SlideShare a Scribd company logo
How To Triple The Range of LoRa
© 2019 Haystack Technologies, Inc.
2
Executive Summary
• Haystack is introducing a new DASH7 implementation for LoRa: XR Mode.
• Using software-only enhancements, DASH7 XR Mode for LoRa provides 2-3x range
improvements over comparable observed LoRaWAN performance at ground level while
preserving multiple year endpoint battery life.
• XR Mode utilizes 2-way LDPC Error Correction as well as Automated Receipt Response for
guaranteed message delivery. Shorter packet sizes and lower data rates than are typically
deployed with DASH7 are also utilized.
• XR Mode running at a higher data rate can achieve the same QoS as LoRaWAN running at a
much lower data rate. This translates to 10-20x the network density and 10-20x the battery
life vs. LoRaWAN, for a given QoS.
• Private beta testing for XR Mode is now underway.
3
Background
• Some LoRa range test results make aggressive claims (“Up to 30 miles!”) which
often do not match real world customer results. These marketing claims are
typically based on “best case” outdoor testing environments and may result in
unhappy customers and harm long-term adoption.
• In comparison, testing LoRa at ground level yields very different results from
testing conducted in such ideal outdoor environments. Yet some Haystack
customer requirements include both gateways and endpoints deployed at or
near ground level.
• To improve range performance at all levels, but in particular at ground level, we
spent the last six months optimizing our OpenTag firmware stack for extended
range and reduced packet error rates over LoRa. The result: DASH7 XR Mode.
DASH7™
4
Technical Features
‣ 2-way LPWAN networking software

‣ Works across many different LPWAN radios

‣ Ultra-low power (years on coin cell)

‣ Low latency (<2 sec.) queries

‣ Broadcast, multicast, P2P messaging

‣ Two-way error correction

‣ Message delivery confirmation

‣ Real-time GPS, A-GPS support

‣ Indoor location via RSSI, TDOA, others

‣ AES128 private-key + public-key crypto

‣ OTA firmware updates, key refreshes

‣ Smart contract support at endpoint 

‣ Invented by Haystack!
OSI Layer
7 Application UDP + OIC + NDEF + iOS, etc.
6 Presentation
DASH7 Core

low power
low latency
low cost
5 Session
4 Transport
3 Network
2 Data Link
1 Physical (Radio) LoRa, Other Options
DASH7: The Most Advanced
Networking Stack for Low Power
Wide Area Networks
LoRa®
5
For more information on LoRa:
https://www.semtech.com/lora/
LoRa is a radio, not a protocol.
To exploit its true potential and performance,
it requires an advanced protocol stack
Basic Attributes
‣ Popular low power, wide area networking radio
from Semtech

‣ Promises multi-year battery life and multi-
kilometer range

‣ Mostly intended for use in unlicensed spectrum
(915 MHz in USA)

‣ Low cost (~$3 in volume today)

‣ Modules available from Murata, Microchip, ST
Micro, others
6
Today: Two DASH7 Options For LoRa
Option 1: DASH7 LAN Mode
Optimal for usage with most software & business logic. 

Includes fully bidirectional data transfers, point-to-point,
broadcasting, and real-time (asynchronous) communication.
Typ. Ground-to-Ground Range: Light-Urban 0.20 miles / 0.33 km
Typical Data Rate 16 kbps
Typical “Real World” Link Budget 130 dB
Endpoint Default Operation Event-driven (asynchronous)
LDPC Error Correction (Coming Soon)
Message Delivery Confirmation Yes
Send Commands, Queries To Endpoint Yes
Real-Time Queries to Endpoint Yes
Packet Size Variable: up to 256 bytes
Multi-Year Battery Life Yes
GPS-based Geolocation (Low Power) Yes: OTA A-GPS
Geofencing Real-Time, Low Power
Broadcast, Multicast Messaging Yes
OTA Firmware Updates Yes
0.64 miles / 1 km Typ. Ground-to-Ground Range: Light-Urban
1.1 kbps Typical Data Rate
153 dB Typical “Real World” Link Budget
Periodic Msg (synchronous) Endpoint Default Operation
Yes LDPC Error Correction
Yes Message Delivery Confirmation
Limited Functionality Send Commands, Queries To Endpoint
No Real-Time Queries to Endpoint
Fixed: 16 bytes Packet Size
Yes Multi-Year Battery Life
Yes: Preloaded A-GPS GPS-based Geolocation (Low Power)
Beacon Geofencing
No Broadcast, Multicast Messaging
No OTA Firmware Updates
Option 2: DASH7 XR Mode
Optimized for delivering LPWAN pub-sub messaging at
maximum theoretical range.
7
Today: Two DASH7 Options For LoRa
Option 1: DASH7 LAN Mode
Optimal for usage with most software & business logic. 

Includes fully bidirectional data transfers, point-to-point,
broadcasting, and real-time (asynchronous) communication.
Typ. Ground-to-Ground Range: Light-Urban 0.20 miles / 0.33 km
Typical Data Rate 16 kbps
Typical “Real World” Link Budget 130 dB
Endpoint Default Operation Event-driven (asynchronous)
LDPC Error Correction (Coming Soon)
Message Delivery Confirmation Yes
Send Commands, Queries To Endpoint Yes
Real-Time Queries to Endpoint Yes
Packet Size Variable: up to 256 bytes
Multi-Year Battery Life Yes
GPS-based Geolocation (Low Power) Yes: OTA A-GPS
Geofencing Real-Time, Low Power
Broadcast, Multicast Messaging Yes
OTA Firmware Updates Yes
0.64 miles / 1 km Typ. Ground-to-Ground Range: Light-Urban
1.1 kbps Typical Data Rate
153 dB Typical “Real World” Link Budget
Periodic Msg (synchronous) Endpoint Default Operation
Yes LDPC Error Correction
Yes Message Delivery Confirmation
Limited Functionality Send Commands, Queries To Endpoint
No Real-Time Queries to Endpoint
Fixed: 16 bytes Packet Size
Yes Multi-Year Battery Life
Yes: Preloaded A-GPS GPS-based Geolocation (Low Power)
Beacon Geofencing
No Broadcast, Multicast Messaging
No OTA Firmware Updates
Option 2: DASH7 XR Mode
Optimized for delivering LPWAN pub-sub messaging at
maximum theoretical range.
DASH7 XR Mode for LoRa
• XR: “eXtended Range” DASH7 protocol
• Runs concurrently with DASH7 LAN Mode
• Improves real-world signal propagation over
LoRaWAN by 10-13 dB (10-20x)
• Retains essential features like real-time GPS
location and multi-year battery life
• Software-based approach:

No HW or antenna modifications required
• XR Mode delivers the longest range
connectivity for battery-powered LoRa
devices in the world today
8
Observed Sensitivity @ 1kbps
110
117.5
125
132.5
140
LoRaWAN DASH7 LAN DASH7 XR
-
-
-
-
-
dBm
How is this gain possible? 

XR Mode algorithms come from very
recent technology for communicating
with probes in deep space.
XR Mode: Shorter Packet Lengths
9
• We reconfigured DASH7 frame
from the conventional, variable size
(up to 256 bytes), to a fixed size of
16 bytes.
• Most LPWANs use small packets,
so we picked a size that gives the
best efficiency for our deep space
error correction algorithm.
• The packet is small, but it is
sufficient for geolocation or sensor
data, and it can support strong
cryptography.
1 2 3 4 65 8 9 10 11 12 13 14 15
Link
Info
Subnet
Info
Tokenized ID
Encrypted
Forward
CRC8
XR Mode 16 Byte Frame
7 16
Data

ID
Data Payload Data

CRC8
XR Mode: Lower Data Rate, Greater Range
10
• Basic Truth 1: lower data rate = more energy per bit = longer range
• Basic Truth 2: lower data rate = longer transmission = device drains more power
• We observed that roughly 1 kbps is the best data rate trade-off (using LoRa) for achieving
maximum range in an LPWAN device with multi-year battery requirement.
‣ DASH7 LAN Mode uses 16 kbps (LoRa SF7, 500 kHz)
‣ XR Mode uses 1.1 kbps (LoRa SF11, 500 kHz, + Haystack LDPC encoding)
• We observed a range improvement of XR Mode vs LAN Mode
‣ 3x: Ground-to-Ground with obstacles (extreme non-line-of-sight)
‣ 6x: Elevated antenna to ground endpoint, with obstacles (moderate non-line-of-sight)
‣ >20x: Line of Sight
XR Mode: More Efficient than LoRaWAN
11
• Basic Truth 1: lower data rate = more energy per bit = longer range
• Basic Truth 2: lower data rate = longer transmission = device drains
more power
• In order to determine energy use to send a message, we need to
establish a baseline QoS. In this case, it is 1% PER (packet loss) at 

146 dB link budget.
• We observed that XR Mode performs significantly better than LoRaWAN
at any data rate. As result, if we set equal QoS, XR Mode is 22 times
more energy efficient than the longest-range LoRaWAN Mode.
‣ “LoRaWAN DR0” | 150 bps | -126 dBm @ PER 30% | 811 mJ per msg
‣ XR Mode | 1074 bps | - 132 dBm @ PER 1% | 36.5 mJ per msg
TPKT Packet Time-on-Air
PTX TX Power
PRX RX Power
XC Cumulative probability of error
EMSG Energy of message
EMSG = TPKT (PTX + PRX) * XC
TPKT 1254 ms
PTX 412.5 mW (@ 20 dBm)
PRX 42.9 mW
XC 1.42
EMSG 811 mJ
LoRaWAN DR0
TPKT 194 ms
PTX 145.2 mW (@ 14 dBm)
PRX 42.9 mW
XC 1.0
EMSG 36.5 mJ
DASH7 XR Mode
12
The greatest advancement
DASH7 XR Mode brings is its

error correction technology.
1. Forward Error Correction (FEC), where sender encodes the data using an error-
correcting code prior to transmission. The additional information (redundancy) added
by the code is used by the receiver to recover the original data in the case of
corruption.
2. Automatic Repeat Request (ARQ), which uses acknowledgements (messages sent by
the receiver indicating that it has correctly received a packet) and timeouts (specified
periods of time allowed to elapse before an acknowledgment is to be received) to
achieve reliable data transmission over an unreliable service. If the sender does not
receive an acknowledgment before the timeout, it usually re-transmits the packet until
the sender receives an acknowledgment or exceeds a predefined number of
retransmissions.
Two Types of LPWAN Error Correction
13
LPWANs & Forward Error Correction
14
LPWANs operating in unlicensed radio spectrum contend with high levels of noise,
unpredictable network congestion, or ground-level gateway deployments where
terrain and other objects negatively impact range.
Packet Error Rate (PER)

Percentage of packets not successfully received by a
gateway.
If we can find a way to reduce PER, we can increase the
range and/or Quality of Service (QoS) of the network.
RangePER QoS=
Forward Error Correction (FEC)

FEC is a common remedy to high PER in virtually all
contemporary wireless networking technologies, however
it is strangely crude or non-existent in most LPWAN
standards.
LoRa HW Built-in Forward Error Correction
15
• The LoRa radio includes a forward error correction (FEC) feature in
the radio hardware.
• The downside is that it is an outdated implementation, and it actually
does more harm than good in most usages (including LoRaWAN)
• The technical details of why it’s a poor design are beyond the scope
of this presentation, but the basics are:
‣ It is 1950’s-era technology. Better technology is available.
‣ LoRaWAN and all known usages of LoRa, apart from DASH7, use
suboptimal configurations of the built-in error correction.
‣ The actual implementation in HW seems to have some bugs that
result in occasional false positives.
• We have replaced LoRa’s HW FEC with our software implementation
of the most advanced FEC technology known to man (as of 2019)
Range
PER
No FEC
LoRa HW FEC

