1. 1
Infrastructure solutions
for large-scale public
and private IoT
networks
iSPHER’s groundbreaking technology enables
network operators to cost-efficiently deploy
scalable IoT networks while securely managing
any volume of data exchanged with the
connected objects.
August 2015
A document published by iSPHER and the Nicolas Bourbaki Center for
Research and Development.
The future at your fingertips.
Powering the Internet of Things
2. 2
SPHER NET | INTRODUCTION PAGE 04
SPHER NET | SOLUTION OVERVIEW PAGE 05
IOT | A RICH ENVIRONMENT DESTINED FOR GROWTH PAGE 06
IOT NETWORK | COMPONENTS OVERVIEW PAGE 07
IOT | CURRENT AND FUTURE CHALLENGES PAGE 08
SPHER NET | A VERSATILE BUSINESS MODEL PAGE 10
ABOUT ISPHER PAGE 11
APPENDIX I:
IOT - THE RELENTLESS ARRIVAL OF A SOCIOECONOMIC UPHEAVAL
PAGE 12
APPENDIX II:
WHAT DIFFERENTIATES AN INTERNET OF THINGS NETWORK
OPERATOR FROM A MOBILE NETWORK OPERATOR?
PAGE 17
APPENDIX III: GLOSSARY OF TERMS PAGE 21
APPENDIX IV: SPHER NET DEPLOYMENT REQUIREMENTS PAGE 22
CONTACT INFORMATION PAGE 26
CONTENTS
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
3. 3
A few years ago, three engineers got together to fulfill
a common vision:
Design superior technology to bring to the market advanced,
yet affordable business and communication solutions.
Now, the dream is becoming reality.
With a staff of over 30 scientific personnel that include PHDs, University Professors and Senior R&D
Engineers, the Nicolas Bourbaki Center for Research and Development, iSPHER’s Research Division, has
successfully developed groundbreaking technologies that include amongst other, an innovative, 100%
European operating system, infrastructure solutions to enable the Internet of Things to become an
affordable reality and the most advanced large surface touch screen technology in the world.
Everyday, our rapidly expanding R&D and engineering team contribute their talent, unique expertise and
passion for technology, to innovate and equip our clients with the advanced technology to provide them with
the competitive edge they need now and in the future.
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
4. 4i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
SPHER NET | INTRODUCTION
If you think that the internet has
changed your life, think again. IoT is
about to change it all over again!
Internet of Things and Low Throughput Networks
The Internet of things (IoT) refers to all technologies enabling
communications between objects commonly referred to as
“smart objects” and Cloud-hosted servers. The global estimated
connected object market based on current forecasts is set to
exceed 26 billion connected units by 2020. It is anticipated
that IoT will generate a global direct and/or derived income
running into several trillion USD per annum through the sale of
objects and services for a multitude of new and innovative
markets and usages such as metering, monitoring, security,
asset tracking, etc.
Low Throughput Networks (LTNs) are technologies that enable
long-range wireless data communication between smart objects
and gateways, with the advantage of requiring only very low
power consumption, meaning an object powered with standard
AAA batteries can communicate with the designated application
or server during several years. LTNs can be applied to building
star-topology networks connecting smart objects that need to
send and/or receive a small number of data packets per day.
There are multiple radio technologies underlying LTNs: Sigfox,
LoRa, Neul (Weightless), etc.
The LoRa Revolution
Contrary to other radio communication systems, LoRa is a
standardized protocol designed by the SemTech Corporation,
which produces LoRa radio chipsets, yet does not operate a
network. As a result, third-party companies are free to deploy
their own public or private network on any scale.
Furthermore, LoRa allows for bi-directional communication, thus
making connected objects truly smart, as well as being extremely
cost-effective in terms of sensors, deployment and
communication costs.
The proliferation of this protocol is being supported by a the
LoRa Alliance consortium that was founded in 2015 to promote
LoRa radio technology. This consortium has currently some 50
industrial members, including equipment manufacturers,
telecommunication operators, system integrators, or sensor
manufacturers. The LoRa Alliance recently published the
LoRaWAN specification defining a standard and secure protocol
to deploy LoRa-based networks.
LoRa’s implication for IoT network deployments is considerable
because what was until now a vision of the future, is fast
becoming an affordable reality, which in turn presents the
industry with one key problem:
How can network managers simultaneously handle the
information being fed back from billions of connected objects
hundreds of times per day and enable users to make sense of all
this information?
What makes SPHER NET unique?
Our R&D team foresaw that IoT’s ubiquitous character would
present network operators with some major challenges:
communication protocols, cost, data volume, data integrity and
scalability among other.
Addressing these challenges guided SPHER NET’s development
in order to enable us to provide a robust solution to mobile
operators or any private network operator, such as an airport.
Early on, we took a bet on LoRa due to its low operating cost as
well as its “standardized” profile, making it easily adoptable.
However, SPHER NET is flexible and designed to integrate any
number of existing protocols, as well as those yet to be invented.
It is extensible. More than 90% of our developments are not
specifically aimed at the LoRa protocol and can be utilized in the
context of another IoT radio protocol.
