NAVIGATION, 70(4) Licensed under CC-B Y 4.0 © 2023 Institute of Navigation
Received: 2 November 2021 Revised: 26 April 2023 Accepted: 4 June 2023
DOI: 10.33012/navi.605
ORIGINAL ARTICLE
GEODNET: Global Earth Observation Decentralized
Network
Mike Horton1 David Chen2 Yudan Yi3 Xiaohua Wen4 James Doebbler5
1 INTRODUCTION
The use of global navigation satellite system (GNSS) receivers to collect real-time
solar and space weather observations is well established (Jakowski et al., 2009).
Over the last 20 years, small- and medium-sized networks of GNSS receivers
have collected multiple signal measurements, enabling important industrial and
scientific applications. However, despite the demonstrated benefits of these net -
works, there are currently no openly accessible global, dense GNSS receiver net -
works (Gakstatter, 2014). Furthermore, GNSS receiver signals are easily jammed
and spoofed, yet there is no public, cryptographically secure, spatially dense GNSS
receiver network to help protect the ~$1T USD in location-aware applications.
GEODNET, a public blockchain-based and token-incentivized network, has been
designed to fully address these challenges.
2 DEFINITIONS
The following terms are used in this paper:
• Base (also base station or station): A GNSS receiver permanently installed
at a known fixed location. A virtual base is a base created by blending data
from multiple bases.
1 Mike Horton: ANELLO Photonics
2 David Chen: GEODNET
3 Yudan Yi: EasyRTK
4 Xiaohua Wen: Tersus GNSS
5 James Doebbler: VectorNav
Correspondence
Mike Horton
Email: mike@anellophotonics.com
Abstract
This paper explains some design and architecture decisions around the
GEODNET network and the GeoDAO decentralized autonomous organization,
which aims to create and operate a truly decentralized public GNSS reference
sensing network. This paper covers the motivation of the network, the capabil-
ities of current and future reference stations, the blockchain and GEOD token
mechanics, and how the network powers applications ranging from climate
change monitoring to real-time centimeter-accurate positioning.
Keywords
blockchain, cors networks, ionospheric modeling, precise point positioning
(ppp), real time kinematics (rtk), space weather
HORTON et al.
• Blockchain: A publicly verifiable distributed record of transactions used in
peer-to-peer networks. Blockchain operations are secured cryptographically,
allowing independent verification of their data without centralized trust.
• CORS: A continuously operating reference station (CORS) is a GNSS station
that collects GNSS measurement data, usually as part of a regional or global
network of similar stations.
• GEOD Token, Blockchain, and Wallet Address: GEOD tokens are issued
to a wallet address based on proof of location (POL) and proof of accuracy
(POA) for GNSS miners and proof of stake (POS) for service validator nodes.
Transaction records are stored in the blockchain.
• GeoDAO: The decentralized autonomous organization (DAO) that governs
the GEODNET network.
• GNSS: Global navigation satellite system (GNSS) is the general term for satellite
navigation systems that provide autonomous geospatial positioning with global
coverage, e.g., global positioning system (GPS), GLONASS, Galileo, and BeiDou.
• Ground Motion: Movement in the surface of the Earth that is naturally
caused by tectonic plate motion and other factors such as climate change,
rising sea levels, and desertification.
• IGS: International GNSS Service (IGS) provides an open network of global
GNSS stations, as well as other products and services in support of the terrestrial
reference frame, Earth observation, and research; positioning, navigation, and
timing; and other applications that benefit science and society.
• NRCAN: Natural Resources of Canada (NRCAN) is a Canadian government
organization. NRCAN provides an online PPP tool used for coordinate
determination with high-accuracy geodetic GNSS receivers.
• NTRIP: Networked transport of real-time correction message (RTCM) via
internet protocol (IP) (NTRIP) is a widely used method for transmitting GNSS
correction data over the internet that is commonly used in RTK applications.
• PPP, PPP-AR: Precise point positioning (PPP) (optionally with ambiguity
resolution [PPP-AR]) is an alternative method for centimeter-accurate
positioning that relies on data derived from a network of global stations instead
of a nearby base (or virtual base).
• Public Key Cryptography: Methods and techniques for encryption and
authentication that do not require pre-shared secrets between parties. Public
key cryptography is also referred to as asymmetric cryptography.