(Best config)
DASH7 XR

FEC
LoRaWAN
Mobile LPWANs Require Better PER
16
Environmental

Variables
Terrain
Diversity
Multipath/
Fading
Distance from
gateway
Distance
traveled/day
Frequency of
Movement
Elevation
We can see that robust error correction not only
makes range greater, but it improves the shape of
the PER curve.
• For LPWANs of fixed-position endpoints, the
endpoints can be adjusted at time-of-deployment
to achieve acceptable PER.
• Mobile LPWAN solutions must contend with
changing or unpredictable variables. PER must be
acceptable for worst-case scenarios.
• If the environmental variables are unpredictable,
you want a PER curve that is as flat as possible.
Range
PER
LoRa HW FEC

(Best config)
DASH7 XR

FEC
LoRaWAN
Robust FEC flattens the PER
curve, making XR Mode more
reliable than LoRaWAN when the
environment is unpredictable
XR Mode Error Correction

DASH7 XR Mode uses the most advanced FEC in conjunction with ARQ
17
Automatic Repeat Requests (ARQ)
• Operates via DASH7 Networking Layer
‣ Devices know to retransmit incorrectly received
frames.
‣ Devices will automatically reduce transmission power
to suit the required link budget (saves energy and
creates less interference for others)
• DASH7 ARQ process operates until:
‣ Packet is corrected
‣ Time-out is reached
‣ Shannon limit reached
Forward Error Correction (FEC)
• Haystack proprietary implementation of CCSDS “Orange
Book” Non-binary LDPC coding
‣ Bidirectional: full decoder implemented on cheap ARM
Cortex M0+ endpoint, as well as gateway.
‣ Performance is very close to theoretical limit
(Shannon)
• Engineering challenge was to get it working in such a
RAM & CPU constrained environment.
• Similar tech is used to enable communication with space
probes out in deep space, and also in 5G Cellular.
18
Important Note:

DASH7 Error Correction is Fully Bidirectional
2Way
• FEC Encoders are relatively simple to implement. 

FEC Decoders are much harder to implement.
‣ For example, Sony Eltres implements uplink LDPC, but does not
implement a downlink to the endpoint.
‣ SigFox also has optional uplink FEC (Convolutional code)
• In order to have guaranteed message delivery, the gateway needs to be
able to send ARQ messages to the endpoint.
‣ Very important for broadcast-with-acknowledgement, which (for
example) Haystack uses to do real-time queries.
• If the endpoint cannot decode FEC, the gateway needs to compensate by
transmitting messages to endpoint with much greater power.
‣ In rare cases this is OK, but most of the time there are regulatory limits
that make it impossible for a gateway to compensate with enough
power to reach the endpoint.
19
We have begun range testing 

XR Mode in the SF Bay Area
and have some results to share
XR Mode Range Test #1
20
Hardware STMicro Discovery Kit (SX1276) with external dipole
TX Power 17 dBm
Environment Light-Urban (San Mateo, CA), mostly flat
Gateway Elevation 1 m / 3.3 feet
Endpoint Elevation 1 m / 3.3 feet
Results best range: 1 km / .64 miles
Gateway
(1m)
Endpoint

(1m)
Propagation Fading Profile for Test #1
< 1% LOS (line of sight)
74% NLOS (non-line of sight)
25% Obstructed
XR Mode Range Test #2
21
Hardware STMicro Discovery Kit (SX1276) with external dipole
TX Power 17 dBm
Environment Light-Urban (San Mateo, CA), subtle rolling terrain
Gateway Elevation 5 m / 16.4 feet
Endpoint Elevation 1 m / 3.3 feet
Results best range: 1.77 km / 1.1 mi
Gateway
(5m) Endpoint

(1m)
Propagation Fading Profile for Test #2
13% LOS (line of sight)
62% NLOS (non-line of sight)
25% Obstructed
XR Mode Range Test #3
22
Hardware STMicro Discovery Kit (SX1276) with external dipole
TX Power 17 dBm
Environment Hilltop, then over SF bay, then suburban
Gateway Elevation 360 m / 1181 feet
Endpoint Elevation 1 m / 3.3 feet
Results best range: 27.5 km / 17.03 miles
Gateway
(360m)
Endpoint

(1m)
Propagation Fading Profile for Test #3
80% LOS (line of sight)
10% NLOS (non-line of sight)
10% Obstructed
XR Mode Range Test #4
23
Hardware STMicro Discovery Kit (SX1276) with external dipole
TX Power 17 dBm
Environment Mountain and Valley
Gateway Elevation 1170 m / 3838 feet
Endpoint Elevation 1 m / 3.3 feet
Results best range: 57 km / 35.9 miles
Gateway
(1170m)
Endpoint

(1m)
Propagation Fading Profile for Test #4
93% LOS (line of sight)
6% NLOS (non-line of sight)
1% Obstructed
XR Mode Range Test Conclusions
24
Gateway
(1170m)
Endpoint

