Você está na página 1de 16

Best Fit Use Cases

for LPWANs
www.abiresearch.com

EXECUTIVE SUMMARY
Low Power Wide Area Networks (LPWANs) are one of the foundations underpinning the Internet of Things (IoT) and its
growth to more than 8 billion M2M and LPWA devices connected by 2021, while LPWA connections excluding the legacy
cellular machine-to-machine (M2M) technologies, which include 2G, 3G and LTE Cat-1 and higher, will reach more than 400
to 500 million worldwide in the same time period. Offering low-power and long-range communications, LPWANs stand at the
verge of enormous growth; however, there are many competing technologies to choose from when deploying an LPWAN. Each
of the LPWA technologies offer a trade-off between factors including transmit range, data rate, frequency, channel bandwidth,
and power consumption, among others. In this highly competitive environment, each technology vies for position, often with
conflicting claims that can be complex to understand.

In this white paper, ABI Research has chosen to clarify the situation by examining six LPWA technologies, outlining their main features and determining which use cases are a best fit. We believe that the proper choice
of LPWA technology for an application can avoid costly mistakes and lead to successful LPWAN deployment
and operation.
There are two main classes of LPWA technology, which are distinguished by whether they operate on unlicensed industrial, scientific, and medical (ISM) frequency bands or licensed cellular frequencies. The LPWA
technologies operating in the unlicensed frequency bands are often proprietary, and the licensed frequency
LPWA technologies are based on versions of cellular LTE. We compare the leading unlicensed LPWA technologies from Ingenu, SIGFOX, and LoRa with the main Mobile IoT (also called LTE IoT) technologies LTE CatM1 (also known as eMTC), NB-IoT (also known as LTE Cat-NB1), and EC-GSM-IoT.

ABI Research believes that the


Random Phase Multiple Access
(RPMA) technology from Ingenu
has the broadest appeal across
diverse use cases

ABI Research believes that the Random Phase Multiple Access (RPMA) technology from Ingenu has the
broadest appeal across diverse use cases and bridges the gap between low-cost LPWA for simple applications to what is anticipated to be high-performance LTE IoT. RPMA offers a carrier-grade LPWAN on par with
LTE IoT. While RPMA was built from the ground up to target the industrial use case, we believe it has proven
its capability to tackle applications in the smart cities, smart buildings, and transport and logistics use cases.
ABI Research also notes that although the specifications for LTE IoT have been finalized, the industry is performing trials on pilot networks now and through 4Q 2016. LTE IoT is targeted to launch commercially in the
first half of 2017 by MNOs including AT&T, Orange, and Vodafone. With significant support from MNOs and
silicon, module, and equipment vendors, time will tell whether LTE IoT lives up to its promises.

LEADING LPWA TECHNOLOGIES


RPMA
RPMA is a technology from Ingenu that operates in the global, license-free 2.4 GHz band and is a technology
designed to offer a secure, large coverage footprint featuring extreme capacity efficiency for unscheduled
small data payloads. The technology makes use of direct sequence spread spectrum modulation (DSSS) and
the channel bandwidth is 1 MHz. With a 1 MHz buffer between channels, the system can support 40 simultaneous channels. Ingenu also uses additional proprietary algorithms to further guarantee message arrival
at up to 50% packet error rate (PER), which means that even if half the message packets drop out, the entire
message can still be decoded. RPMA also uses transmit power control so that as more access points and end
points are added to the network, they can adjust accordingly for network scalability.

SIGFOX
SIGFOX is an ultra-narrowband (UNB) LPWA radio technology operating in the sub 1 GHz ISM bands. The
narrow (100 Hz) bandwidth allows the technology to transmit over very long ranges. The SIGFOX system

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

sends very small amounts of data (up to 12 bytes per message and up to 140 messages per day) and sends
that data very slowly at 100 bps using BPSK. As a result, this low-duty cycle and very slow data transmission
is capable of being detected over long ranges and can reach underground installations and buried objects.
The SIGFOX technology is primarily uplink, but it does offer an acknowledgment or response scheme that
allows a device that has sent a message to receive a response.