Being an infrastructure software solution especially designed for
mobile operators that need to deploy public LTNs to generate
new revenues, SPHER NET’s development focus was on large
data volume management and reliability. As such we address
efficiency, security and high availability, regardless of data
volume, thus providing a highly reliable and scalable system.
SPHER NET is operational and can already integrate mobile
operator ecosystems enabling them to capitalize on the IoT in
addition to providing them with a platform primed for Big Data
analysis.
The strength of our technology has also resulted in a highly
versatile system capable of integrating private LTNs, thus
substantially increasing configuration types and market reach.
We are currently rolling out pilot SPHER NET LTNs for large car
park operators.
Today, under one percent of the 1.5
Trillion connectable objects are
actually connected. The scope for
growth is almost endless.
As the most advanced IoT network management system
available today, SPHER NET will seamlessly integrate a variety
of sectors that include:
Environment: Smart Cities, Environmental Management, Water
Management, Electricity, Gas, etc.
Security: Homeland Security and Border Control, Emergency
Response Services, etc.
Sales & Retail: Sales Monitoring and Management, Stock
Management, etc.
Industry and Agriculture: Smart Logistics, Infrastructure
Management, Field monitoring, etc.
Just-in-Time Supply Management: Logistics, Stock
Management and Monitoring, etc.
Other: Healthcare, Home Appliances, Domestic Monitoring,
Smart-Houses, etc.
5. 5
SPHER NET | SOLUTION OVERVIEW
CUSTOMER APPLICATIONS
NETWORK PROVIDERS
GATEWAYS
CONNECTED OBJECTS
Software solutions for
private Network Operators
SHPER NET’s versatility allows
it to perform equally in private
or public LTN configurations
Bespoke Applications
In private network
configuration, we provide
the Gateway using
protocols such as
LoRaWAN, WMBus, etc.
Although we have the capability to develop
client applications, SPHER NET is a network
management software designed to be
integrated in the solutions of applications
specialists such as SAP for instance.
Software solutions for public Network
Operators
SPHER NET is a robust and versatile LTN and
infrastructure management system designed
to integrate the infrastructures of large mobile
operators such as Orange, Bouygues, BT,
AT&T, etc. We enable such operators to offer
large-scale connected objects networks to
their customers.
SPHER NET’s ability to handle any radio
communication protocol means that it
integrate any radio protocol into a large-scale
network infrastructure for the likes of Cisco,
Kerlink,, Link Labs, etc. SPHER NET can
easily be adapted to handle future IoT radio
protocols.
SPHER NET’s ability to handle millions of
simultaneous data feeds allows sensor
manufacturers such as Adeunis, IMST,
Libellium, Microchip, to widen their global
client offer.
SCALABILITY
SPHER NET simultaneous processes millions of data feeds
from connected objects in the field.
HIGH AVAILABILITY
SPHER NET guarantees 99.999% availability.
SECURITY
SPHER NET encrypts all communications to ensure the
highest level of data security.
MULTIPROTOCOL
SPHER-NET can handle multiple communication protocols
and technologies.
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
SPHER NET | THE WORLD’S MOST ADVANCED SOLUTION TO DEPLOY LARGE-SCALE IOT NETWORKS
In private network
configuration, we ensure
availability of suitable
sensors to connect objects
to the Gateway.
6. 6
Wearables
Connected
Cars
Connected
Homes
Connected
Cities
Industrial
Internet
Transportation
Oil & Gas
Healthcare
IOT | A RICH ENVIRONMENT DESTINED FOR GROWTH
Smart Cities: Intelligent parking, monitoring of
buildings, sound monitoring, personal detection,
traffic management, public lighting management,
household waste management, public information
display, etc.
Smart Environment: Fire detection, air pollution
monitoring, snow level monitoring, avalanche
prevention, floods and drought management,
earthquake warnings, etc.
Smart Water: Monitoring of drinking water quality,
chemical contamination detection, tracking pools,
pollution levels of the seas and oceans, leak
detection, flood monitoring, etc.
Smart metering/Smart Grid: Electricity, gas and
water smart metering, monitoring of photovoltaic
systems, water flow monitoring, calculation of stock in
silos, etc.
Tracking: vehicles, bicycles, valuables, animals,
people, etc.
Security and rescue: Hazardous area analysis,
detection of hazardous liquids, radiation levels,
explosive substances detection, etc.
Trade: Supply chain control, NFC payments, smart
shopping, ray product radiation, etc.
Logistics: Transport conditions monitoring, parcel
tracking, storage mismatch detection, fleet tracking,
etc.
Industrial control: Monitoring of machines,
equipment status, quality of indoor air, temperature
monitoring, detection of ozone level, equipment and
indoor products localization, remote vehicles
diagnostic, etc.
Smart Agriculture: Monitoring of vines, greenhouses
monitoring, golf courses irrigation control, weather
stations, compost, animal tracking, etc.
Smart Animal Farms: Grazing tracking, monitoring
of toxic gas levels, animal development monitoring,
hydroponic installation monitoring, etc.
Home Automation and Smart Buildings: Use of
water and electricity, remote control, intrusion
detection, smoke detection, monitoring of valuable
assets, etc.
e-Health: Fall detection, drug storage, sports
monitoring, patient monitoring, ultraviolet radiation,
etc.
