the internet of important things - smart grid: an intelligent test case
DESCRIPTION
Helios Adviser white paper Author: Richard Womersley [email protected] _______________________________________________________________________ Follow Helios via Linkedin, www.twitter.com/askhelios and www.facebook.com/askheliosTRANSCRIPT
Analys is & commentary for
decis ion makers in the telecoms industry
The Internet of
Important Things
Smart grid: an intelligent test case
HELIO
SADVISER
About Helios
Helios is an independent consultancy providing business, regulatory and technical
advice to the ICT and transport sectors. The company specialises in the
development, application, exploitation and regulation of terrestrial (fixed and
wireless) and satellite-based communications, surveillance, broadcast and
navigation technologies and also has significant expertise in aviation and associated
markets.
We provide high quality consultancy encompassing everything from concept
development to regulatory impact assessment; from technology roll-out and
commercialisation to business case analysis and investment appraisal.
We support businesses, governments, regulators and other institutions. Our
customers usually work in complex regulatory domains, in safety critical industries
and are supported by advanced technology. Our aim is to improve corporate
performance.
Our success has been recognised through two Queen’s Awards for Enterprise (in 2004
and 2009).
Get in touch…
For further information please contact:
Richard Womersley
Helios
29 Hercules Way
Aerospace Boulevard
AeroPark
Farnborough
Hampshire
GU14 6UU
UK
T +44 1252 451 651
F +44 1252 451 652
W www.askhelios.com
“It has long been an axiom of
mine that the little things
are infinitely the most
important.”
Sir Arthur Conan Doyle
“[The] next step in [the]
development [of the Internet]
is to progressively evolve
from a network of
interconnected computers to
a network of interconnected
objects, from books to cars,
from electrical appliances to
food, and thus create an
‘Internet of Things’.”
Internet of Things — An
action plan for Europe
Offering a new means of
delivering connectivity for the
smart grid is potentially
lucrative for telecoms service
providers
The ‘Internet of Things’
Introduction
Telecommunications began by enabling people to talk with other people. As
technology has progressed people started to communicate in a variety of
ways using ‘machines’ (computers) as intermediaries. Machines are
effectively communicating with other machines, largely guided by human
operators. It is logical therefore, that machines will begin to communicate
with other machines without human intervention. As smaller and smaller
machines become able to communicate, a new Internet paradigm emerges:
an Internet of machines, sometimes called machine-to-machine
communication but increasingly being called the Internet of Things (IoT).
One of the great questions relating to the IoT is exactly how devices will
connect to each other, given the large number of potential entities on a
network. Wiring things together would be cumbersome and expensive and
therefore the obvious means of achieving connectivity is through the use of
wireless networks. This paper considers the opportunity for cellular
operators to capitalise from the impending growth in the IoT. Arguably, the
IoT starts with the Internet of important Things and first amongst the
important uses of the IoT is likely to be smart meters. Using smart meters
as an example, we consider the size and scale of the opportunity for
cellular operators and how a joint cellular-mesh network approach may
prove to be the ‘dream team’.
Smart grid and the Internet of Things
It is widely recognised1 that the next stage in the evolution of the Internet
is to move from connecting people (using computers) to connecting objects.
This concept of interconnected objects has been termed the ‘Internet of
things’. In reality, however, the IoT is an ‘Internet of Internets’ in which
like objects may form internets (or intranets) of their own, which are then
in turn joined together to form the wider IoT.
Machine-to-machine (M2M) communication, which has typically been
addressed by GSM-based technologies, represents one widely recognised
example of a part of the IoT, but this is just one small part of a much larger
jigsaw. Each of the different elements which will together complete the IoT
jigsaw will have varying sizes and connectivity requirements. One issue yet
to be addressed, however, is whether there is a ‘one size fits all’ solution
to connecting devices together. It seems highly likely that low power, small
size devices, may require different data transmission protocols from larger,
higher power devices. This means that a ‘one size fits all’ approach may
not be efficient.
Some ‘Internets’ are also more easily implemented than others due to size
and power availability: developments in ‘smart grid’ technology which
interconnects utility meters are therefore a logical first step towards the
growth of the IoT, given the availability of local power sources and of space
within meters to accommodate a connecting device. Offering a means of
delivering connectivity for the smart grid, and for future M2M devices
therefore delivers a potentially lucrative business opportunity for
telecommunication service providers.
