Sie sind auf Seite 1von 9

3/31/2016

IoT A thing of our


interest?
Personal Views & Analysis

Ajay Vachhani

Ajay Vachhani

IoT A thing of our interest?


Personal Views & Analysis

IoT Internet of Things


Lately, this IoT as a new technology trend and market buzzword grabbing the
phenomenal attention of mobile communication world. This paper analyses the IoT with
perspective of wireless communication.
Simply put, its a network of things or better say `connectivity of things`. Here Things
could be Man or Machine. While M2M was originally defined Machine to Machine
connection, IoT redefines the M2M being relationship between Man 2 Man, Man 2
Machine, Machine 2 Machine so a wider definition of connected things.
Now what this `connectivity of things` could mean to us and impact our lives going
forward I would let you play with your creativity in imagining & painting your own
world, sorry `smart` world (with IoT everything turns Smart ). There is endless
material on web talking through numerous use cases of IoT on several different verticals
and segments. So we shall stay away from that discussion in this paper.
Lets call these Things as End Points and Connectivity as Network Infrastructure. The
Network Infrastructure could be of wireline or wireless nature. The Wireless Network
necessitates the need of having radio spectrum to connect end points.
So what do an IoT need

Network Infrastructure (with Radio Spectrum)


End Points (sensors, actuators, smart devices, ..) requiring wireless modem as
Tx/Rx unit to connect through network infrastructure

With above definition IoT could be of two types


1) A low cost, low power end points (Tx/Rx unit) with low bandwidth (narrow band) and
long range of spectrum & network this is called Massive-IoT. Industry is
enthusiastically making claims that there would be around 20billion or more (some
predicting up to 50billion) connected end points by 2020.
Example Use Cases Smart Metering, Automation, Smart City etc

Ajay Vachhani

2) Another, with high end points requiring sending/receiving of large chunk of


information necessitating high bandwidth and high performance network this is
termed as Mission Critical IoT
Example Use Cases Video Surveillance, Industrial Plant and Processes etc

What Infrastructure to use for IoT?


Clearly if you ask mobile network operators, they would unanimously say cellular
network, to be precise LTE cellular network, a future proof network where they are
investing billions of dollars in build out hence would look to leverage and reap the
benefit.
That said, but is LTE network a right fit?
For mission critical IoT answer is YES as LTE is indeed a high data rate, high
performance network built to enhance QoS and QoE so it qualifies and suits to mission
critical IoT.
Its this massive IoT which puts forth and demand different characteristics like

Low Cost modem devices such that connectivity of massive 50billion devices
can be practically realized. It has to be low cost to succeed (industry expectation
<$10).
Low Power devices the power of 2AA battery should last minimum for 10 years.
Low Data Rates it just require sending/receiving of information on changes &
updates
Long Range - thats necessary to keep the infrastructure cost low

Lets cross examine suitability of LTE network for massive IoT


LTE is developed to squeeze in more data in available spectrum for higher spectral
efficiency leading to complex design of modem costing much higher (around $40) than
expected price point (around $10). The wideband spectrum of LTE is more power
hungry hence failing to meet low power demand of IoT devices. The range is fine as
800MHz spectrum can be used with LTE.
At this stage, a critical reader (or call it a Smart reader ) may ask this question why
not use GPRS network instead its a narrow band network available for years giving
economies with its simple modem design and low power consumption

Ajay Vachhani

Rightly asked, but MNOs are starting to turn off their ageing GPRS networks, as they
want to use the spectrum for more 4G. There is a very logical rationale for that. The
same amount of spectrum used for 4G supports around three to five times the number
of users as 3G and ten+ times the number of 2G data users. The need of hour is high
data rate but massive IoT is demanding contrary to this hence creating a hole in
technology.
Now this hole means an opportunity for others, hence paved the way for plethora of new
and existing options mushrooming in the market, some being

Sigfox
LoRa
Weighless
Telensa
On Ramp

Out of these the two are gaining good popularity and market tractions - Sigfox and
LoRa. Refer the appendix for comparison of them, as an additional information.
So the takeaway so far is, LTE technology is struggling to meet IoT demand for
infrastructure hence other options are emerging in the market.

What Spectrum to use for IoT?


The technologies like Sigfox and LoRa, with their own proprietary protocol stack,
promote usage of unlicensed ISM bands 915MHz in US and 868MHz in Europe. The
Weightless suggests usage of white space spectrum of 400-800 MHz in TV Guard band
but regulators have not opened the use of these bands as yet.
A valid question again crops up Is unlicensed spectrum good enough for massive
number of devices, expected to be lasting for years?
No, it does not look good. What it means is that MNOs have significant role to play for
IoT to be successful as they are the ones who own licensed spectrum.
Ok, but then MNOs are stuck at this stage with that technology hole, as mentioned
earlier. So its a deadlock, no?