• Rover: A GNSS receiver whose location is not fixed, e.g., a smartphone with
a built-in GNSS receiver. The accuracy of the rover can be improved with data
from a base or a network of bases and a suitable positioning engine algorithm.
• RTK: Real-time kinematic (RTK) positioning is a technique for locating a rover
using nearby base data with centimeter accuracy in a global coordinate frame.
The method uses differences in the satellite observations of a base and nearby
rover to remove common errors, including the influence of space weather.
• Smart Contract: A self-executing contract with the terms of agreement
between the buyer and seller written directly into lines of code.
• Space Weather: Changes in the ionosphere and troposphere that are naturally
caused by solar activity and other factors such as air pollution. Space weather
can be tracked via changes to GNSS signals.
• TEQC: Translation – Editing – and Quality Check (TEQC). TEQC is a family
of software tools used for many GNSS preprocessing problems and quality
analysis, available from the UNAVCO website (Estey et al., 1999).
HORTON et al.
3 NEED FOR A DECENTRALIZED GLOBAL GNSS
NETWORK
3.1 Current Centralized Commercial Networks
Currently, the majority of GNSS reference stations fall into two categories: pub -
licly available CORS installations and commercial real-time networks. CORS sta -
tions are generally run by governmental and research entities, providing individual
or regional coverage, and do not form a unified globally consistent, spatially dense
station network.
In developing regions, there is little incentive to deploy a costly network as a
commercial entity or to invest in infrastructure as a governmental or non-profit
organization. Consequently, there are many areas in the world with poor GNSS
reference network coverage.
Volunteer and “crowd-sourced” network models have existed for many years, but
they are not truly decentralized because there is a central organization controlling
and sponsoring the core infrastructure to which the volunteers contribute (server
costs, software development, etc.). Moreover, the lack of incentivization causes the
growth and stability of such arrangements to be slow and uncertain. Additionally,
these networks cannot be developed without significant existing infrastructure
because the volunteer “nodes” rely on a centralized infrastructure.
3.2 Goals of a Decentralized GNSS Reference Network
The following goals are proposed for a valuable alternative to current GNSS
networks:
• Distributed network with no single point of trust or failure
• Robustness against (non-malicious) low-quality data
• Robustness against malicious manipulation (gaming or spoofing)
• Rapid network development without centralized investment (incentivization)
3.3 GEODNET Architecture Overview
The architecture proposed for GEODNET aims to meet the above goals by uti -
lizing modern proven distributed consensus and reward technologies in the form
of blockchain and tokens. This section presents a high-level overview of the com -
ponents and technologies used by GEODNET. Detailed descriptions of the roles of
the nodes as well as the blockchain and network architecture are presented later
in the paper.
GEODNET consists of multiple physical and virtual nodes interacting with each
other and with the blockchain. Copies of the blockchain are stored on many nodes
as a distributed ledger, and updates to the blockchain are performed in a distrib -
uted manner using consensus protocols with no reliance on a centralized or spe -
cific set of nodes. The physical GNSS reference stations (“miners”) generate GNSS
observation data in exchange for tokens. Unlike many other mining approaches
that use an energy-intensive proof of work (POW), where the output of the pro -
cessing work is used to secure transactions on the network, GEODNET utilizes a
POS for validating transactions on the blockchain.
HORTON et al.
The miners are responsible for generating unique and authenticable geoloca -
tion data (GNSS measurements) whereas the validator nodes independently verify
measurement quality, confirm authenticity, and detect intentional manipulation.
Service nodes interface between the miner nodes and external users of the network
in order to provide transport and/or derived products from the miner data. Note
that a single node may perform both validator and service roles. A user may inter-
act with GEODNET through service providers using established protocols (i.e.,
NTRIP) with no knowledge of internal GEODNET workings, allowing immediate
adoption by existing applications.
Incentivization comes in the form of GEOD tokens, which are awarded to differ-
ent node types in exchange for their respective contributions. The real-time value
of the GEOD token is available at https://coinmarketcap.com and other public
blockchain-related websites. Network participants have a vested interest in the suc-
cess of the network, and the awarded GEOD tokens can be securely exchanged for
other tokens, including popular USD-pegged stable tokens (e.g., USDC tokens) on
decentralized exchanges. Once the network is fully operational as a public service,
GEOD tokens will be exchangeable for GEODNET data services via blockchain
transactions.