IoT will be used in the following sectors …
26Billion connected
objects by 2020
1%of 1.5 T connected
object global
potential actually
connected today
25Billion US Dollar
industry by 2020
14Trillion US Dollar
estimated derived
IoT market
47Thousand jobs
created around IoT
in Barcelona alone
over the past 7 years
7. 7
LTN stands for Low Throughput Networks, which are technologies that enable long-
range wireless data communication (up to 15km) between smart objects and
gateways, requiring only very low power consumption. There are multiple radio
technologies underlying LTNs: Sigfox, LoRa, Neul (weightless), etc.
Connected objects
Connected Smart Objects embed a radio chipset, a micro-controller, and a set of
sensors. The micro-controller executes a software that periodically sends data packets
containing sensor values, such as temperature, GPS location, power consumption and
more. Various radio chipsets are available: LoRa, Sigfox, Neul (Weightless), etc.
The current estimation is for 25 Billion connected objects by 2020, each
transmitting data multiple times every 24 hours, 365 days a year.
Gateways
They embed a multi-channel radio chipset (aka concentrator) and a tiny computer. The
latter executes a software that forwards packets received from smart objects to the
datacenters. Gateways also take in charge communications from the datacenters to
smart objects.
There will be as many gateways as there will be networks, multiplied by the
number of antennas nneded to provide adequate coverage.
Datacenters
They are composed of a set of interconnected computers and servers. Datacenters
have three main functions:
• Process the data packets forwarded by Gateways.
• Implement management functions for Gateways and Connected Objects.
• Implement administrative functions for Customer Relationship Management,
such as smart object registration, accounting, billing, etc.
Globally, there will likely be several hundred mobile operators operating
public LTNs in addition to 10’000s private LTNs.
Applications
Receive data packets forwarded by datacenters and implement high-level functions
for customers, such as reporting values of smart object sensors, triggering actions
based on received packets, etc.
There will be 100’000s of applications utilized by billions of users.
IoT NETWORK
This infrastrure layer allows for the exploitation of the data collected from any number
of sensors by private or public LTNs, thus enabling individuals and companies to
incorporate the Internet of Things into their ecosystems.
DATA MANAGEMENT AND ANALYSIS
IOT NETWORK | COMPONENTS OVERVIEW
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
GATEWAY
LTN Access-Point. A LAP is a radio base station that receives messages sent by LEPs
and forward them to LTN servers (using Ethernet, 3G modems, etc.).
OBJECT
LTN End-Point. This is an object with LTN modem running the LTN radio protocol (i.e.
LoRaWAN).
8. 8
What problems does SPHER NET solve?
Now that IoT networks are affordable and can be deployed in a
cost-effective manner, they provide mobile operators and large
verticals unprecedented opportunities for the generation of new
revenues either through the provision of new services to
customers or by optimizing processes, thus achieving bottom-line
savings.
The key problems facing private and public LTN managers are
the processing of large data volumes, system reliability,
scalability, data integrity and security as well as sustainability.
Network managers simultaneously need to handle vast amounts
of information being fed back from millions of connected objects
per operator, hundreds of times per day while enabling users to
leverage this information.
This sector will enjoy sustained multi-year growth because
processing the anticipated volume of data generated by
thousands of private and public networks and billions of
connected objects will indeed become a very big business. The
infrastructure needed to simultaneously process such volumes
will generate revenues from bandwidth provision, data
processing, and will require lots of hardware to operate.
SPHER NET is a highly available and scalable software stack
for network operators
We developed a fully operational LoRaWAN-compatible software
stack targeting network operators. In order to ensure that SPHER
NET possesses a solution profile best adapted to their needs:
• We focus on implementing the software for gateways
and datacenters.
• We do not develop smart object code.
Our solution is fully LoRaWAN-compatible and can be used with
any smart object implementing LoRaWAN specification. We have
successfully tested SPHER NET with objects developed by three
separate manufacturers, and as a matter of course are
expanding the compatibility program in order to meet any variety
of LTN configurations.
Our solution targets large-scale networks with stringent
requirements in terms of performance, availability, security, and
scalability. SPHER NET fulfills this commitment to clients by
guaranteeing the following product characteristics:
High performance: SPHER NET sustains a very high throughput
(number of messages handled in any given time unit) and critical
network operations are processed using a very low response
time path. This is achieved via a set of techniques, including:
dynamic data sharding, differentiation of service, custom intra-
datacenter network communication protocols, detection of
service-level agreement violation, etc.
High-availability: SPHER NET is designed to ensure 99,999%
availability (aka “five nines of availability”). It does not exhibit any
single point of failure. Moreover, it can survive an entire
datacenter outage. This is achieved via a set of techniques,
including: state-machine replication, automatic fail-over, fault-
detection, rolling upgrades, custom geo-replication protocols, etc.
Security: SPHER NET provides end-to-end security guarantees.
This is achieved through encrypted communications, network
traffic separation, detection of denial of service attacks, intrusion
detection, etc.