1
The size of the opportunity is vast: the European Commission estimates
that there may be around 50 to 70 billion machines across Europe that need
to be connected2. This equates to an estimated 5 billion in the UK alone,
which is more than the current worldwide total number of cellular
subscribers3. This is before many of the new ideas come to fruition, such as
those suggested by thought groups such as the European Commission’s
Future Internet 2020 task force. All these objects will require data
connectivity and it is envisaged that this connectivity will be almost
universally wireless. Further, the capacity required to connect objects
together will be very significantly more than is currently available using
2.5G and 3G technology today.
Given its immediacy, the remainder of this paper uses the concept of smart
meters as a case study to examine the opportunities that may be presented
to cellular operators by the IoT. It is important to remember, however, that
today’s smart grid (the Internet of important Things) is only one part of
tomorrow’s IoT and represents the tip of the iceberg insofar as its likely
eventual development.
UK smart meter plans
There are currently 28.5 million electricity meters in the UK4, together with
a similar number of gas meters, and around 8 million water meters (though
this is expected to expand to 16 million by 20305). Thus there are well over
60 million utility meters in use in the UK today.
The UK government has committed to roll out smart energy metering to all
households by 20206. In parallel, the UK water industry regulator (Ofwat) is
hoping to piggyback on this roll-out to connect water meters too. The plans
for this smart grid require two-way connectivity and thus there will be a
need to provide solutions for these 60 million units over the next 10 years.
In some countries, the use of wired means of connecting smart meters into
the necessary management and monitoring systems using power line
telecommunications (PLT) is being considered, however this has a number
of drawbacks: as well as only being directly applicable for electricity
meters (as they are the only devices immediately connected to the
electricity infrastructure), there
has been strong vitriol in the UK
against PLT devices7. They are
unpopular due to the amount of
radio interference which they
generate. Unpopularity aside,
PLT has not been proven to meet
the reliability or throughput
requirements necessary for a
truly smart grid, therefore the
UK’s smart grid will be largely
wireless. This represents a
significant opportunity for
cellular network providers across
both the UK and Europe to secure additional new revenue streams,
assuming that a suitable business model can be found in which the
connectivity can be delivered profitably.
“[The Internet of Things]
requires truly ubiquitous
wireless capacity that can
handle several magnitudes
more data.”
Future Internet 2020
The UK government has
committed to roll out smart
energy metering to all
households by 2020
Wireless offers advantages
over wired connectivity
2
Wireless smart metering
offers cellular operators the
chance to capitalise on their
existing network infrastructure
Each meter will generate
<2kByte of traffic per day
As networks are upgraded,
modems will need to be
replaced. This could mean
120m visits by 2020
The Average Revenue Per User
from embedded devices is
likely to be very small
Smart grid using a cellular network
Advantages of a cellular solution
Wireless smart metering offers cellular operators the chance to capitalise
on their existing network infrastructure by offering a connectivity solution
to the 60 million UK utility meters, thereby generating additional revenues.
However, such connections will inevitably deliver low Average Revenue Per
User (ARPU), while consuming network resources such as signalling
congestion and upstream bandwidth.
The straightforward means to access this opportunity is to place a SIM and
associated wireless card (whether 2G, 3G, 4G or beyond) into each meter.
Given the small volume of traffic that each meter is likely to generate
(< 2kByte per day), this would not present a significant growth in network
traffic (assuming that the load is spread throughout the day). Of note, most
pricing plans offered to date are restricted to off-peak meter reads, which
discourages a more interactive smart grid.
Managing this number of subscribers should not present a significant
headache. Not only can modern networks deal with these kinds of numbers
of subscribers, but the fact that the meters are static means that the
frequency of location updates can be reduced (or forced updates only
employed) further reducing the overall load on the network.
Disadvantages of a cellular solution
A large obstacle in a cellular solution arises as technology is upgraded. If
each meter is fitted with a suitable radio modem, as networks are
modernised, these modems would need to be replaced. With 60 million
meters in the UK, even if it is simply a case of replacing each module, this
would represent a significant effort on the part of either the network or the
meter supplier or fitter.