(1m)
Gateway
(360m)
Endpoint

(1m)
Gateway
(5m) Endpoint

(1m)
Gateway
(1m)
Endpoint

(1m)
1 km
1.77 km
27.5 km
57 km
• Increasing gateway elevation has a major
impact in improving range, because it increases
the amount of the link that is Line of Sight.
• LOS (Line of Sight) portions are shown in green.
• The ground itself is a major cause of signal
fading (attenuation). This is NLOS (Non-line-of-
Sight) and is shown in orange.
‣ Visible light (~566,000 GHz) is a much higher
frequency than LoRa (0.9 GHz).
‣ Lower frequencies can travel further, but they
have a larger wavefront that is more effected
by the ground.
• Obstructions cause even more attenuation of
the link. Obstructions are shown in red.
Antenna Design & Placement
• Gateway antenna placement makes a huge difference when optimizing for range.
• Nearly all LPWAN range tests with very long range claims (e.g., 10 miles+) use a
combination of:
1.High-gain antennas located very high above the ground,
2.Data rates that are too low to be suitable for low-power devices
3.Very sophisticated receiver hardware.
‣ Haystack XR Mode outperformed these systems despite being tested with higher data
rate, simple antennas, and low cost, low-power receivers
• Even a small increase in gateway height can make a significant difference in range
performance. In recent testing, the average difference between 2 meters and 5 meters was
nearly 200%.
• At these extreme sensitivities, it becomes very important to minimize electromagnetic
noise generated in the electronics. This is an engineering challenge. STMicro’s LoRa dev
kit, for example, is way too noisy. We needed to isolate the antenna in order to conduct
the tests.
25
Option A Option B
Guess which gateway option results in
better range?
Final Thoughts
• Haystack enables not just the longest range LoRa connectivity available today, but also the most
reliable and the lowest power.
• Few LPWAN networking stacks appear to have been designed for mobile use cases as most are
only found being applied to fixed use cases and leave developers to “hack” mobile solutions or
ignore them altogether. LoRa’s utility in mobile use cases is limited without robust two-way error
correction and other features like multicast. Haystack XR Mode is directly targeted to mobile
developers.
• Haystack is offering XR mode demo software to a limited number of beta testers. Register here:
http://bit.ly/XRModebeta.
• Contact us: info@haystacktechnologies.com
26
How To Triple The Range of LoRa

More Related Content

What's hot (20)

PDF
7. log distance and log normal shadowing
JAIGANESH SEKAR
 
PPT
STM32 MCU Family
Premier Farnell
 
PPTX
Vblast
Amal Shekar
 
PDF
vlsi projects using verilog code 2014-2015
E2MATRIX
 
PPTX
Mastering Microcontroller : TIMERS, PWM, CAN, RTC,LOW POWER
FastBit Embedded Brain Academy
 
PDF
Equalization in digital communication
Pei-Che Chang
 
PDF
Field Programmable Gate Array: Building Blocks and Interconnections
Dr. Saravanakumar Umathurai
 
PPTX
Radix-2 DIT FFT
Sarang Joshi
 
PDF
Water Level Monitoring System using IOT
IRJET Journal
 
PPTX
Microblaze
Krunal Siddhapathak
 
PPTX
Peripherals and interfacing
RAMPRAKASHT1
 
PPTX
Unit 2 sdr architecture
JAIGANESH SEKAR
 
PPTX
Types of AM Receiver
Waqar Ahmed
 
PPTX
Hspa and hsdpa
Manish Srivastava
 
PPTX
TMS320C6X Architecture
Shweta Tripathi
 
PPTX
Sample and hold circuit
Prathamesh Kolekar
 
PDF
DSP_FOEHU - Lec 07 - Digital Filters
Amr E. Mohamed
 
DOCX
Filtros pasivos de primer orden
Francesc Perez
 
PPTX
4G Handovers || LTE Handovers ||
ankur tomar
 
PPTX
Passive infrared based human detection alive robot
Sidharth Mohapatra
 
7. log distance and log normal shadowing
JAIGANESH SEKAR
 
STM32 MCU Family
Premier Farnell
 
Vblast
Amal Shekar
 
vlsi projects using verilog code 2014-2015
E2MATRIX
 
Mastering Microcontroller : TIMERS, PWM, CAN, RTC,LOW POWER
FastBit Embedded Brain Academy
 
Equalization in digital communication
Pei-Che Chang
 
Field Programmable Gate Array: Building Blocks and Interconnections
Dr. Saravanakumar Umathurai
 
Radix-2 DIT FFT
Sarang Joshi
 
Water Level Monitoring System using IOT
IRJET Journal
 
Peripherals and interfacing
RAMPRAKASHT1
 
Unit 2 sdr architecture
JAIGANESH SEKAR
 
Types of AM Receiver
Waqar Ahmed
 
Hspa and hsdpa
Manish Srivastava
 
TMS320C6X Architecture
Shweta Tripathi
 
Sample and hold circuit
Prathamesh Kolekar
 
DSP_FOEHU - Lec 07 - Digital Filters
Amr E. Mohamed
 
Filtros pasivos de primer orden
Francesc Perez
 
4G Handovers || LTE Handovers ||
ankur tomar
 
Passive infrared based human detection alive robot
Sidharth Mohapatra
 

Similar to How To Triple The Range of LoRa (20)

PDF
Haystack Technology Overview
Haystack Technologies
 
PDF
LoRaWAN vs Haystack
Haystack Technologies
 
PDF
Bringing Better Networking to LTE IoT
Haystack Technologies
 
PDF
Open fest2018 LoRa, LoRa WAN, The Thing Network - capabilities, myths, applic...
Olimex Bulgaria
 
PDF
Cellular lpwan paris nov 2015
robert huynh
 
PDF
LoRaWAN101_What is it
Birdz
 
PDF
LoRa General Presentation LoRa General Presentation.pdf
ssuser262297
 
PDF
LoRaPresentation20255555555555555555555555555555555555.pdf
WidedAhmadi
 
PPTX
LoRa General Presentation FOR MASSIVE DEPLOYMENT.pptx
WidedAhmadi
 
PPTX
LoRa General Presentation FOR MASSIVE DEPLOYMENT.pptx
WidedAhmadi
 
PDF
Haystack + DASH7 Security
Haystack Technologies
 
PPTX
Long-Range Low-power Radio (LoRa).pptx
Ajay Gangakhedkar
 
PDF
IoT_standards
João Santos
 
PDF
The IoT Hunger Games 2015
Haystack Technologies
 
PDF
Overview of Low Power Wide Area Networks
Maarten Weyn
 
PDF
LPWAN London Meetup: Solving Urban Challenges
Digital Catapult
 
PDF
LoRa online training for utility guys
Nikolay Milovanov
 
PDF
Understanding limitations of lo rawan
mafole
 
PDF
Analysis of Energy Consumption in MCU for IoT System
IRJET Journal
 
PDF
FIWARE Global Summit - Introduction to LoRa Alliance and Technology
FIWARE
 
Haystack Technology Overview
Haystack Technologies
 
LoRaWAN vs Haystack
Haystack Technologies
 
Bringing Better Networking to LTE IoT
Haystack Technologies
 
Open fest2018 LoRa, LoRa WAN, The Thing Network - capabilities, myths, applic...
Olimex Bulgaria
 