LoRa
LoRa (short for long range), like RPMA, is a proprietary spread spectrum RF technology from the LoRa Alliance that offers low-cost, low-power transceivers for long-range wireless connectivity. The LoRa RF interface
also makes use of the license-free ISM bands, and according to the LoRa Alliance, it can also be adapted to
other frequencies. LoRa uses chirp spread spectrum (CSS), which is one form of spread spectrum modulation and which allows LoRa systems to resolve or demodulate signals that are up to 20 dB below the noise
floor for extended range.
A variety of bandwidths are available: 7.8 kHz, 10.4 kHz, 15.6 kHz, 20.8 kHz, 31.2 kHz, 41.7 kHz, 62.5 kHz, 125
kHz, 250 kHz, and 500 kHz. The required bandwidth can be selected according to the data requirements of
the application, as well as the link conditions. To conserve battery power data rates are also adaptive, using
an algorithm to determine optimum spreading factor depending on link conditions. LoRa has three classes
of edge node: one that opens a small downlink window after each upload (Class A), another that schedules
an uplink time slot (Class B), and one that listens for downlink messages occurring at any time (Class C). The
edge node class chosen has an impact on battery life, with Class C nodes consuming the most power and
Class A the least. Class A nodes are intended for battery-powered sensors with no constraints on latency,
Class B for battery-powered actuators, and Class C for AC mains-powered actuators.

LTE Cat-M1
LTE Cat-M1, also known as Cat-M, or eMTC, is a 3GPP Release 13 specification that reduces modem complexity and power when compared with a typical handset modem. Cat-M1 also increases link budget by approximately 15 dB, allowing the RF signal to penetrate further for enhanced coverage and giving better coverage
deep inside buildings, at the cell edge, or in subterranean locations. Cat-M1 implements half-duplex data
rates of a maximum of 1 Mbps in both uplink and downlink in a reduced channel bandwidth of 1.4 MHz.
Although 1.4 MHz is available, Cat-M1 uses only six out of the eight available 180 kHz LTE physical resource
blocks (PRBs), giving an actual bandwidth of 1.08 MHz. The remaining two PRBs are used as guard bands to
limit interference. The fact that the modem uses this narrowband channel gives a substantial reduction in
complexity to around 20% to 25% of that of a typical Cat-4 mobile broadband device, with consequent reductions in power and cost when compared to Cat-4 solutions. Also, the reduction in maximum transmit power
to 20 dBm is expected to lead to single-chip modem solutions, since the power amplifier can be integrated
with the modem for lower cost.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

For power saving, Cat-M1 uses power savings management (PSM), which allows the device to skip the regular
page monitoring cycles to increase the length of time the device is in sleep mode. The device is unreachable
when it is in sleep mode so PSM is best used for device-originated or scheduled data transfers such as found
in smart meters and sensor applications.
A second power-saving technique is also incorporated into Cat-M1. The extended discontinuous reception
(eDRX) technique optimizes battery life by extending the maximum time between data transmissions, allowing the device and the network to synchronize sleep periods. This maximum time can be up to 40 minutes
but with a consequent increase in latency. Since the network and device are able to synchronize during
sleep periods, eDRX is best used for device-terminated applications such as asset tracking and smart grid
applications.
With a maximum transmit power allowed of 20 dBm, the Cat-M1 link budget is approximately 156 dB, resulting in improved coverage versus typical M2M Cat-1. Configuring an LTE network for Cat-M1 involves a
straightforward software upgrade to most existing LTE radio interfaces with no new hardware or spectrum
required.

NB-IoT
NB-IoT, also called LTE Cat-NB1, is an evolution of LTE Cat-M1 and is also specified within 3GPP Release 13.
NB-IoT, like its GSM equivalent EC-GSM-IoT, uses an ultra-narrow 180 kHz channel (1 PRB) and supports
uplink data rates of up to 150 kbps. NB-IoT achieves a 23 dB improvement in link budget when compared
to standard LTE, which results in a link budget of 164 dB for maximum area coverage. NB-IoT also leverages
the PSM and eDRX features for power efficiency and implements new logical control and data channels for
low overhead data transmission.
NB-IoT is designed to be tightly integrated and interwork with LTE, and can be deployed in the LTE guard
band, embedded within a normal LTE carrier or as a standalone carrier in, for example, GSM bands. NB-IoT
reduces chip complexity levels below those of LTE-M1 to around 15% of a Cat-4 LTE modem and equivalent
complexity to EC-GSM-IoT. It is estimated that 85% of LTE infrastructure can be configured by a software
upgrade to the existing LTE radio interface; however, the LTE footprint is primarily restricted to largely urban
areas. The evolved packet core (EPC) requires the addition of the service capability exposure function (SCEF)
for NB-IoT; however, SCEF can be virtualized and is therefore also a software update.