This is a significant challenge: whilst normal cellular users will
automatically migrate to new technologies as their handsets are upgraded,
fixed installations of any kind will require wholesale manual replacement.
The level of work involved would be similar, for each technology upgrade,
as for the initial roll-out of services and whilst a gradual migration from one
technology to another could take place, the need to revisit 60 million sites
still represents a significant undertaking. This is a very real possibility
should GPRS be chosen as the sole smart metering communications
technology in the UK. Even if this is limited to one visit to each of the UK’s
approximately 27 million residences, due to the strong likelihood of GPRS
obsolescence within 7 to 10 years, there may be a second visit prior to the
2020 mandate deadline.
An arguably larger problem is that of coverage: smart metering and smart
grids require ubiquitous coverage. Meters are located indoors and in hard-
to-reach locations. Handsets are mobile, meters and transformers are not.
Another problem posed by a solution of this type is that the necessary
investments may not yield significant financial returns, especially if the
annual line rental charged for each device is small. The ARPU from
3
embedded devices of this type is likely to be very small. In some instances
it may be so small that the amortised cost of the SIM and wireless card may
take longer to recover than the life of the card itself. A UK government’s
Impact Assessment (Baringa Partners) estimates the yearly GPRS backhaul
cost as £4.83 for one off-peak meter read per day. Evidence exists that
competitive pressures will further drive this price down. And once-a-day,
instead of true on-demand interaction, is an impediment to a truly useful
smart grid.
As mentioned, coverage may be a concern, not least because many meters
will be situated in locations where coverage is difficult to achieve, notably
indoors and in particular in basements. GSM coverage of the UK population
is upwards of 98% and as such most domestic and commercial properties
should be within an area of coverage. However, whether this would be
available indoors, particularly in less dense non-urban areas, is less certain
and it is likely that some additional sites would be needed to make
coverage ‘deeper’. The penetration loss at 900MHz8 for a signal passing
through two walls is around 19dB which is higher than the 10dB building
penetration loss which operators normally plan for when attempting to
deliver indoor coverage. An additional 10% indoor penetration requires 4 to
6dB of additional signal strength, requiring an increase in site density of 3
to 4 times. To deliver an additional 9dB of signal strength may therefore
require up to 10 times the number of cell sites.
The use of femtocells may alleviate some of this problem in the longer
term. However, it is unclear how the problem of lack of coverage would be
dealt with in areas where this technology is not available if a cellular
solution were to be universally adopted, and whether or not a ‘non 100%
compliant’ solution would be politically acceptable, leaving some users
unable to use smart metering.
Wireless mesh networks
Advantages of mesh networks
Mesh networks are communication networks in which each mesh radio (or
node) is capable of connecting to one or more other nodes and of injecting
data into the network, receiving data destined for it, and passing data from
one node to which it is connected to another. In the case of smart
metering, think of each meter as a unique picocell. Mesh networks are also
capable of self-healing: in the event that any node is lost, traffic can
usually be re-routed around it. In essence, wireless mesh networks form a
‘wireless internet’.
ZigBee is an example of an early mesh technology and this and other
meshing technologies using 2.4GHz spectrum have proven adequate for
short range, intra-home applications. However, due to range limitations
using 2.4GHz, mesh networks has proven impractical to deploy at scale due
to coverage issues and infrastructure costs.
Much work has been conducted to assess the effectiveness and efficiency of
wireless mesh networks9 and they are already delivering significant
benefits, in particular in applications such as smart metering. The coverage
Many meters will be situated
in locations where coverage is
difficult to achieve
With mesh networks, each
node connects to many other
nodes, and networks are
capable of self-healing
4
Much work has been
conducted to assess the
effectiveness and efficiency
of wireless mesh networks
and they are already
delivering significant benefits
There is no need to change
the radio device unless the
requirements of the meters or
the devices themselves change
As networks grow, the number
of nodes connected through
any Access Point can begin to
exceed its capabilities
One solution might be to put
a GPRS or 3G card in every
meter
of a mesh network is determined by the geometry of the nodes themselves
as they are rolled-out. In an application such as metering, nodes will be
situated in domestic and commercial properties and coverage will be
extended as new nodes are
installed. If appropriate radio
frequency and power levels
are used, the short distance
between nodes means that it
becomes relatively
straightforward to provide
connectivity in otherwise
difficult to access locations
such as in basements or deep
inside properties. As the node
density increases, the number of nodes that any particular device will be
able to connect to will increase providing alternative connectivity options
and resilience of coverage in the event of any failures.