Cellular lpwan paris nov 2015
robert huynh
 
LoRaWAN101_What is it
Birdz
 
LoRa General Presentation LoRa General Presentation.pdf
ssuser262297
 
LoRaPresentation20255555555555555555555555555555555555.pdf
WidedAhmadi
 
LoRa General Presentation FOR MASSIVE DEPLOYMENT.pptx
WidedAhmadi
 
LoRa General Presentation FOR MASSIVE DEPLOYMENT.pptx
WidedAhmadi
 
Haystack + DASH7 Security
Haystack Technologies
 
Long-Range Low-power Radio (LoRa).pptx
Ajay Gangakhedkar
 
IoT_standards
João Santos
 
The IoT Hunger Games 2015
Haystack Technologies
 
Overview of Low Power Wide Area Networks
Maarten Weyn
 
LPWAN London Meetup: Solving Urban Challenges
Digital Catapult
 
LoRa online training for utility guys
Nikolay Milovanov
 
Understanding limitations of lo rawan
mafole
 
Analysis of Energy Consumption in MCU for IoT System
IRJET Journal
 
FIWARE Global Summit - Introduction to LoRa Alliance and Technology
FIWARE
 
Ad

Recently uploaded (20)

PDF
2025-07-15 EMEA Volledig Inzicht Dutch Webinar
ThousandEyes
 
PDF
Sustainable and comertially viable mining process.pdf
Avijit Kumar Roy
 
PPTX
Machine Learning Benefits Across Industries
SynapseIndia
 
PDF
Productivity Management Software | Workstatus
Lovely Baghel
 
PDF
Log-Based Anomaly Detection: Enhancing System Reliability with Machine Learning
Mohammed BEKKOUCHE
 
PPTX
Darren Mills The Migration Modernization Balancing Act: Navigating Risks and...
AWS Chicago
 
PDF
HydITEx corporation Booklet 2025 English
Георгий Феодориди
 
PDF
How Current Advanced Cyber Threats Transform Business Operation
Eryk Budi Pratama
 
PDF
Ampere Offers Energy-Efficient Future For AI And Cloud
ShapeBlue
 
PDF
UiPath vs Other Automation Tools Meeting Presentation.pdf
Tracy Dixon
 
PDF
Arcee AI - building and working with small language models (06/25)
Julien SIMON
 
PPT
Interview paper part 3, It is based on Interview Prep
SoumyadeepGhosh39
 
PDF
Wojciech Ciemski for Top Cyber News MAGAZINE. June 2025
Dr. Ludmila Morozova-Buss
 
PDF
Novus Safe Lite- What is Novus Safe Lite.pdf
Novus Hi-Tech
 
PDF
Women in Automation Presents: Reinventing Yourself — Bold Career Pivots That ...
DianaGray10
 
PPTX
Extensions Framework (XaaS) - Enabling Orchestrate Anything
ShapeBlue
 
PDF
Building Resilience with Digital Twins : Lessons from Korea
SANGHEE SHIN
 
PDF
NewMind AI Journal - Weekly Chronicles - July'25 Week II
NewMind AI
 
PPTX
Building and Operating a Private Cloud with CloudStack and LINBIT CloudStack ...
ShapeBlue
 
PDF
Upgrading to z_OS V2R4 Part 01 of 02.pdf
Flavio787771
 
2025-07-15 EMEA Volledig Inzicht Dutch Webinar
ThousandEyes
 
Sustainable and comertially viable mining process.pdf
Avijit Kumar Roy
 
Machine Learning Benefits Across Industries
SynapseIndia
 
Productivity Management Software | Workstatus
Lovely Baghel
 
Log-Based Anomaly Detection: Enhancing System Reliability with Machine Learning
Mohammed BEKKOUCHE
 
Darren Mills The Migration Modernization Balancing Act: Navigating Risks and...
AWS Chicago
 
HydITEx corporation Booklet 2025 English
Георгий Феодориди
 
How Current Advanced Cyber Threats Transform Business Operation
Eryk Budi Pratama
 
Ampere Offers Energy-Efficient Future For AI And Cloud
ShapeBlue
 
UiPath vs Other Automation Tools Meeting Presentation.pdf
Tracy Dixon
 
Arcee AI - building and working with small language models (06/25)
Julien SIMON
 
Interview paper part 3, It is based on Interview Prep
SoumyadeepGhosh39
 
Wojciech Ciemski for Top Cyber News MAGAZINE. June 2025
Dr. Ludmila Morozova-Buss
 
Novus Safe Lite- What is Novus Safe Lite.pdf
Novus Hi-Tech
 
Women in Automation Presents: Reinventing Yourself — Bold Career Pivots That ...
DianaGray10
 
Extensions Framework (XaaS) - Enabling Orchestrate Anything
ShapeBlue
 
Building Resilience with Digital Twins : Lessons from Korea
SANGHEE SHIN
 
NewMind AI Journal - Weekly Chronicles - July'25 Week II
NewMind AI
 
Building and Operating a Private Cloud with CloudStack and LINBIT CloudStack ...
ShapeBlue
 