High scalability: SPHER NET is able to dynamically adapt to an
increasing load and limits the use of over-provisioning to the strict
minimum. This is achieved through server consolidation, micro
service deployment, continuous monitoring, self-sizing, load
prediction algorithms, etc.
Finally, SPHER NET is designed to be extensible. More than 90%
of our developments are not specifically geared towards the
LoRa protocol and can be reused in the context of other IoT radio
protocols. To illustrate this, we are currently implementing the
support for another protocol, called WMBUS (Wireless M-Bus).
Embedding economic empowerment into our development philosophy
Our overriding objective is to achieve technological excellence in order to develop versatile solutions that not only withstand the test of
time, but also can be adapted to many different technical and/or business configurations, now and tomorrow. This is especially relevant
when developing for the Internet of Things sector which in many ways is in its infancy and is taking shape daily.
Consequently, we have developed SPHER NET around what we know to be the most likely business and implementation models
based on the feedback we obtain from industry, IoT network components manufacturers, applications development companies, mobile
operators and likely private network operators.
In parallel to its technology, SPHER NET empowers mobile operators to offer ahead of time a rich IoT environment and service platform
to their customers, capable of generating new revenues and adapting to future commercial needs.
SPHER NET extends this rich environment to key actors such as government agencies, local authorities and a host of private sectors.
In many instances, SPHER NET can also stand as the processing backbone for IoT network component manufacturers, who can
incorporate the data processing and network management into their overall product offering to access markets that were previously
inaccessible due to the lack of a turnkey solution.
IOT| CURRENT AND FUTURE CHALLENGES
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
9. 9
GE estimates that
convergence of
machines, data and
analytics will
become a $200
billion global
industry over the
next three years.
10. 10
SPHER NET | A VERSATILE BUSINESS MODEL
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
Configuration 1
SPHER NET provides the data
processing backbone of Public
LTNs for Mobile Operators
In this configuration, iSPHER works in partnership with the
mobile operator providing SPHER NET as the data
processing backbone to create the LTN layer needed for a
full connected object network deployment.
iSPHER services include system customization, 24/7
support as well as ongoing software upgrades.
The mobile operator will be in a position to offer a full IoT
offer to his customers, who pay for this service on the basis
of billing per connected object per month, or data transfer,
or a combination of both depending on the operator’s
commercial approach.
In any event, iSPHER will aim for a minority participation in
the revenues generated by the operator while operating his
LTN with SPHER NET technology.
Configuration 2
SPHER NET provides the
processing backbone of Private
LTNs within a consortium of
service providers
iSPHER works in partnership with manufacturers of
hardware used for operating IoT networks and will provide
SPHER NET as the data processing backbone to create the
LTN layer needed for a full connected object network
deployment.
iSPHER services include system customization, 24/7
support as well as ongoing software upgrades.
The consortium will be in a position to offer a full IoT offer to
customers, who pay for this service on the basis of billing
per connected object per month.
iSPHER will aim to benefit from an equal share of the
revenues generated by the consortium.
Configuration 3
SPHER NET provides the data
processing backbone for LTNs as
a white label solution
Configuration 4
SPHER NET acts as a Private
LTN operator
As a white label solution, iSPHER works in partnership with
the operator, who can be a business application specialist
for instance, providing SPHER NET as the data processing
backbone to create the LTN layer needed for a full
connected object network deployment.
iSPHER services include system customization, 24/7
support as well as ongoing software upgrades.
iSPHER will license SPHER NET and where possible, aim
for a minority participation in the revenues generated by the
by operating LTNs with SPHER NET technology.
In this configuration, iSPHER creates a consortium
constituted by regional companies specializing in IoT
network components and provides the data processing
backbone to create the LTN layer needed for a full
connected object network deployment.
iSPHER services include system customization, 24/7
support as well as ongoing software upgrades.
iSPHER will aim for a majority participation in the revenues
generated by the consortium since the firm will often take
the business development lead.
11. 11
We have the ability to integrate
market needs and reality into our
development programs. Our
capacity to understand future trends
have yielded groundbreaking
technologies that we are now
bringing to market.
iSPHER excels at creating, developing and implementing
sophisticated technology solutions that enhance Man’s
ability to interact and connect with the surrounding
environment.
We are dedicated to the development, promotion and sale of
enhanced user-friendly communication solutions and tools
characterized by innovation, as well as by a deep understanding
of the markets we serve. Our experienced team along with the
outstanding researchers and scientists of the Nicolas Bourbaki
Research and Development Center apply their unique talents to
foresee today the market needs of tomorrow.
Our areas of expertise are communication and connectivity
solutions where our state-of-the art technology developments
provide our clients with a competitive edge in their respective
markets. Our technology and solutions integrate seamlessly into
a wide variety of leading sectors such as Media &
Entertainment, Social Media, Medical, Education, as well as in a
host of industrial areas. Our user-friendly applications include
tactile technologies for large-scale surfaces, end-to-end
communication gateway to power the Internet of Things, novel
graphic interfaces, etc.