Ajay Vachhani

How 3GPP is trying to fill in that hole?


3GPP is working on fixing this since Release-12 onwards.
Release 12, published in March 2015, and introduced a new lower complexity Cat-O
device. This has a reduced peak rate of 1mbps for both downlink and uplink, achieved
via reduced transport block size, half duplex capability and only one antenna.
3GPP estimates Rel. 12 can reduce the bill of material of a Cat-O LTE module by 50%.
Rel.12 also introduced Power Saving Mode (PSM), enabling devices to enter sleep mode,
thereby saving power, for long periods of time.
Release13 (due out 2016) will reduce the complexity still further by defining a Cat-M
device, which will offer even lower uplink and downlink peak rates, half duplex and with
a modem 75% less complex and less costly than standard devices.
There is another variant introduced in Release-13 called NB-IoT (Narrow Band IoT)
which is expected to operate in less than 200KHz bandwidth.

Ok, 3GPP is doing its job but when can devices be commercially
made available?
Firstly there are two vendor camps within 3GPP standardization with different
viewpoints. Huawei (after their acquisition of an IoT company called Neul) argues that
its very hard to achieve low power long range network through the evolution of existing
cellular radio access technologies and that this is because the IoT requirements are so
different from mobile broadband. So we should start fresh and develop what they call it
Clean Slate solution.
Other camp including Nokia and Ericsson claim that they can do plumbing of LTE to
make it fit for IoT.
Huawei approach is more expensive for MNOs while Nokia/Ericsson suggestion only
require software update to existing network infrastructure hence MNOs prefer
Nokia/Ericsson approach to fix LTE.
Nevertheless, a first commercial device can not be expected until 2018-19 even if
standardization completes its job in 2016. And moreover without economies of scale it
can not meet the price expectation of market from its introduction.
Fair enough, so then what MNOs are doing here? Whats their approach?

Ajay Vachhani

What is MNOs current approach towards IoT market?


Well, they are left with two choices Either wait and develop device ecosystem &
accelerate the device availability, or adopt any existing option e.g. Sigfox or LoRa or
other.
The MNOs like Telefonica, T-Mobile have invested in Sigfox while Orange, Swisscom,
SK Telecom, Du (Dubai) have gone on putting their bet on LoRa. Its expected that list
on both sides will grow further in coming months. While others continue to wait for
LTE-M and NB-IoT options to become ready.
Though Sigfox or LoRa are completely new infrastructure unlike LTE based options
(LTE-M and NB-IoT), nevertheless MNOs are well placed to leverage their
infrastructure as sites, power, mast and other necessities (e.g. management system and
back office) in building Sigfox or LoRa network. So it does not look foolish that MNOs
are going ahead in adopting Sigfox and LoRa after all early birds and first movers
generally get better market share.

What is the real challenge in IoT market?


No matter which standard Sigfox, LoRa or LTE sets its root and gains market traction
the real challenge lies in providing the cheapest possible module shipped in the greatest
possible volumes to achieve economies of scale. It requires a right business model to
scale irrespective of technology.
Most of these IoT devices may not cross national boundaries, the majority will probably
be static, so it does not need to be global. So there is room for multiple technology
options to set its foot and flourish in their targeted geographical market.

Finally, what it means for NEC Small Cell Business?


REMOVED.

===========================================================

Ajay Vachhani

Appendix Sigfox Vs LoRa

(Source: Web)

SigFox and LoRa have been competitors in the LPWAN space for several years. And
while the business models and technologies behind the companies are quite different,
the end goals are very similar. The goal of the LoRa Alliance, LoRaWAN adopters, and
SigFox is that mobile network operators adopt their technology for IoT deployments
over both city and nationwide low power, wide-area networks (LPWANs). But there are
some prominent differences between how each technology plans to achieve this goal and
which applications the technology is best suited for.
Technology
SigFox
SigFox is a narrowband (or ultra-narrowband) technology. It uses a standard radio
transmission method called binary phase-shift keying (BPSK), and it takes very narrow
chunks of spectrum and changes the phase of the carrier radio wave to encode the data.
This allows the receiver to only listen in a tiny slice of spectrum which mitigates the
effect of noise. It requires an inexpensive endpoint radio and a more sophisticated
basestation to manage the network.
SigFox communication tends to be better if its headed up from the endpoint to the
basestation. It has bidirectional functionality, but its capacity going from the basestation
back to the endpoint is constrained, and youll have less link budget going down than
going up. This is because the receive sensitivity on the endpoint is not as good as the
expensive basestation.
LoRa & LoRaWAN
LoRa is a spread-spectrum technology with a wider band (usually 125 kHz or more). Its
frequency-modulated chirp utilizes coding gain for increased receiver sensitivity.
LoRaWAN looks at a wider amount of spectrum than SigFox (and thus gets more
interference). However, because its looking for a very specific type of communication,
the elevate noise due to a larger receiver bandwidth is mitigated by the coding gains.
Practical link budgets are about the same for SigFox and LoRaWAN.
Unlike SigFox, both the endpoint and the basestation are relatively inexpensive with
LoRa-enabled devices. This is primarily because you can use the same radio for a
6