Upgrading to z_OS V2R4 Part 01 of 02.pdf
Flavio787771
 
Ad

How To Triple The Range of LoRa

  • 1. How To Triple The Range of LoRa © 2019 Haystack Technologies, Inc.
  • 2. 2 Executive Summary • Haystack is introducing a new DASH7 implementation for LoRa: XR Mode. • Using software-only enhancements, DASH7 XR Mode for LoRa provides 2-3x range improvements over comparable observed LoRaWAN performance at ground level while preserving multiple year endpoint battery life. • XR Mode utilizes 2-way LDPC Error Correction as well as Automated Receipt Response for guaranteed message delivery. Shorter packet sizes and lower data rates than are typically deployed with DASH7 are also utilized. • XR Mode running at a higher data rate can achieve the same QoS as LoRaWAN running at a much lower data rate. This translates to 10-20x the network density and 10-20x the battery life vs. LoRaWAN, for a given QoS. • Private beta testing for XR Mode is now underway.
  • 3. 3 Background • Some LoRa range test results make aggressive claims (“Up to 30 miles!”) which often do not match real world customer results. These marketing claims are typically based on “best case” outdoor testing environments and may result in unhappy customers and harm long-term adoption. • In comparison, testing LoRa at ground level yields very different results from testing conducted in such ideal outdoor environments. Yet some Haystack customer requirements include both gateways and endpoints deployed at or near ground level. • To improve range performance at all levels, but in particular at ground level, we spent the last six months optimizing our OpenTag firmware stack for extended range and reduced packet error rates over LoRa. The result: DASH7 XR Mode.
  • 4. DASH7™ 4 Technical Features ‣ 2-way LPWAN networking software ‣ Works across many different LPWAN radios ‣ Ultra-low power (years on coin cell) ‣ Low latency (<2 sec.) queries ‣ Broadcast, multicast, P2P messaging ‣ Two-way error correction ‣ Message delivery confirmation ‣ Real-time GPS, A-GPS support ‣ Indoor location via RSSI, TDOA, others ‣ AES128 private-key + public-key crypto ‣ OTA firmware updates, key refreshes ‣ Smart contract support at endpoint ‣ Invented by Haystack! OSI Layer 7 Application UDP + OIC + NDEF + iOS, etc. 6 Presentation DASH7 Core
 low power low latency low cost 5 Session 4 Transport 3 Network 2 Data Link 1 Physical (Radio) LoRa, Other Options DASH7: The Most Advanced Networking Stack for Low Power Wide Area Networks
  • 5. LoRa® 5 For more information on LoRa: https://www.semtech.com/lora/ LoRa is a radio, not a protocol. To exploit its true potential and performance, it requires an advanced protocol stack Basic Attributes ‣ Popular low power, wide area networking radio from Semtech ‣ Promises multi-year battery life and multi- kilometer range ‣ Mostly intended for use in unlicensed spectrum (915 MHz in USA) ‣ Low cost (~$3 in volume today) ‣ Modules available from Murata, Microchip, ST Micro, others
  • 6. 6 Today: Two DASH7 Options For LoRa Option 1: DASH7 LAN Mode Optimal for usage with most software & business logic. 
 Includes fully bidirectional data transfers, point-to-point, broadcasting, and real-time (asynchronous) communication. Typ. Ground-to-Ground Range: Light-Urban 0.20 miles / 0.33 km Typical Data Rate 16 kbps Typical “Real World” Link Budget 130 dB Endpoint Default Operation Event-driven (asynchronous) LDPC Error Correction (Coming Soon) Message Delivery Confirmation Yes Send Commands, Queries To Endpoint Yes Real-Time Queries to Endpoint Yes Packet Size Variable: up to 256 bytes Multi-Year Battery Life Yes GPS-based Geolocation (Low Power) Yes: OTA A-GPS Geofencing Real-Time, Low Power Broadcast, Multicast Messaging Yes OTA Firmware Updates Yes 0.64 miles / 1 km Typ. Ground-to-Ground Range: Light-Urban 1.1 kbps Typical Data Rate 153 dB Typical “Real World” Link Budget Periodic Msg (synchronous) Endpoint Default Operation Yes LDPC Error Correction Yes Message Delivery Confirmation Limited Functionality Send Commands, Queries To Endpoint No Real-Time Queries to Endpoint Fixed: 16 bytes Packet Size Yes Multi-Year Battery Life Yes: Preloaded A-GPS GPS-based Geolocation (Low Power) Beacon Geofencing No Broadcast, Multicast Messaging No OTA Firmware Updates Option 2: DASH7 XR Mode Optimized for delivering LPWAN pub-sub messaging at maximum theoretical range.
  • 7. 7 Today: Two DASH7 Options For LoRa Option 1: DASH7 LAN Mode Optimal for usage with most software & business logic. 
 Includes fully bidirectional data transfers, point-to-point, broadcasting, and real-time (asynchronous) communication. Typ. Ground-to-Ground Range: Light-Urban 0.20 miles / 0.33 km Typical Data Rate 16 kbps Typical “Real World” Link Budget 130 dB Endpoint Default Operation Event-driven (asynchronous) LDPC Error Correction (Coming Soon) Message Delivery Confirmation Yes Send Commands, Queries To Endpoint Yes Real-Time Queries to Endpoint Yes Packet Size Variable: up to 256 bytes Multi-Year Battery Life Yes GPS-based Geolocation (Low Power) Yes: OTA A-GPS Geofencing Real-Time, Low Power Broadcast, Multicast Messaging Yes OTA Firmware Updates Yes 0.64 miles / 1 km Typ. Ground-to-Ground Range: Light-Urban 1.1 kbps Typical Data Rate 153 dB Typical “Real World” Link Budget Periodic Msg (synchronous) Endpoint Default Operation Yes LDPC Error Correction Yes Message Delivery Confirmation Limited Functionality Send Commands, Queries To Endpoint No Real-Time Queries to Endpoint Fixed: 16 bytes Packet Size Yes Multi-Year Battery Life Yes: Preloaded A-GPS GPS-based Geolocation (Low Power) Beacon Geofencing No Broadcast, Multicast Messaging No OTA Firmware Updates Option 2: DASH7 XR Mode Optimized for delivering LPWAN pub-sub messaging at maximum theoretical range.
  • 8. DASH7 XR Mode for LoRa • XR: “eXtended Range” DASH7 protocol • Runs concurrently with DASH7 LAN Mode • Improves real-world signal propagation over LoRaWAN by 10-13 dB (10-20x) • Retains essential features like real-time GPS location and multi-year battery life • Software-based approach:
 No HW or antenna modifications required • XR Mode delivers the longest range connectivity for battery-powered LoRa devices in the world today 8 Observed Sensitivity @ 1kbps 110 117.5 125 132.5 140 LoRaWAN DASH7 LAN DASH7 XR - - - - - dBm How is this gain possible? 
 XR Mode algorithms come from very recent technology for communicating with probes in deep space.
  • 9. XR Mode: Shorter Packet Lengths 9 • We reconfigured DASH7 frame from the conventional, variable size (up to 256 bytes), to a fixed size of 16 bytes. • Most LPWANs use small packets, so we picked a size that gives the best efficiency for our deep space error correction algorithm. • The packet is small, but it is sufficient for geolocation or sensor data, and it can support strong cryptography. 1 2 3 4 65 8 9 10 11 12 13 14 15 Link Info Subnet Info Tokenized ID Encrypted Forward CRC8 XR Mode 16 Byte Frame 7 16 Data
 ID Data Payload Data
 CRC8
  • 10. XR Mode: Lower Data Rate, Greater Range 10 • Basic Truth 1: lower data rate = more energy per bit = longer range • Basic Truth 2: lower data rate = longer transmission = device drains more power • We observed that roughly 1 kbps is the best data rate trade-off (using LoRa) for achieving maximum range in an LPWAN device with multi-year battery requirement. ‣ DASH7 LAN Mode uses 16 kbps (LoRa SF7, 500 kHz) ‣ XR Mode uses 1.1 kbps (LoRa SF11, 500 kHz, + Haystack LDPC encoding) • We observed a range improvement of XR Mode vs LAN Mode ‣ 3x: Ground-to-Ground with obstacles (extreme non-line-of-sight) ‣ 6x: Elevated antenna to ground endpoint, with obstacles (moderate non-line-of-sight) ‣ >20x: Line of Sight
  • 11. XR Mode: More Efficient than LoRaWAN 11 • Basic Truth 1: lower data rate = more energy per bit = longer range • Basic Truth 2: lower data rate = longer transmission = device drains more power • In order to determine energy use to send a message, we need to establish a baseline QoS. In this case, it is 1% PER (packet loss) at 
 146 dB link budget. • We observed that XR Mode performs significantly better than LoRaWAN at any data rate. As result, if we set equal QoS, XR Mode is 22 times more energy efficient than the longest-range LoRaWAN Mode. ‣ “LoRaWAN DR0” | 150 bps | -126 dBm @ PER 30% | 811 mJ per msg ‣ XR Mode | 1074 bps | - 132 dBm @ PER 1% | 36.5 mJ per msg TPKT Packet Time-on-Air PTX TX Power PRX RX Power XC Cumulative probability of error EMSG Energy of message EMSG = TPKT (PTX + PRX) * XC TPKT 1254 ms PTX 412.5 mW (@ 20 dBm) PRX 42.9 mW XC 1.42 EMSG 811 mJ LoRaWAN DR0 TPKT 194 ms PTX 145.2 mW (@ 14 dBm) PRX 42.9 mW XC 1.0 EMSG 36.5 mJ DASH7 XR Mode
  • 12. 12 The greatest advancement DASH7 XR Mode brings is its
 error correction technology.
  • 13. 1. Forward Error Correction (FEC), where sender encodes the data using an error- correcting code prior to transmission. The additional information (redundancy) added by the code is used by the receiver to recover the original data in the case of corruption. 2. Automatic Repeat Request (ARQ), which uses acknowledgements (messages sent by the receiver indicating that it has correctly received a packet) and timeouts (specified periods of time allowed to elapse before an acknowledgment is to be received) to achieve reliable data transmission over an unreliable service. If the sender does not receive an acknowledgment before the timeout, it usually re-transmits the packet until the sender receives an acknowledgment or exceeds a predefined number of retransmissions. Two Types of LPWAN Error Correction 13
  • 14. LPWANs & Forward Error Correction 14 LPWANs operating in unlicensed radio spectrum contend with high levels of noise, unpredictable network congestion, or ground-level gateway deployments where terrain and other objects negatively impact range. Packet Error Rate (PER)
 Percentage of packets not successfully received by a gateway. If we can find a way to reduce PER, we can increase the range and/or Quality of Service (QoS) of the network. RangePER QoS= Forward Error Correction (FEC)
 FEC is a common remedy to high PER in virtually all contemporary wireless networking technologies, however it is strangely crude or non-existent in most LPWAN standards.
  • 15. LoRa HW Built-in Forward Error Correction 15 • The LoRa radio includes a forward error correction (FEC) feature in the radio hardware. • The downside is that it is an outdated implementation, and it actually does more harm than good in most usages (including LoRaWAN) • The technical details of why it’s a poor design are beyond the scope of this presentation, but the basics are: ‣ It is 1950’s-era technology. Better technology is available. ‣ LoRaWAN and all known usages of LoRa, apart from DASH7, use suboptimal configurations of the built-in error correction. ‣ The actual implementation in HW seems to have some bugs that result in occasional false positives. • We have replaced LoRa’s HW FEC with our software implementation of the most advanced FEC technology known to man (as of 2019) Range PER No FEC LoRa HW FEC
 (Best config) DASH7 XR
 FEC LoRaWAN
  • 16. Mobile LPWANs Require Better PER 16 Environmental
 Variables Terrain Diversity Multipath/ Fading Distance from gateway Distance traveled/day Frequency of Movement Elevation We can see that robust error correction not only makes range greater, but it improves the shape of the PER curve. • For LPWANs of fixed-position endpoints, the endpoints can be adjusted at time-of-deployment to achieve acceptable PER. • Mobile LPWAN solutions must contend with changing or unpredictable variables. PER must be acceptable for worst-case scenarios. • If the environmental variables are unpredictable, you want a PER curve that is as flat as possible. Range PER LoRa HW FEC
 (Best config) DASH7 XR
 FEC LoRaWAN Robust FEC flattens the PER curve, making XR Mode more reliable than LoRaWAN when the environment is unpredictable
  • 17. XR Mode Error Correction