One of the advantages of mesh technology over cellular in a smart metering
context is that the mesh radio node can be designed with the exact data
rate needed, and transmitter duty cycle in mind, so that it is optimised for
the throughput of the meter or other device to which it is connected (and
to supporting connected nodes). The knock-on effect of this is that there is
no need to change the radio device unless the requirements of the meters
or other underlying devices themselves change. Should any changes in
metering take place, which would require a change to the meters
themselves, the associated mesh device could be replaced at the same
time. There is thus no need to replace mesh radios unless the underlying
device to which they are connected also changes.
Disadvantages of mesh networks
Unlike mobile networks wherein each user inherently has a way into and
out of the mobile network to connect to third parties, as a stand-alone
network, wireless mesh delivers limited connectivity, only permitting
connection between nodes on the network. To enable ingress and egress
from a mesh network, mesh access points (APs) need to be deployed at
appropriate locations (the equivalent to cellular base stations in a mobile
network). Typically 1000 or more mesh nodes can be served by each AP −
the exact number depending on the density of node installations and radio
range (a function of spectrum and transmit power). These APs, or
gateways, provide points for network management as well as for collecting
and distributing data amongst the nodes.
In theory, APs can be located anywhere amongst the various nodes, and it
helps if the AP can connect with multiple neighbouring nodes. The more
nodes it can connect to, the shorter the path between remote nodes and
the central AP, minimising delay and latency, and providing additional
resilience. If the number of APs is fixed, as networks grow, the number of
nodes connected through any AP can begin to exceed its capabilities such
that additional APs would need to be established. One solution to this might
be to put a GPRS (or 3G) card in every meter. If each meter were fitted
with both a mesh node and a mobile device some meters could act as APs
and in others, the GPRS (or 3G) device need not be activated.
5
Connecting the APs to the central management function and data back-haul
network can be achieved through a variety of means, and commonly fixed
or wireless broadband or cellular connections are used.
• Fixed broadband connections are limited in that they require the AP to
be situated in a location where there is a telephone line available.
These locations are often not optimum for providing connections to
multiple nodes without additional cabling between the telephone line
and a vantage point providing better coverage from the AP.
• The use of a bespoke wireless network may be suitable where nodes
can be situated in line of sight positions to a central concentrator site,
or in line of sight situations to each other. This restricts flexibility and
though any AP located in a position with a good line of sight is likely to
be able to connect to multiple nodes, it does restrict the choice of
sites.
• A cellular solution is a particularly flexible option to provide the
necessary connectivity for mesh network APs. Not only is coverage
likely to be present in locations which are also good sites for accessing
multiple neighbouring nodes, but the additional flexibility of being able
to move APs easily offers the opportunity to optimise the location of
APs as the mesh network grows. Even with 1000 nodes connected to a
AP, the total volume of data generated (from a smart meter network)
is very much within the capability of even 2.5G cellular networks and
would present a very small load on a 3G or 4G network.
Thus, there are strong synergies between mesh APs and cellular networks.
In many cases, the cell sites themselves would represent ideal locations for
the positioning of the APs, providing an alternative model both for cellular
operators and for mesh operators. Instead of using the cellular radio
interface for the backhaul from the mesh APs, the data could be routed
directly over the UTRAN and either interconnected at the RNC or continue
to the core network to a point of interconnection further towards the edge
of the network. From the point of interconnection, mesh traffic could be
routed to a separate interconnection point and thus the cellular network
would act as a virtual tunnel. The advantage of a connectivity model of this
kind would be to overcome the restrictions on gateway capacity.