What differentiates us is our ability to integrate market needs
and reality into our development programs. This capacity to
understand future trends have led us to undertake important
long-term developments that have yielded groundbreaking
technologies that include a proprietary operating system
enabling simultaneous multi-user multi-application use of large-
scale tactile surfaces, state-of-the-art hardware and software
developments to enhance user experience and lower unit
energy consumption, sophisticated content distribution systems,
the first operational end-to-end LoRa network solution, and
more.
Our strength is our ability to integrate
key business skills, resources and
unique technological knowledge, thus
allowing us to fulfill our commercial
vision and achieve our stated goals.
The Nicolas Bourbaki R&D Team in brief
All our R&D team members have been handpicked from a pool
of hundreds of top-level research and engineering candidates
from leading French universites and laboratories.
Our full-time R&D staff include:
• 7 PhDs specialized in advanced software development
• 3 Professors specialized in big data and distributed
systems.
• 17 research engineers
• 2 Technicians
Unique scientific knowledge needed to create today
the solutions of tomorrow
Our investment into Innovative R&D drives iSPHER's
commercial approach and strategy. The key to our success is
our in-house capability to generate genuine innovation and
develop solutions that do not draw on any fully developed or
existing technology.
We are dedicated to becoming market leaders by
sharing our knowledge and fostering a development
community.
We are committed to obtaining market leadership in hardware
and components design, software development and content
management systems, allowing us to durably serve commerce
and industry by delivering superior communication solutions that
are always adapted to human needs.
ABOUT ISPHER
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
12. 12i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
IoTThe relentless arrival of a
socioeconomic upheaval
A high-stake challenge for all
economic sectors.
August 2015
The future at your fingertips.
13. 13i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
IoT: the relentless arrival of a
socioeconomic upheaval
What’s at Stake?
The Internet of Things, generally known as IoT, its accronym, refers to all technologies allowing
the communication between connected objects, commonly called "smart objects" and servers
hosted in the Cloud.
These objects are and will be used in a numerous sectors:
These objects are already, and will increasingly be used in a multitude of sectors and
applications:
Smart City: intelligent parking, dynamic monitoring of the status and condition of buildings,
dynamic mapping of noise levels, dynamic measurement of electromagnetic energy levels,
dynamic management of road and pedestrian traffic, management of street lighting, garbage
collection management, etc.
Intelligent Environmental Management: detection of forest fires, pollution level control,
monitoring snowfall levels, detection and prevention of landslides, detection and prevention of
avalanches, monitoring of sysmic activity and earthquake warning, volcanic activity monitoring,
etc.
Intelligent Water Management: quality control of water, detection of chemical leaks into rivers,
control of the quality of pool water, control the level of pollution of the seas and oceans,
detecting leaks water around the tanks and pipes, level control of rivers and canals, etc.
Intelligent Management of Remote Measurements: energy consumption monitoring and
management of public or private users, monitoring of water tanks, gas filling level or gas
monitoring, photovoltaic facility monitoring, monitoring of water pressure levels in the water
supply system, etc.
Safety and Emergencies Management: access control for restricted access areas, people
detection in sensitive and prohibited areas, corrosion prevention in sensitive buildings, radiation
level control at nuclear power plants and in surrounding areas, detection of gas leaks in
chemical plants and surrounding areas, etc.
The appearance of the LoRa
protocol that enables the mass
deployment of connected
objects, coincides with
favorable estimates that all
agree on an annual turnover of
several trillion US dollars
generated from 2020 onwards
by this new market.
14. 14i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
Intelligent Sales Management: monitoring of storage conditions, storage management for the
supply of products, product tracking to ensure traceability, expiry date control, detection of
allergenic components, end-user smart shopping applications, etc.
Intelligent Logistics: control of the conditions of transported goods - sensing vibrations,
detection of a break in the refridgeration chain -, locating objects in large warehouses,
detection of storage mismatch between products containing flammable and explosive
substances, monitoring and control of routes used for the transport of large or dangerous
objects, etc.
Intelligent Industrial Control: self-diagnosis and preventive maintenance of industrial machinery,
control of the workers working conditions, temperature control within industrial refrigerators,
ozone level control in food plants, automatic location of goods within of buildings, etc.
Smart Agriculture: sugar level control in wine, state of vineyards control, monitoring of micro-
climatic conditions to maximize quantity and quality of fruit and vegetable production, smart
irrigation management through the automtic detectinon of arid surfaces, remote monitoring of
weather conditions, quality control, etc.
Intelligent Farms: animal tracking, air quality control, presence of toxic gas detection emenating
from excrements, etc.
Intelligent Health: remote monitoring of the elderly, vaccines control and grafts storage
conditions, control of vital indicators for large public sports centers, etc.
Since the appearance of the LoRa protocol, that enables the massive deployment of connected
objects networks, all estimates point towards an annual turnover of several trillion US dollars by
2020 generated by this new market. The stakes are so important that we are witnessing today
in an unprecedented media battle between telecom operators in support of a substantial
fundraising effort.
Public Relations vs. Technological Reality
There is not a day that goes by without an operator announcing the deployment of a network
dedicated to IoT and connected objects. What is the really behind these announcements?
To answer this question we need to know if Mobile Network Operators (MNO) are
technologically equipped to face the oncoming challenge of the Internet of Things. As of today,
the answer is clearly "no" (see Annex 2 explaining the main reasons for this lack of
preparedness).