EC-GSM-IoT
EC-GSM-IoT, or extended coverage GSM (often referred to EC-GSM), has also been specified in the 3GPP Release 13 specification. EC-GSM-IoT uses a narrow 180 kHz channel on existing GSM networks to enhance link
budget by an additional 20 dB compared to legacy GSM / GPRS technologies. EC-GSM-IoT can reuse exiting
2G chipsets with a firmware upgrade, so it will not require new chipset or module designs. EC-GSM-IoT was

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

designed as a high-capacity, long-range, low-energy, and low-complexity cellular system for IoT communications. Battery life of up to 10 years can be supported for a wide range of use cases.
The 20 dB coverage extension is achieved by utilizing the concept of repetitions and signal combining
techniques. EC-GSM-IoT also makes use of eDRX and PSM, and also implements new logical channels for
low overhead data transmission. Uplink data rate is up to 10 kbps, and using a maximum transmit power
level of 23 dBm, EC-GSM-IoT has a link budget of 164 dB for maximum coverage. These simplifications give
an overall level of complexity of roughly 15% when compared to a Cat-4 mobile broadband modem, and
configuring a GSM network for EC-GSM-IoT involves a straightforward software upgrade to the existing GSM
radio interface.
Since GSM technology is still the dominant cellular protocol in many markets, the vast majority of todays
cellular M2M applications use GPRS / EDGE for connectivity. As a result, GSM will continue to underpin global
cellular connectivity for the IoT well into the future. Recognizing this, the 3GPP created EC-GSM-IoT, which will
be of interest to those MNOs with a large M2M GSM installed base as are found in Europe; however, many
operators in North America and Asia are sunsetting their 2G networks to re-farm spectrum for LTE, with
other regions around the world to follow suit in the coming years.

COMPARING LPWA TECHNOLOGIES


SPECTRUM
An LPWA technologys attractiveness can be judged by its ability to be deployed anywhere worldwide without
modification. The choice of a frequency band available worldwide enables this. The global 2.4 GHz ISM band
is one such band and is commonly used for Wi-Fi, but has been adopted by Ingenu for RPMA. The 2.4 GHz
band does have the drawback of being unlicensed, so any radio communications in the band must tolerate
or mitigate interference. Also, since the path loss of the communications link increases with frequency, the
2.4 GHz band will have a shorter range than an equivalent transmitter in the sub 1 GHz bands. Any LPWA
technology must compensate for this by careful design of the radio.
To achieve long range, some LPWA technologies have adopted the sub 1 GHz bands. These bands are
highly fragmented, with typically the frequencies between 902 to 928 MHz, referred to as the 915 MHz band
offered in North America and Australia. The 433 MHz and 868 MHz bands are offered in Europe and
Australia and 780 MHz in Asia. Each regional authority imposes different specifications for the equipment
transmitting on the communications link. Factors such as output power, bandwidth, channel spacing, and
duty cycle must all be considered. This leads to region-specific hardware. SIGFOX and LoRa operate in the
sub 1 GHz bands. In particular, SIGFOX redesigned its originally European equipment for North American
compliance. This required a new architecture and may be the reason why SIGFOX LPWANs are deploying at
a slower rate in the United States. When compared to the ETSI 868 MHz band, the FCC 915 MHz band is also
subject to much higher levels of interference, which could be a drawback for technologies not designed with
interference in mind.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

The 3GPP licensed cellular bands are also highly fragmented and come with the added complexity that each
carrier must certify equipment for use on its frequencies and infrastructure. With more than 30 LTE bands
and, according to the GSA, 521 commercially launched LTE networks in 170 countries, equipment certification is an additional step that an LPWA vendor must take before equipment can be deployed on a carriers
network. In return for managing this complexity, cellular-based solutions offer a secure, highly available
network operating in clean or interference-free licensed spectrum around the clock for applications that can
scale from the very simple to the most demanding and in a short timeframe. Cellular-based solutions also
offer a standards-based mature ecosystem that guarantees interoperability across networks and devices.
LTE IoT removes complexity and cost by eliminating the complex multiband radio found in typical handsets,
using a single antenna and operating in half-duplex mode only. Also, it turns out that the band support for
an LTE IoT radio can be greatly simplified using, for example, band 12 (FDD 700 MHz), which is compatible
with AT&T, or band 13 (FDD 700 MHz) for Verizon in the United States. Band 20 (FDD 800 MHz) is popular
in Europe, the Middle East, and Africa (EMEA), and band 8 (FDD 900 MHz) is compatible with networks in
Europe and Japan.

COVERAGE
In an efficient LPWAN, each node and basestation covers the maximum distance possible so that infrastructure investment is minimized. This means that the range of the link must be as high as possible. Modern
LPWA links have between 140 dB to 168 dB link budgets (or maximum coupling loss), depending on technology, transmit power, receiver sensitivity, and other factors. Receiver sensitivity is a major contributor to link
budget and receiver sensitivities of better than -140 dBm are common in LPWA technologies, compared with
Receiver sensitivities of better than 140 dBm are common in LPWA
technologies, compared with -90 to
-110 dBm seen in many traditional
wireless technologies, such as found
in typical cellular handsets.