There are a range of solutions
for connecting the Access
Points to the central
management function
The cellular network can be
configured to act as a virtual
tunnel for mesh traffic,
overcoming restrictions on
gateway capacity
There are strong synergies
between mesh Access Points
and cellular networks
6
Radio
Network
Controller
(RNC)
Radio
Network
Controller
(RNC)
Node B +
Mesh AP
Node B +
Mesh AP
Node B +
Mesh AP
UMTS Terrestrial Access Network (UTRAN)
to Core Network
to Mesh NetworkRadio
Network
Controller
(RNC)
Radio
Network
Controller
(RNC)
Node B +
Mesh AP
Node B +
Mesh AP
Node B +
Mesh AP
UMTS Terrestrial Access Network (UTRAN)
to Core Network
to Mesh Network
Operators can remove the risk
of technology upgrade and
enjoy deployment flexibility
by using cellular connections
or sites to activate mesh
network Access Points
Mesh networks adapt with
each new connection
They permit new types of
devices to be connected
The radio connection is only
changed when the device
itself changes
A hybrid cellular-mesh solution
The ideal solution?
There are undoubtedly cost benefits to using a wireless mesh network to
provide the connectivity into and out of buildings given the self-provided
coverage that mesh networks deliver. Further, there are clearly capacity
benefits in optimising the
use of scarce radio
resources for a particular
application by tailoring the
radio technology to the
requirements of the device
to which it is connected.
Smart metering and smart
grids have limited data and
throughput requirements,
often resembling a process
control application; here,
reliability trumps
throughput. Using cellular connections or sites to activate mesh network
APs provides an unsurpassed amount of flexibility in deployment and
importantly removes the risk to operators of technology upgrade.
The combination of mesh and cellular technology provides a ‘dream team’
for smart metering in that it is:
• Scalable: Alone, cellular networks would begin to creak as the number
of IoT devices grew to the kind of levels foreseen by the IoT. Mesh
networks adapt with each new connection, but would require an
increasing number of easily configurable, and well located APs.
Importantly, mesh networks are today being designed to leverage the
vast address space made available using IPv6. The two together allow
for planned and for ad-hoc expansion to the kinds of hundreds of
millions of devices which are likely to emerge.
• Adaptable: Cellular networks are restricted to certain message sizes
which may not suit the variety of different applications that will make
up the IoT, making inefficient use of the valuable network resources.
On the other hand, mesh networks can be specifically combined to deal
with certain traffic types. Combined they offer the ultimate in
adaptability, permitting new types of devices to be connected in the
most effective and efficient manner.
• Able to deal with obsolescence: Replacing the immense number of
wireless devices the IoT may comprise, would be an almost impossible
task. With mesh technology, the radio connection is only changed when
the device itself changes, minimising the effort of replacement and
reducing the carbon footprint caused by equipment scrappage. The
obsolescence cycle for smart meters, for instance, is 15 to 25+ years,
markedly longer than that of mobile technologies.
• Of sufficient coverage: Cellular network coverage, whilst widespread,
is insufficient to ensure the depth of penetration required for IoT
devices. Coupling cellular’s star topology architecture with a meshed,
7
“Telecommunication operators are usually
called for with regards to the long distance part
of the picture … However, the other area where
we think that operators can add value is the
local data collection network … GSM might not
be appropriate in all cases to connect the meter
up to the information system … and also it
might not be economically viable to roll out
based on GPRS. So, we think there is a business
case for advanced radio technologies - mesh
networks.”
Valerie Le Peltier, Director M2M Vertical,
Orange Group10
peer-to-peer architecture extends coverage without needing to build
additional base stations, offering a cost attractive way ahead.
• Profitable: Putting all IoT devices onto a cellular network clearly
maximises revenue potential, however this revenue comes at a high
cost and thus may be low profit. Working together with mesh
technology to deliver a blended solution may reduce revenue, but at
much reduced risk and cost and thus potentially higher profitability.
Smart metering: how each solution compares
The following table highlights the advantages and disadvantages of the
three potential solutions to cellular operators.
There are clear benefits for cellular operators to work closely with mesh
providers, and close integration of the networks/technology may deliver
the significant additional benefits of securing mesh traffic and thus
revenues to one particular network. Whilst in the first instance, such close
integration of mesh and cellular technologies may seem unnecessary, as the
IoT grows, such a model may be the most effective and perhaps the only
means of ensuring the bandwidth and flexibility necessary for growth.