Knowing this, we can assume that they have opted for this strategy due to these key
commercial drivers:
1. The Media Battle
As mentioned above, MNOs do not hesitate to announce the deployment of IoT networks with
the full knowledge they are following a communication strategy that is clearly desiged to mask
infrastucture deficiencies. The truth is that in their current state, MNOs and mobile telephony
have very little, even nothing to do with the communication between connected objects, data
centers and user applications. The reality is the only technological advantage MNOs possess
today for deployment of connected objects networks is their park of masts supporting the
antennas used for mobile telephony.
2. Equity Investment into Startups
Equity investment in the startups that could help them develop a viable solution for becoming
an IoT operator are often considered by MNOs, wrongly, as a proof of their unpreparedness
and the ensuing pressure that this new unknown market is causing them.
The difficulty of implementing an IoT solution is not in an antenna design study, whose cost is
far inferior to 150’000 required for their antenna masts, or that of the connectivity of the objects,
but rather in the development of the IT infrastructure needed for large-scale data processing
that only specialists in distributed computer systems and technologies are able to implement.
The expertise needed for this task is very different to the skillsets that are not possessed by RF
engineers (antenna developers and connectivity for objects) who simply do not have the
competence to design or implement this type of infrastructure.
The stakes related to the
market for connected objects
are so important that we are
currently witnessing an
unprecedented media battle
between Mobile Network
Operators in view of raising
substantial investment funds.
15. 15i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
The Nicolas Bourbaki Research & Development Center at the request of iSPHER, in order to
complement its IoT network data processing infrastucture solution, launched the study for an
antenna possessing more powerful capabilities than those existing now. Initial studies are
proving highly promising, and should we not encounter any substantial technolgical barrier,
iSPHER will be able to produce its own antenna from January 2016 onwards, which will provide
us with the added benefit of emancipating ourselves from a limited number of antenna
manufacturers commanding high-end pricing.
The Nicolas Bourbaki R&D Center developed for iSPHER, the SPHER NET IoT data
processing solution, a end-to-end solution that includes the software for antennas and what is
without a doubt, the single hardest challenge, which is a reliable solution for application servers
processing vast amounts of data to and from connected objects.
The networks deployed and operated using our SPHER NET solutions enable secure bi-
directional communication between connected objects and applications hosted in the Cloud. In
addition, Spher NET provides its customers with tools to:
• Monitor their fleet of connected objects
• Implement complex data processing workflows exploiting the data data packets
received from their connected objects
To ensure the best quality of service, SPHER NET has developed a reliable and effective
antenna park supervision and management system that are linked to the servers constituting
the data processing infrastructure.
SPHER NET solutions for operators of IoT Networks are based on three system components:
SPHER NET Gateway: this component is deployed on the masts. It supports bi-directional
communication between the connected objects and servers deployed in the cloud.
SPHER NET Operator: this component is run on servers hosted in either SPHER NET’s data
center and/or at the operator’s premises. It is responsible for the supervision and administration
of the deployed antenna park.
SPHER NET Interface: is a component that runs on servers that can be hosted either in
SPHER NET data centers, or in the operator’s own server farm. This component provides the
framework needed for the creation of applications using the data produced by the objects.
Example: storage, analysis, implementation of complex processing workflows, etc.
To ensure the highest service
quality, SPHER NET has
developed a reliable and
effective system for the
supervision and administration
of antenna parks, in addition to
server farms that provide the
data processing infrastructure.
16. 16i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
Tamedia
iSPHER together with Tamedia, the leading publishing group in Switzerland founded in 1893
and listed on the SMI, the Swiss Stock Exchange, since 2000, are currently studying the
deployment of a network of connected objects using SPHER NET, our IoT Network solution, in
the city of Lausanne.
This network will be used by Tamedia to track the physical sales of newspapers published by
the Group that are sold through self-service Points of Sale (POS) boxes in public places
(streets, public transport stations, etc.). The objective of the network deployment is to monitor
sales and optimize the supply of newspaper to the POSs. The intended outcome of the SPHER
NET deployment will be:
• A reduction in the number of unsold newspapers in the POSs,
• Early-detection of product shortage at the POS.
When fully deployed, each Tamedia self-service POS will be equipped with a connected object
approximately the size of a cigarette packet. This object will carry an ultrasonic sensor used for
measuring the height of the stack of newspapers in POS’s tray. The measurements are sent,
via antennas deployed on high points throughout the city of Lausanne, to servers hosted in
SPHER NET data centers. Consequently, Tamedia managers have real-time access to the
collected data via their user-friendly interface with a number of key indicators: instantaneous
percentage of newspapers present in each tray, supply history for each POS, etc.
Each Tamedia distribution tray
will eventually be equipped with
a connected sensor the size of
a cigarette packet.
This object will carry an
ultrasonic sensor for measuring
the height of the stack of
newspapers present in the tray,
thus providing real-time
supply status.
17. 17i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
What differentiates an
Internet of Things
network operator from
a mobile network
operator?