-90 to -110 dBm seen in many traditional wireless technologies, such as found in typical cellular handsets.
The high link budgets translate to many kilometers of range, since, for example, a receiver with a sensitivity
of -130 dBm can complete the link with a signal that is 10,000 times less than the weakest signal that can be
detected with a receiver sensitivity of -90 dBm. Ingenus RPMA supports an uplink receiver sensitivity of -145
dBm worldwide with no restrictions based on the regulations for the 2.4 GHz band. The equivalent number
for SIGFOX uplink sensitivity is between -134 dBm in the Americas and Australia regions and -142 dBm for
EMEA, while LoRa is between -132 dBm (Americas and Australia) and -137 dBm (EMEA). The RPMA receiver
is a factor of 13X to 20X more sensitive than SIGFOX or LoRa in the Americas and a factor of 2X to 13X
more sensitive than SIGFOX or LoRa in EMEA. This is one of the factors contributing to RPMAs large
coverage footprint.
According to Ingenu, RPMAs link budget can reach 177 dB in the United States and 168 dB in EMEA, which
allows it to cover an area up to 176 square miles for urban settings in the United States, where antenna gains
can be higher than in Europe where urban coverage is 33 square miles. By comparison, SIGFOX offers a link
budget of 149 dB, while LoRa offers a 157 dB link budget.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

LTE Cat-M1 increases link budget to 155 dB, which is an increase of 15 dB when compared to the LTE Cat-1
M2M technology, and both NB-IoT and EC-GSM-IoT have a link budget of 164 dB. This is a 15 to 20 dB higher
link budget, allowing the RF signal to penetrate deeper than the M2M-based LTE Cat-1 for enhanced area
coverage and deep in-building, subterranean, and cell-edge penetration.

CAPACITY AND SCALABILITY


In addition to coverage, the capacity of the network is a key performance metric and both coverage and
capacity play a major role in the economics of the LPWAN. It makes a poor business model for the LPWAN if
the network has great coverage but insufficient capacity to serve every end point in the network. Conversely,
it also makes for poor economics if the network has great capacity but insufficient range or penetration to
make the best use of it. So, an LPWA technology must carefully balance the tradeoffs between coverage and
capacity for best economics or lowest cost for the LPWAN.
With 1,000 simultaneous demodulations possible in RPMA, unlike LoRa or SIGFOX, which use an Aloha
protocol for channel access and repeat transmissions to ensure minimum bit and packet errors, and based
on total application throughput, RPMA on a single frequency on a single sector can demodulate thousands
of data packets per hour, a value which is many times superior to SIGFOX or LoRa. The ability of the LPWAN
to scale capacity is also a key success factor. In a successful LPWAN, if the original capacity is exceeded,
additional basestations must be deployed to scale up the capacity. Transmit power control is the feature
that allows for end points better covered by additional infrastructure to connect to the added basestation
and not continue to load capacity on the original basestation. This feature is very similar to the 3GPP LPWA
protocols and allows RPMA to easily scale up.
Since LTE IoT is an overlay to pre-existing
Since LTE IoT is an overlay to
pre-existing LTE networks without
requiring additional infrastructure,
its potential coverage is worldwide
on 512 commercially launched LTE
networks in 170 countries today

LTE networks without requiring additional


infrastructure,

its

potential

coverage

is

worldwide on 512 commercially launched LTE


networks in 170 countries today, depending
on the age of the MNOs equipment. ABI
Research expects that most LTE MNOs will
have equipment that is compatible with LTE
IoT. As a result, LTE IoT can easily scale up to
cover the majority of an MNOs network with
a straightforward software update to the
infrastructure equipment. LTE IoT can also
easily scale up to support high-throughput
use cases and scale down to support
low-performance