4 BENEFITS OF BLOCKCHAIN FOR PHYSICAL
NETWORKS
4.1 Blockchain Project Characteristics
Blockchain protocols are the backbone of secure permissionless and decentral -
ized networks. Blockchain projects encourage open standards and open-source
development while economically rewarding participation in network development.
Large communities have developed around blockchain and cryptocurrency proj -
ects, as evidenced by the emergence of Bitcoin and other tokens as a trillion-dollar
asset class and the more than 68M verified accounts on Coinbase (a mainstream
digital currency trading platform) as of its initial public offering (SEC, 2021). The
ability of a successful blockchain project to rapidly deploy distributed infrastruc -
ture makes it an excellent candidate for a global, dense, and secure GNSS receiver
network.
4.2 Related Example: Helium
The Helium network (Haleem et al., 2018) is an existing example realization
of the idea of rapidly deploying a distributed physical network using blockchain
technologies. The Helium project was created to establish decentralized wireless
infrastructure and is now the world’s largest wide-area low-power wireless net -
work implementing the long-range wide-area network (LoRaWAN) standard.
Within two years of the public availability of Helium miners (hotspots), more
than 200,000 hotspots were online, providing dense and inexpensive LoRaWAN
coverage in many of the major cities of the world, and as of September 2022,
there were over 950,000 hotspots registered to the network. GEODNET’s goal
is to provide a similar benefit for access to secured GNSS measurement obser -
vation streams. Helium’s blockchain is based on a proof-of-coverage concept,
whereas GEODNET is based on POL and POA algorithms, which are both
explained in Section 3 of this paper.
HORTON et al.
4.3 Blockchain vs. Tokenomics
A blockchain distributed ledger offers significant technical benefits that enable
decentralized and secure operation of a large-scale CORS network. From a research
perspective, these benefits are separable from the specific economic mechanics of a
cryptocurrency token, i.e., “tokenomics.”
However, it is also a practical necessity to consider the overall operational costs
of a large-scale CORS network with the potential to exceed 100,000 stations.
Assuming a data plan of $100 per month for reliable real-time data delivery, the
recurring operational cost alone may easily exceed $100M USD per year. Thus,
decentralized economics enabled by a blockchain utility token are needed to facili-
tate the transfer of value from network users to station operators in a decentralized
manner. The GEOD utility token is issued to facilitate this goal. The tokenomics
model is beyond the scope of this paper.
4.4 Why Now?
Several trends favor the immediate development of GEODNET:
• Increasing mass-market applications for precise location service
• Availability of low-cost high-precision multi-band multi-constellation
receivers
• Success of related blockchain projects
As a result of these trends, a community-initiated, decentralized network of
high-accuracy GNSS receivers can quickly scale to “mine” space weather data.
5 SPACE WEATHER GNSS MINERS
High-quality yet affordable space weather (GNSS) miners utilize affordable
components with state-of-the-art GNSS receiver hardware. Miners output the
same data as traditional CORS stations, but include additional security protections
to reduce the probability of fake, relayed, or spoofed data entering the network.
Miner designs must be certificated and secured specifically for use on GEODNET.
GeoDAO-approved miner hardware is sold globally for installation on homes and
buildings with clear sky visibility and stable internet connections. An example of
such an installation is shown in Figure 1.
FIGURE 1 Example installation at a residence in Perth, AUS
HORTON et al.
The GEODNET network depends on the delivery of honest GNSS data, and
several steps are taken in the protocol design to prevent fake or simulated
GNSS data from entering the network, while simultaneously rewarding stable
high-quality installations.
Reference designs that include additional ground and atmospheric sensing
modalities are encouraged. Examples include meteorological sensors (i.e., tem -
perature, pressure, and humidity sensors), acceleration sensors for strong motion
measurement, visual or infrared sensors for celestial imaging, lidar units for spa -
tial information capture, and software-defined radio units for tracking low-earth
orbit signals of opportunity and radiofrequency (RF) spectrum measurement.
Additional RF interface options, such as the inclusion of Helium LoRaWAN net -
work compatibility and RF time-difference-of-arrival beaconing, are also being
considered to provide redundant communication channels and non-GNSS back-up
location services.
