWAN Module NFC
Be a maker!
A movement to “democratize” knowledge, hardware kit, and software to help makers to start making in hardware.
Disclosure: it’s initiated and supported by DycodeX, but it’s Community!
ENABLING MAKER MOVEMENT
Learning
Shop
Community
Software
Projects, tutorials, videos, news, professional trainings
Hardware marketplace: kits, devboards, maker tools
IoT Cloud infrastructure, software libraries, sample code
Offline meetups, seminars, training/hands-on, challenges
makestroid
makestroid
makestroid
How can we help?
for your next IoT endeavour
AndriCEO
HelmiCMO
Get in Touch
[email protected] | http://dycodex.com
IoT & maker movement enabler