applications

using

the

same infrastructure.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

BATTERY LIFE
Many use cases in the IoT require battery life in excess of 10 years on a single cell, so device power
consumption becomes a key success factor for LPWANs. In massive deployments of low-cost devices, it is
cost-prohibitive to locate and change batteries. In LTE IoT, reduced modem complexity and new low-power
enhancements are implemented to reduce power consumption. Compared to LTE Cat-1, the Release 13
LTE-Cat-M1 reduces complexity by adopting a single antenna instead of the two antennas for a typical LTE
Cat-1 2 x 2 MIMO device. This leads to simpler architectures for the modem, including reduced layer 2 buffer
size, which reduces both modem power and cost. Also, by specifying a low-bit-rate half-duplex operation,
the modem design can eliminate the duplex filter and use a single RF chain for transmit and receive, and use
less processing power and memory. The fact that the modem uses a 1.08 MHz narrowband channel also
gives a substantial reduction in complexity, all of which results in a design-complexity reduction to around
20% to 25% when compared to a Cat-4 mobile broadband modem, with consequent reductions in power
and cost. Similarly, with its reduced data rate, NB-IoT reduces device complexity levels below those of LTE
Cat-M1 to around 15% of a Cat-4 LTE modem and equivalent complexity to EC-GSM-IoT. Also, the optional
maximum transmit power reduction to 20 dBm is expected to lead to single-chip modem solutions, since
the power amplifier can be integrated with the modem for lower cost. With PSM and eDRX, LTE IoT technologies reduce power with reduced paging cycles and reduced wake time of the end device, further improving
power efficiency.
Battery life is also a function of coverage. Increased coverage is obtained by lowering the data rate so that the
transmit power is shared among fewer bits and coverage is extended. However, for a given message size, the
transmit time can become very long, drawing significant power when compared to technologies with higher
bit rates. For example, with SIGFOX, its slow transmission rate requires long transmission times that draw
significant power when compared to other LPWA technologies that can transmit an equivalent message in a
shorter time thus lowering the transmit duty cycle and saving power.
RPMA was also designed to maximize battery life by minimizing overhead or control and signaling data.
This is a key consideration in LTE IoT since LPWA data payload is low. While both Cat-M1 and NB-IoT can
be deployed in existing infrastructure and spectrum, Cat-M1 devices will leverage legacy LTE synchronization signals while introducing new control and data channels that are more efficient for low-bandwidth operations. NB-IoT takes this one step further and improves efficiency by adopting a new set of narrowband
control and data channels to reduce overhead and signaling. Also, by maximizing data rate to a relatively high
level, transmit times are minimized. By contrast, LoRa and NB-IoT cannot adequately optimize modulation
mode and the SIGFOX technology uses a fixed data rate. The net result is that power consumption characterization is something that can only be done within the framework of the actual use case after the system
is deployed in real-world scenarios.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

TWO-WAY
DATA

INTERFERENCE
ROBUSTNESS

LOCATION
SUPPORT

MOBILITY
SUPPORT

AUTHENTICATION
AND SECURITY

MODULE
COST

TIME TO
MARKET

ADDITIONAL KEY SUCCESS FACTORS


Interference Robustness
Interference robustness is an important factor particularly for technologies operating in unlicensed
spectrum. Failure to take this into consideration can reduce the link budget considerably. In unlicensed
systems, shared channel operation can result in mutual interference and blocking of the uplink channel
unless the interfering system has very few simultaneous users. Successful coexistence schemes will be
required as unlicensed systems proliferate. However, licensed 3GPP LPWA technologies are not subject to
the same limitations, since channel access is scheduled.
RPMA DSSS spreading is highly resistant to the type of interference found in the 2.4 GHz ISM band since the
media access control (MAC), which was designed from the ground up, is highly tolerant to packet loss. RPMAs
proprietary network planning techniques also de-rate coverage based on predicted, and then measured,
interference so coverage tends to be conservative.

Two-way Data
Two-way data is also an important feature. All LPWA systems uplink data but downlink is important since
throughout the (10+ year) life of the end node, a firmware update for system upgrade or security patch
may be required, and without sufficient downlink capacity, this may not be possible. Both the SIGFOX
technology and LoRa technologies are restricted to a 1% duty cycle that makes any downlink very time
consuming for large payloads. In addition, LoRas bidirectional capability depends on the edge node class,
with Class A opening a small downlink window after each upload, Class B scheduling an uplink time slot, and
Class C listening for downlink messages occurring at any time.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

Broadcast is a closely related feature and is uniquely supported by RPMA, which can broadcast a
firmware image to all of the devices that need it at the same time. Cellular LPWA solutions do not support
this capability in Release 13, but this feature is under consideration for Release 14. In addition to this, the
broadcast feature becomes useful for certain use cases, such as simultaneously turning on (or off) a large
number of street lights or for utility load shedding during periods of peak demand.

Authentication and Security


Authentication and security is critical in IoT systems since rogue data can have catastrophic consequences
for a LPWAN. The industry standard 128-bit Advanced Encryption Standard (AES-128) has been defined
by the National Institute of Standards and Technology for applications until 2030. This means that 128 additional bits or 16 bytes must be transmitted for a sufficiently strong digital signature. RPMA has been designed
to accommodate this with a flexible packet size in such a way that it has minimal impact to the system.
SIGFOX offers device authentication, but the protocol does not natively encrypt the data. Since SIGFOX is
constrained to a maximum transmit duration based on its spectrum selections, its frame structure cannot
support a 16 byte signature. A hash is sent with each message that, along with a 16-bit private key specific
to the device, provides authentication. It is left to the user to encrypt the data (within the 12-byte payload) to
avoid being vulnerable to so-called man in the middle attacks, for example. LoRa implements authentication using 32-bit keys.
LTE IoT leverages the well-established 3GPP authentication and encryption standards and uses 128- to 256bit keys with DES, 3DES, and IDEA encryption algorithms for end-to-end security. The embedded SIM card
or eSIM standard may be adopted in some LTE IoT applications to simplify subscription management and
provisioning for end nodes. Using eSIM embedded during manufacture, the subscription credentials for
the mobile network can be downloaded or provisioned remotely, thus greatly simplifying end-node
management. The eSIM chip will be part of the radio module and as such will add cost.