Ajay Vachhani

receiver on the basestation and at the endpoint. While the LoRaWAN basestation tends
to be more expensive than the endpoint, its inexpensive in comparison to a SigFox
basestation.
Business Model
SigFox
The SigFox business model takes a top-down approach. The company owns all of its
technologyfrom the backend data and cloud server to the endpoints software. But the
differentiator is that SigFox is essentially an open market for the endpoints. SigFox gives
away its endpoint technology to whatever silicon manufacturer or vendor wants it so
long as certain business terms are agreed upon. Large manufacturers like
STMicroelectronics, Atmel, and Texas Instruments make SigFox radios. SigFox thinks
that allowing the application to be really inexpensive is the way to drive people to its
market
SigFox endpoints use commodity MSK radios, and they are relatively inexpensive. You
can get a chip for a few dollars and a module for less than $10 in high volumes, so
SigFox partners arent bringing in much money from the hardware itself. SigFox makes
its money by getting network operators to pay royalties on reselling its technology stack
to customers. In other words, SigFox gives away the hardware enablers but sells the
software/network as a service. In some cases, the company actually deploys the network
and acts as the network operator. This is the case in France and in the US; when you buy
LPWAN service there, youre operating on a SigFox network.
SigFoxs ultimate goal is to get large network operators from all over to world to deploy
its networks. It has raised more than 100 million to do this and has great global reach.
SigFox has been around since 2009 (longer than almost everyone else in the space), and
its likely the most aggressive marketer in IoT.
SigFox is of the opinion that its easier to work with mobile network operators or deploy
networks itself and charge a small recurring fee than to sell expensive hardware at the
endpoint. However, there are some challenges associated with this business model. For
one, if you want to deploy a SigFox network, you have to work directly with SigFox
there isnt another option. Additionally, only one SigFox network can be deployed in an
area, because the company has exclusive arrangements with network operators when
they work together.
LoRa

Ajay Vachhani

The LoRa Alliance has a different strategy. They would say theyre more open than
SigFox, strictly because the specification that governs how the network is managed is
relatively open. You can download the specifications and join the LoRa Alliance, and any
hardware or gateway manufacturer can build a module or gateway that conforms with
the LoRa specifications. The catch is that the only company that makes the radio for
LoRa is Semtech. (Theyve announced licensing to other silicon manufacturers in the
future, but Semtech is the only option right now.) So while the ecosystem itself is open,
it does have a closed element.
One nice thing about LoRas open standard is its potential to be very flexible; its not
going to be driven by a specific company. In practice, this does force development to be
a little slower, because youre developing standards by committee.
The LoRa Alliance believes that openness creates adoption, so members stress that
anyone can join the Alliance and build hardware to support it. The trick here is how
companies who adopt LoRaWAN can add value. Just like SigFox, the LoRa Alliance
wants network operators to deploy the LoRa networkbut they also want private
companies and startups to do so. To allow for this, theyve developed some discussion
around roaming network to network. The business and technology around this idea isnt
fleshed out yet, so one of the next steps will be to figure out how to allow for roaming
from public network to public network and private network to private network.
Use Cases For LoRa & SigFox
LoRa is likely the better option if you need true bidirectionality because of the
symmetric link. So if you need command-and-control functionalityfor, say, electric
grid monitoringLoRa is your best option. (Full disclosure: Link Labs is a proud
member of the LoRa Alliance, and we build on this technology.
With SigFox, you could use bidirectional command-and-control functionality, but to
work appropriately, network density would need to be higher (due to the asymmetric
link). Therefore, it is better for applications that send only small and infrequent bursts
of data (like alarms and meters).
Other than these minor differences, SigFox and LoRa serve similar markets. Its worth
noting that both technologies were originally designed for the European regulatory
bands between 865 and 868 mHz, and theyve both faced challenges in coming over to
the regulatory markets in the U.S. Progress is being made, and both technologies are
working toward optimization for FCC use.