 DASH7 XR Mode uses the most advanced FEC in conjunction with ARQ 17 Automatic Repeat Requests (ARQ) • Operates via DASH7 Networking Layer ‣ Devices know to retransmit incorrectly received frames. ‣ Devices will automatically reduce transmission power to suit the required link budget (saves energy and creates less interference for others) • DASH7 ARQ process operates until: ‣ Packet is corrected ‣ Time-out is reached ‣ Shannon limit reached Forward Error Correction (FEC) • Haystack proprietary implementation of CCSDS “Orange Book” Non-binary LDPC coding ‣ Bidirectional: full decoder implemented on cheap ARM Cortex M0+ endpoint, as well as gateway. ‣ Performance is very close to theoretical limit (Shannon) • Engineering challenge was to get it working in such a RAM & CPU constrained environment. • Similar tech is used to enable communication with space probes out in deep space, and also in 5G Cellular.
  • 18. 18 Important Note:
 DASH7 Error Correction is Fully Bidirectional 2Way • FEC Encoders are relatively simple to implement. 
 FEC Decoders are much harder to implement. ‣ For example, Sony Eltres implements uplink LDPC, but does not implement a downlink to the endpoint. ‣ SigFox also has optional uplink FEC (Convolutional code) • In order to have guaranteed message delivery, the gateway needs to be able to send ARQ messages to the endpoint. ‣ Very important for broadcast-with-acknowledgement, which (for example) Haystack uses to do real-time queries. • If the endpoint cannot decode FEC, the gateway needs to compensate by transmitting messages to endpoint with much greater power. ‣ In rare cases this is OK, but most of the time there are regulatory limits that make it impossible for a gateway to compensate with enough power to reach the endpoint.
  • 19. 19 We have begun range testing 
 XR Mode in the SF Bay Area and have some results to share
  • 20. XR Mode Range Test #1 20 Hardware STMicro Discovery Kit (SX1276) with external dipole TX Power 17 dBm Environment Light-Urban (San Mateo, CA), mostly flat Gateway Elevation 1 m / 3.3 feet Endpoint Elevation 1 m / 3.3 feet Results best range: 1 km / .64 miles Gateway (1m) Endpoint
 (1m) Propagation Fading Profile for Test #1 < 1% LOS (line of sight) 74% NLOS (non-line of sight) 25% Obstructed
  • 21. XR Mode Range Test #2 21 Hardware STMicro Discovery Kit (SX1276) with external dipole TX Power 17 dBm Environment Light-Urban (San Mateo, CA), subtle rolling terrain Gateway Elevation 5 m / 16.4 feet Endpoint Elevation 1 m / 3.3 feet Results best range: 1.77 km / 1.1 mi Gateway (5m) Endpoint
 (1m) Propagation Fading Profile for Test #2 13% LOS (line of sight) 62% NLOS (non-line of sight) 25% Obstructed
  • 22. XR Mode Range Test #3 22 Hardware STMicro Discovery Kit (SX1276) with external dipole TX Power 17 dBm Environment Hilltop, then over SF bay, then suburban Gateway Elevation 360 m / 1181 feet Endpoint Elevation 1 m / 3.3 feet Results best range: 27.5 km / 17.03 miles Gateway (360m) Endpoint
 (1m) Propagation Fading Profile for Test #3 80% LOS (line of sight) 10% NLOS (non-line of sight) 10% Obstructed
  • 23. XR Mode Range Test #4 23 Hardware STMicro Discovery Kit (SX1276) with external dipole TX Power 17 dBm Environment Mountain and Valley Gateway Elevation 1170 m / 3838 feet Endpoint Elevation 1 m / 3.3 feet Results best range: 57 km / 35.9 miles Gateway (1170m) Endpoint
 (1m) Propagation Fading Profile for Test #4 93% LOS (line of sight) 6% NLOS (non-line of sight) 1% Obstructed
  • 24. XR Mode Range Test Conclusions 24 Gateway (1170m) Endpoint
 (1m) Gateway (360m) Endpoint
 (1m) Gateway (5m) Endpoint
 (1m) Gateway (1m) Endpoint
 (1m) 1 km 1.77 km 27.5 km 57 km • Increasing gateway elevation has a major impact in improving range, because it increases the amount of the link that is Line of Sight. • LOS (Line of Sight) portions are shown in green. • The ground itself is a major cause of signal fading (attenuation). This is NLOS (Non-line-of- Sight) and is shown in orange. ‣ Visible light (~566,000 GHz) is a much higher frequency than LoRa (0.9 GHz). ‣ Lower frequencies can travel further, but they have a larger wavefront that is more effected by the ground. • Obstructions cause even more attenuation of the link. Obstructions are shown in red.
  • 25. Antenna Design & Placement • Gateway antenna placement makes a huge difference when optimizing for range. • Nearly all LPWAN range tests with very long range claims (e.g., 10 miles+) use a combination of: 1.High-gain antennas located very high above the ground, 2.Data rates that are too low to be suitable for low-power devices 3.Very sophisticated receiver hardware. ‣ Haystack XR Mode outperformed these systems despite being tested with higher data rate, simple antennas, and low cost, low-power receivers • Even a small increase in gateway height can make a significant difference in range performance. In recent testing, the average difference between 2 meters and 5 meters was nearly 200%. • At these extreme sensitivities, it becomes very important to minimize electromagnetic noise generated in the electronics. This is an engineering challenge. STMicro’s LoRa dev kit, for example, is way too noisy. We needed to isolate the antenna in order to conduct the tests. 25 Option A Option B Guess which gateway option results in better range?
  • 26. Final Thoughts • Haystack enables not just the longest range LoRa connectivity available today, but also the most reliable and the lowest power. • Few LPWAN networking stacks appear to have been designed for mobile use cases as most are only found being applied to fixed use cases and leave developers to “hack” mobile solutions or ignore them altogether. LoRa’s utility in mobile use cases is limited without robust two-way error correction and other features like multicast. Haystack XR Mode is directly targeted to mobile developers. • Haystack is offering XR mode demo software to a limited number of beta testers. Register here: http://bit.ly/XRModebeta. • Contact us: [email protected] 26