Location Support
LoRa is the only LPWA technology to natively support geolocation; all other technologies rely on GPS or
LoRa is the only LPWA technology to
natively support geolocation

other positioning technology, which requires additional electronics in the module and adds cost. LoRa uses
a non-GPS geolocation technology to calculate position. Implemented in the LoRa gateway, it is based on an
ultra-high-resolution timestamp added to each received LoRa data packet. Each basestation reports the time
of arrival and uses differential time of arrival (DTOA) algorithms to determine the end-node position. LoRa
geolocation has a major cost and power benefit to the end node because it does not require any additional
hardware. However, given that signals will be received from multipath reflections as well as the direct path,
this type of geolocation may not be as accurate as GPS-based systems.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

10

Mobility Support
Mobility support may also be required by some applications such as asset tracking where the end node is
expected to move. Even if the end node is stationary, it may be moved for a variety of reasons. However,
IoT devices do not require the rapid cross-cell mobility that a fast-moving cellular terminal or handset would
require. In LoRa, all basestations collaborate and receive data from any device in range, so no handover
procedure, which can consume additional power, is required between network cells.
In RPMA, a device makes an initial choice of access point on power-up and then joins that access point. If the
link is lost (for example, because the node moved), then the node exits and performs a scan to find a new
access point. For improved capacity, RPMA also features a second mechanism called background re-scan,
which has the ability to periodically monitor the channel for access point signal strength. If another access
point has better signal strength, then a handover to that access point will be triggered. SIGFOX makes no
provision to support mobility.
LTE Cat-M1 will support full mobility in both connected and idle modes by leveraging the existing handover
mechanisms. However, NB-IoT and EC-GSM-IoT support limited mobility since there is no link measurement
or reporting. The degree of mobility support is determined by the eDRX and PSM settings. In eDRX cell
reselection may be delayed due to the increased latency as link measurements were paused during sleep
mode. When PSM is implemented, there is no mobility support and the radio node only selects a cell when
leaving PSM.

Module Cost
With legacy M2M modules costing around US$15 today, it is an expensive proposition to deploy many thouNB-IoT reduces device complexity
levels below those of LTE Cat-M1
to around 15% of a Cat-4 LTE
modem and equivalent complexity
to EC-GSM-IoT

sands of end points over a wide area. LPWA vendors simplify their solutions as a way to reduce costs and
all target sub US$10 or lower. Current LTE modules have separate ICs for power management, baseband,
memory, RFIC, and power amplifier functions and contain discrete components for filtering and switching.
For LTE IoT, reduced complexity is obtained by using a single antenna instead of two, thus only one RF chain
is required, and half-duplex operation eliminates the duplex filter. The lower bit rate half-duplex operation
reduces the memory requirements for the modem, and PSM and eDRX eliminate the need for a separate
power-management chip. The option of 20 dBm transmit power (instead of 23 dBm) will permit the integration of the PA into a single chip SoC, further reducing cost, all of which results in a design-complexity
reduction for Cat-M1 to around 20% to 25% when compared to a Cat-4 mobile broadband modem. With
its reduced data rate, NB-IoT reduces device complexity levels below those of LTE Cat-M1 to around 15% of
a Cat-4 LTE modem and equivalent complexity to EC-GSM-IoT.
However, complexity reduction may not directly translate to cost reduction since in the advanced semiconductor technology nodes that will be used for LTE IoT SoCs, the designs may well be pad limited to a fixed
minimum die area and cost. As a result, some silicon vendors are including both LTE Cat-M1 and NB-IoT in

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

11

the LPWA SoC, the selection between LPWA technologies being made by software. For the purposes of this
white paper, we have estimated the module costs for each of RPMA, SIGFOX, and LoRa and LTE IoT based on
our estimates of complexity.

Time to Market
RPMA, SIGFOX, and LoRa are all available now with a time-to-market advantage and are deploying in multiple
markets, while LTE IoT Cat-M1 and NB-IoT will be available in pre-commercial trial networks in 4Q 2016, with
the first commercial implementations as early as 1H 2017. All of these features are summarized in Table 1
below and highlight for each technology its compatibility with the performance goal.