5.1 Receiver Hardware
The hardware architecture of a miner is shown in Figure 2. The GNSS chipset is
secured, and authenticated data are transferred as described in the next section. The
miner may provide Ethernet, WiFi, and/or cellular communication for the transfer
of data to the internet. The miner must meet the following minimum requirements:
• Calibrated triple-band, full-constellation GNSS antennae
• Minimum dual-frequency, optional triple-frequency GNSS signal tracking
• At least three of the four major constellations: GPS, GLONASS, Galileo, and
BeiDou
• > 50 channels
• < 10 W of total power consumption
• GNSS chipset with a software development kit available for custom secure
firmware
• Cryptographically secure key storage for certification
Miners communicate data to service provider nodes using user datagram proto -
col (UDP) for continuous real-time space weather data and to validators via trans-
mission control protocol (TCP)/IP for configuration and oversight (UDP and TCP/
IP are standard internet protocols). To ensure reliable real-time transmission, the
number of service provider connections per miner is limited by GEODNET. Miners
may be differentiated in the ecosystem based on channels received, additional
functionality, and price.
FIGURE 2 Space weather (GNSS) miner architecture
HORTON et al.
All certificated receiver miner hardware is required to ship with high-quality,
phase-center calibrated triple-band antenna. Use of the POA protocol multipath
and positional instability are penalized in the form of reduced token rewards.
5.2 Secure Receiver: Certification and Encryption
Hardware miner designs and manufacturers are first approved by the GeoDAO.
Upon approval, the manufacturer creates a root X.509 certified by the GeoDAO.
Each miner device receives a unique private and public key pair that is signed with
the manufacturer’s root certificate. The miner’s private key is stored securely within
the device and used to authenticate all communication and data from the miner.
Certification and data encryption secure GEODNET against obvious fake min -
ers, simulated observation data, or use of data relayed from other base station
networks. Authenticating the hardware platform ensures that miner hardware
meets minimum quality standards, and securing the entire receiver processing
and communication path protects the GNSS observation data from manipulation
and guarantees authenticity. Thus, a simulated GNSS RF signal remains the only
(somewhat) practical way to generate authenticated fake data, which is addressed
as part of the POL check.
A GNSS miner is uniquely associated with a wallet address. Miners who do
not have an interest in GEOD tokens or are unable to receive tokens can transfer
earned tokens to the wallet address of the GeoDAO or an alternative destination.
5.3 Proof of Location (POL)
The POL process builds on the fact that the miner has a secured data communi-
cation channel from the GNSS receiver to the GEODNET network and is designed
to achieve two goals:
1. Verify that the GNSS observation data supplied by the miner node match the
stated location (low drift), are of high quality (low noise), and are consistent
with nearby stations.
2. Ensure that the miner is not synthesizing observation data for the purpose of
gaining rewards without contributing unique observation data to the network.
An initial GNSS miner POL validation occurs after a series of successful valida -
tion messages are received within a time threshold, tvalidation. GNSS miners are inval-
idated and removed from GEODNET when they fail to deliver timely validation.
The POL time challenge prevents a dishonest space weather miner from utilizing
simulated RF data, e.g., indoor installation connected to a GNSS RF simulator. For
a miner to remain joined to GEODNET and eligible for token rewards, the miner
must deliver periodic validation messages to the validator nodes within a threshold
time tvalidation, as shown in Figure 3.
The validation message is a pseudorandom code derived from a combination
of datum in each of the four major constellations’ navigation messages. The navi -
gation messages are updated regularly but without advance public notice. An RF
simulator attempting to generate dishonest GNSS measurements must first receive
current navigation messages from actual satellite vehicles in multiple GNSS con -
stellations and then replay them. In general, the reception and replay of navigation
data will incur a delay greater than the threshold time, resulting in the detection of
dishonest GNSS data.
HORTON et al.
Because inexpensive real-time GNSS simulators do not allow low-latency
changing broadcast navigation message bits between nominal bit boundaries, a
reasonable minimum expected latency is a single navigation message bit length.
At typical GNSS message rates of 25–50 bits/s, a simulation latency of at least 20
ms is expected. Assuming that latency estimates for the node are accounted for
first (estimated by roundtrip ping times and also us프로젝트로 돌아가기