A brief overview of the key differences between
IoT Network Operators and Mobile Network
Operators impacting equipment, network
architecture and data treatment.
August 2015
A document published by iSPHER and the Nicolas Bourbaki Center for
Research and Development.
The future at your fingertips.
18. 18i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
Mobile Network Operators vs.
IoT Network Operators
Is the Mobile Network Operator equipped to face the
challenges of the Internet of Things (IoT)?
The answer is no for the following fundamental reasons:
1. The Mobile Network Operator (MNO) has a radically different
hardware infrastructure.
For the MNO, the major challenge is to enable two people to communicate (voice or
SMS). Consequently, his infratructure takes the form of a strongly meshed network (a
kind of "peer-to-peer" network) that establishes a communication channel between
two people regardless of their location (the diagram below illustrates how the mesh
network allows two devices to communicate via mobile broadband).
An IoT network requires a radically different physical infrastructure. In fact, the role of
an IoT infrastructure is not to allow "objects" to communicate with each other, but to
transmit data from the objects to one or more point(s) of treatment. Therefore, an IoT
network infrastructure takes the form of trees whose root is a set of data centers
(commonly known as "The Cloud") and whose leaves are "objects" (see figure below:
the flow of data goes from the "end nodes" to the "application servers").
The MNO infratructure takes
the form of a strongly meshed
network that establishes a
communication channel
between two people regardless
of their location.
The IoT Network is radically
different. Its role is not to allow
"objects" to communicate with
each other, but to transmit data
from the objects to one or more
point(s) of treatment.
Telecoms Network Architecture
19. 19i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
2. A radically different treatment of data
The data processing carried out by the MNO is radically different to the data
processing that must be accomplished by the IoT Network Operator:
An MNO’s main purpose and role is to transfer data from one point (a mobile phone)
to a second point (another mobile phone) and is not required to do any further
processing on transferred voice data. For message communications (SMS), the
operator provides message storage for a limited period. The data processing
performed by an MNO is relatively basic: it is mainly storage.
An IoT Network Operator must perform complex data processing of the data packets
generated by the connected objects. Indeed, the purpose of an IoT network is to
extract relevant information from the masses of data received from the objects. Here
are some examples of the type of data processed: data aggregation, pattern
detection, data evolution prediction such as prevent landslides through extrapolations
on data reported by motion sensors, etc. Successfully accomplishing this type of data
processing is notoriously complex and calls for the installation of a large-scale
hardware and software infrastructure. As such, the data processing carried out by an
IoT network operator are closer to the type of data treatment made by major Web
operators on data generated by users (i.e. Google, Facebook, Twitter, etc.) than those
made by MNOs.
3. A radically different network sizes
The amount and scale of equipment as well as the massive data volume managed is
radically different for the MNO than for the IoT Network Operator.
In terms of scale of equipment, a MNO runs one mobile phone per customer. For
example, leading French MNOs service between 10 and 25 million customers. An IoT
Network Operator will have to handle a vastly higher equipment volume (i.e. the
objects). This is due to the fact that at term, each of us will have many connected
objects (for his car, his bike, his garden, his house, etc.). Furthermore, hundreds of
millions, even billions of objects will be used in almost all sectors: smart cities, smart
farming, smart industry, etc.
A MNO’s main purpose and
role is to transfer data from one
mobile phone to a second
mobile phone and is not
required to do any further
processing.
An IoT Network Operator must
perform complex data
processing of the data packets
generated by the connected
objectsin the IoT network.
20. 20i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
An IoT Network Operator in a country like France should therefore will be required to
process several hundred million objects at least.
Regarding the volume of data processed, the "objects" in an IoT network will generate
far more data than mobile phones managed by the MNO. Indeed, the data generated
by mobile phones are the direct result of user activity (i.e. sending an SMS).
Conversely, most of the objects belonging to an IoT network will produce continuous
data (i.e. periodic activity resulting in data packets every ten seconds), which will
generate a much higher data volume to be processed.
4. Closed Network vs open network
A MNO is "closed": only the MNO uses the data generated by mobile phones.
Conversely, an IoT network is "open": a wide variety of customers will use the data
produced by the connected objects. For example, an IoT Network Operator serves
both:
A parking management company wishing to deploy sensors in its car parks and
process related data (such as displaying the number of available parking places,
usage statistics, occupancy prediction, etc.).
An energy supplier wishing to deploy telemetry sensors (eg gas meter reading) onsite
with their customer to offer them consumption statistics, predict the amount of future
bills, or formulate consumption-based advice based on analysis of historical data.
A city wishing to deploy sensors to monitor air quality or the operational state of public
lighting and providing visualization tools to municipal employees to assist them in
their monitoring activities.
The infrastructure of an IoT network must allow a wide range of customers the access
to data and exploit this data. Moreover, the integration of client applications must be
made in the very heart of the network, to enable applications to interact in real time
with the connected objects (eg to send commands to an object acting as actuator).
Bibliography
Some articles on the link between telecom and IoT:
http://telecoms.com/163802/mnos-will-not-make-the-iot-connection/
http://www.analysysmason.com/About-Us/News/Newsletter/IoT-M2M-operator-strengths-
Jan2015/
http://www.analysysmason.com/About-Us/News/Insight/M2M-operator-strengths-Nov2014/
A MNO is closed with only the
MNO using the data generated
by mobile phones.