Table 1
Sigfox

LoRa

EC-GSM-IoT

NB-IoT

LTE Cat-M1

RPMA

Bandwidth

100Hz

125kHz

600khz

180kHz

1.08MHz

1MHz

Coverage

149dB

157dB

164dB

164dB

160dB

177dB

Capacity*

<25,000/cell

<40,000/cell

<190,000/cell

<200,000/cell

<1M/cell

<500,000/cell

Battery Life

10 Years+

10 Years+

10 Years+

10 Years+

10 Years+

10 Years+

Throughput

100bps

290bps - 50kbps

473kbps

250kbps

1Mbps

624kbps

2-Way Data

No

Class dependent

Yes

Yes

Yes

Yes

Security

16bit

32bit

3GPP (128-256bit)

3GPP (128-256bit)

3GPP (128-256bit)

AES 128bit

Scalability

Low

Medium

High

High

High

High

Mobility Support

No

Yes

Idle mode

Idle mode

Connected+Idle mode

Yes

Location Support

No

Yes

Needs GPS

Needs GPS

Needs GPS

Needs GPS

$2.00

$12.00

$5.00

$5.00

$10.00

$12.00

Node Cost
* Note: Depends on usage profile

Above average

Average

Below average

TECHNOLOGY BEST FIT


OVERVIEW
The potential applications for an LPWAN run into the tens of thousands or more with the requirements for
each application varying across a diverse range. Requirements such as bandwidth, coverage, capacity, battery life, security, two-way data, low cost, mobility support, and location support all assume a greater or lesser
degree of importance depending on the use case and application.
Some end nodes will only be required to uplink a few messages per day such as required to monitor temperature, while others are required to uplink high-throughput data streams as might be required in industrial process control. Also, mobility and location support is required for vehicle and asset tracking and fleet
management, for example.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

12

Many high-performance applications will require two-way data or both uplink and downlink so that the LPWAN can also control equipment as well as monitor it, such as may be found in smart buildings for building
automation systems. After sensing a parameter like temperature, the data is uplinked to the gateway, network operations center, system controller, or the cloud where an action is computed based on that value.
For example, in an HVAC system, the action may be to turn on the air conditioning to lower the temperature.
The network operations center must then downlink that command over the LPWAN to actuate or turn on
the air conditioning. These higher-value systems follow a sense-compute-actuate paradigm and generally
require higher levels of performance than simple environmental sensing and monitoring applications.
Downlink capability is also important to prolong the useful life of an end node by downloading over the air
firmware updates to the latest specifications or for security patches. Security is also a very important parameter and some applications may demand the highest levels, such as may be required in a nuclear power
station or critical infrastructure, for example. The consequences of a security breach in such applications
could be disastrous.

USE CASE SEGMENTATION


For the purposes of this white paper, ABI Research has grouped applications into four generic use cases:
transport and logistics, smart cities, smart buildings, and industrial.
The transport and logistics use case includes applications such as fleet management, logistics, and tracking. This is a use case where mobility and location support are required. Coverage and long battery life will
also be highly desirable. However, the bidirectional data requirement for downlink data is not often required
since the end-node data traffic is uplink dominated.
The smart cities use case includes applications for parking sensors, smart lighting, traffic management,
waste management, and environmental monitoring. In this use case, coverage and range are important as is
location support. However, mobility support is not required because these LPWANs are mostly static, operating parking sensors or traffic and street lights, for example. A downlink capability will be relatively important
in some applications, such as operating street lights, for example.
ABI Research defines the smart buildings use case to include applications ranging from smoke detectors
and alarm systems to building management and automation. Some applications in smart buildings will require high throughput as in video surveillance, however, others, such as smoke detectors, will not. Security
of the data is important since many applications involve alarm systems. Two-way communication is also
required for building automation.
The industrial use case includes applications such as process monitoring and control, maintenance monitoring, smart metering, and smart grid. In this use case, capacity, bandwidth, security, and two-way communications are all required. However, mobility and location awareness are not normally needed.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

13

With these broad definitions, we can distinguish between each use case and map the LPWA technologies to
each of the requirements and assign a probability that the technology will meet or exceed the requirement,
meet the requirement in some cases, or not meet the requirement. The results of this exercise are illustrated
in Table 2 below. Although we have not included every possible criterion for selection of an LPWAN technology, we see that in each of our four use cases some technologies are more of a best fit than others.