Beyond smart metering
At a concentration level of 5000:1 as in systems deployed in the Americas
and Australia, meeting the UK’s smart metering requirements would require
12,000 APs to support a mesh network (coincidentally similar to the number
of cell sites currently deployed by each of the UK operators).
For cellular operators, using mesh networks as the ‘last mile’ connection
presents a number of distinct advantages:
• An enormous reduction in the necessary investment in additional
infrastructure to provide coverage
Coupling cellular coverage
with a meshed peer-to-peer
architecture is practical and
cost-effective
8
Network type Pros Cons
Cellular only Large anchor tenant providing guaranteed income.
Control of M2M/IoT connectivity maximising long-
term growth opportunities.
Management of the tens or hundreds of millions of
connections.
Widespread, deep coverage requirement requiring up
to 10 times as many sites.
Potential for expensive replacement programme
when technology upgraded.
Limits in the number of simultaneous connections
may be a problem with millions of devices.
Integrated
cellular and mesh
network
Exceptionally straightforward and flexible roll-out.
Shared network management ensures best tools for
specific connection.
Achieves indoor coverage at low cost.
Additional coverage can be provided quickly and
with minimal investment.
Equipment upgrade only when underlying service
changes.
Optimised for particular device to which it is
connected thus low replacement cycle.
Requires cellular operators to permit access to some
elements of UTRAN to mesh traffic.
Close working relationship with suitable mesh
provider needed.
Ensures revenue from mesh network stays with one
operator.
Mesh using
cellular as back-
haul
Mesh provider would be able to use alternative (non-
cellular) connection means, reducing revenue
potential.
The challenge for UK and
European cellular operators is
to ensure that the benefits of
a ‘best fit’ approach to
delivering smart metering,
M2M and IoT connectivity are
realised
• Thousands instead of millions of devices to replace in the event of
network technology upgrade
• Thousands of high ARPU connections instead of millions of tiny ARPU
connections
• Simpler subscriber and network management
• Lower risk of network overload due to multiple simultaneous
connections
However, as the IoT develops from smart metering, initially into M2M
connectivity and beyond, the advantages of a joint solution would become
increasingly large.
The challenge for UK and European cellular operators is therefore to ensure
that the benefits of a ‘best fit’ approach to delivering smart metering, M2M
and IoT connectivity are realised.
How Helios can help Helios appreciates the challenges facing both those implementing smart
grid and those whose role is to provide the necessary connectivity. We
understand the business, technical and regulatory environment and are
ideally positioned to support organisations wishing to be some of the first
movers in the race to deliver the benefits of smart grid and of the Internet
of Important Things.
References 1. http://ec.europa.eu/information_society/policy/rfid/documents/commiot2009.pdf
2. http://www.future-internet.eu/news/view/article/future-intenet-2020.html
3. http://www.itu.int/ITU-D/ict/newslog/
ITU+Sees+5+Billion+Mobile+Subscriptions+Globally+In+2010.aspx
4. http://www.parliament.the-stationery-office.co.uk/pa/cm200607/cmhansrd/cm070518/
text/70518w0013.htm
5. http://www.defra.gov.uk/environment/quality/water/industry/walkerreview/documents/
walker-call-for-evidence.pdf
6. http://www.decc.gov.uk/en/content/cms/consultations/smart_metering/
smart_metering.aspx
7. http://www.rsgb.org/news/pla_dispute_law.php
8. http://www.its.bldrdoc.gov/pub/ntia-rpt/94-306/94-306.pdf
9. http://www.ofcom.org.uk/research/technology/research/emer_tech/mesh/
10. http://www.telecomengine.com/newsglobe/article.asp?HH_ID=AR_5940
9
The content of this document is intended for general guidance only and, where relevant, represents
our understanding of current status of telecoms industry matters. Action should not be taken without
seeking professional advice. No responsibil ity for loss by any person acting or refraining from action
as a result of the material in this document can be accepted and we cannot assume legal l iabil ity for
any errors or omissions this document may contain.
© Hel ios Technology Ltd - June 2010
Al l rights reserved.
Helios
29 Hercules Way
Aerospace Boulevard | AeroPark
Farnborough | Hampshire | GU14 6UU | UK
T +44 1252 451 651
F +44 1252 451 652
W www.askhelios.com