Conversely, an IoT network is
open with a wide variety of
customers using the data
produced by the connected
objects.
21. 21i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
Glossary of Terms
3G Third Generation
3GPP(2) Third Generation Partnership Project (2)
API Application Programming Interface
AS Application Server
BGCF Breakout Gateway Control Function
CAMEL Customized Applications for Mobile Enhanced Logic
CSCF Call Session Control Function
BICN Bearer Independent Core Network
GGSN Gateway GPRS Support Node
GPRS General Packet Radio Service
GSM Global System for Mobility
IMS IP Multimedia Subsystem
IP Internet Protocol
IP-CAN IP- Connectivity Access Network
ISIM IMS Subscriber Identity Module
IWF Inter-Working Function
LAN Local Area Network
QoS Quality of Service
RACS Resource and Admission Control Subsystem
RAN Radio Access Network
SCM Session Control Manager
SGSN Serving GPRS Support Node
SIP Session Initiation Protocol
SLA Service Level Agreement
UMTS Universal Mobile Terrestrial Access
23. 23
Deploying a SPHER NET networks requires
the deployment of two distinct entities:
Base Stations
(also known as Gateways or antennas)
Datacenters
Hereunder the requirements for these two types of entities.
Base Stations
How many base stations are
required?
The number of base stations needed to provide a fully
operational network is roughly the same as the number of
GSM base stations required by a mobile phone operator.
Approximative coverage: In dense urban areas, the coverage of a Base station is
between 500m a 2km.
In rural areas, the coverage of a Base station is about 10km
Where should base stations
be installed?
Base stations should be installed in the highest possible
points. The ideal location is the top of dedicated pylons as
is commonly done for GSM antennas in rural areas (see
figure below).
When installing a pylon is not feasible, Base stations can
be deployed on the roof of buildings (see figure below).
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
SPHER NET Deployment Requirements
24. 24
How big is a base station?
LoRa Base stations are rather
small. They are composed of
two parts:
The Base station itself, which has the following dimesions:
• ~ 30cm x 10cm x 10cm, weight is about 2kg
(see figure below)
• An antenna, which ranges from 30cm to
180cm (on the picture above, a small antenna
is used; on the picture below a bigger antenna
is shown: 165cm).
Which facilities are required
for a Base station?
Base stations should have an access to Internet. The
available bandwidth on the link does not need to be high.
Typically, 1Mb/s is enough (10Mbit/s is preferred).
• There should be a power source (220V AC).
• The gateway must be protected against lightning.
Datacenters
Several datacenters need to be installed in each country in which SPHER NET is deployed.
In this section we describe the requirements regarding these datacenters.
What is a datacenter? A datacenter (see figure below) is a facility used to house
computer systems and associated systems (e.g. storage
systems, network switches). A datacenter provides:
• Uninterruptible Power Supply (UPS)
• Environmental controls (air conditionning, fire
suppression)
• Security devices
• High-speed Internet access
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
25. 25
Is it necessary to build dedicated datacenters?
It is not necessary to build dedicated datacenters for SPHER NET. In order to decrease
operational costs, datacenters are usually shared by several companies, including the largest
ones in the world. In each country, several companies are specialized in building datacenters
and renting “white spaces” in these datacenters. Examples of such companies include:
http://www.telecitygroup.fr/
http://www.tasfrance.com/datacenter/
http://www.green.ch/de-ch/home.aspx
Which hardware should be deployed in each datacenter?
SPHER NET requires deploying about 40 computers in
each datacenter. Depending on its role in the Sphernet
network, each computer comprises:
• from 8 to 48 cores
• from 32GB to 1TB of RAM
• from 1 to 10 network cards.
Furthermore, SPHER NET requires deploying a set of
storage and network devices (e.g. infiniband switches).
The hardware required for SPHER NET in each datacenter
fits in a so-called “cage” (see figure on the left).
How
many
datacenters
should
be
used?
SPHER NET implements geo-replication mechanisms for
two purposes:
• Ensuring high availability despite datacenter outages
• Ensuring an as-low-as-possible latency between
smart objects and applications hosted in datacenters
Therefore, in each country in which SPHER NET is
deployed, at least two datacenters must be used. In large
countries (e.g. USA, Russia), SPHER NET requires
deploying more than two datacenters.
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S
26. 26
Contact Information
For any further information regarding SPHER NET, please
contact by email info@ispher.com.
For sales enquiries, please contact the iSPHER
Representative Office at +41 800 99 88 33.
iSPHER Headquarters
Avenue de Tivoli 3
1700 Fribourg
Switzerland
iSPHER Representative Office
Chemin des Aulx 21
1228 Plan-les-Ouates
Geneva, Switzerland
T. +41 22 741 00 40
Nicolas Bourbaki R&D Center
294 route des grands bois
74370 Villaz
France
T. +33 4 50 60 60 90
E. info@ispher.com
www.ispher.com
i S P H E R | T H E F U T U R E A T Y O U R F I N G E R T I P S