Table 2
Transport and Logistics
Bandwidth

Requirement

Sigfox

LoRa

EC-GSM-IoT

NB-IoT

LTE Cat-M

RPMA

Medium

Smart Cities

Requirement

Bandwidth

Low
High

Coverage

High

Coverage

Capacity

Medium

Capacity

Battery Life
Security
2-Way Data
Low Cost

High
Medium
Low
Medium

Medium

Security

Medium

2-Way Data

Medium

Low Cost

Medium

High

Mobility Support

Low

Location Support

High

Location Support

High

Smart Buildings

Requirement

Sigfox

High

LoRa

EC-GSM-IoT

NB-IoT

LTE Cat-M

RPMA

Industrial

Requirement

Bandwidth

High

Coverage

Medium

Coverage

Low

Capacity

High

Capacity

High

Battery Life

Medium

LoRa

EC-GSM-IoT

NB-IoT

LTE Cat-M

RPMA

Sigfox

LoRa

EC-GSM-IoT

NB-IoT

LTE Cat-M

RPMA

Low

Battery Life

Mobility Support

Bandwidth

Sigfox

Battery Life

Low

Security

High

Security

High

2-Way Data

High

2-Way Data

High

Low Cost

High

Low Cost

High

Mobility Support

Low

Mobility Support

Low

Location Support

Low

Location Support

Low

Compatibility
Meets or exceeds

Use case dependent

No match

In the transportation and logistics use case, RPMA meets or exceeds our requirements seven times out of
nine, followed closely by LTE Cat-M1. Other technologies may be a good fit, but the lack of location support
without external GPS technology coupled with limited mobility support is a drawback. LoRa does offer native location support, but its coverage and capacity performance may not be compatible with this use case.
Depending on the type of mobility support required in the application, i.e., sporadic, nomadic, or full mobility,
EC-GSM-IoT and NB-IoT may offer adequate performance.
In the smart cities use case, again RPMA stands out by meeting or exceeding requirements seven times out
In the smart cities use case, again
RPMA stands out by meeting or
exceeding requirements seven times
out of nine and matching the
performance of the LTE IoT
technologies

of nine and matching the performance of the LTE IoT technologies. The lack of embedded location support
is a limitation for RPMA and LTE IoT. LoRa and SIGFOX closely follow LTE IoT in matching requirements, with
LoRa support for location a stand-out feature. Depending on the application, security may well be the deciding factor, with LoRa offering better performance.
The smart buildings use case has both RPMA and LTE Cat-M1 offering the best fit, with both technologies
meeting or exceeding requirements seven times out of nine. EC-GSM-IoT and NB-IoT may offer a good fit for
the application, but do not have the throughput for some applications.

www.abiresearch.com

BEST FIT USE CASES FOR LPWANs

14

In the industrial use case, RPMA again stands out seven times out of nine and is on par with LTE Cat-M1.
In the industrial use case, RPMA
again stands out seven times out of
nine and is on par with LTE Cat-M1

The other LTE IoT technologies are close seconds and are distinguished by their reduced throughput.
SIGFOX and LoRa may not have the two-way data performance necessary for industrial applications;
however, with LoRa, the proper selection of edge class can be used to adapt its performance to match the
two-way requirements.

SUMMARY AND CONCLUSIONS


The selection of an LPWA technology is a complex process involving many factors ranging from the technical
RPMA technology has the broadest
appeal across diverse use cases and
bridges the gap between low-cost
LPWA for simple applications to
what is anticipated to be highperformance LTE IoT

aspects of the protocol and radio link, which we have discussed in this white paper, to the economics, the IoT
platform and analytics, silicon vendor support, and roadmap evolution. Nevertheless, based on our analysis,
RPMA technology has the broadest appeal across diverse use cases and bridges the gap between low-cost
LPWA for simple applications to what is anticipated to be high-performance LTE IoT. ABI Research believes
that RPMA offers a carrier-grade LPWAN on par with LTE IoT. RPMA was built from the ground up to target
the industrial use case, but we believe it has proven a capability to tackle applications in the smart cities,
smart buildings, and transport and logistics use cases.
We also note that although the specifications for LTE IoT have been finalized, the industry is performing tri-

LTE IoT is targeted to launch


commercially in the first half of 2017
by MNOs including AT&T, Orange,
and Vodafone

www.abiresearch.com

als on pilot networks now and through 4Q 2016. LTE IoT is targeted to launch commercially in the first half
of 2017 by MNOs including AT&T, Orange, and Vodafone. With significant support from MNOs and silicon,
module, and equipment vendors, time will tell whether LTE IoT lives up to its promises.

BEST FIT USE CASES FOR LPWANs

15

Published August 2016


2016 ABI Research
Post Office Box 452 249 South Street
Oyster Bay, New York 11771 USA
Tel: +1 516-624-2500 | Fax: +1 516-624-2501

www.abiresearch.com

2016 ABI Research. Used by permission. Disclaimer: Permission granted to reference, reprint or reissue ABI Research products is expressly
not an endorsement of any kind for any company, product, or strategy. ABI Research is an independent producer of market analysis and insight and this
ABI Research product is the result of objective research by ABI Research staff at the time of data collection. ABI Research was not compensated in any
way to produce this information and the opinions of ABI Research or its analysts on any subject are continually revised based on the most current data
available. The information contained herein has been obtained from sources believed to be reliable. ABI Research disclaims all warranties, express or
implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.

www.abiresearch.com

Você também pode gostar