855
1 EGNOS L1 MARITIME SERVICE
EGNOS is the satellite-based positioning service over
Europe which provides better accuracy with respect
to GPS standalone position. In addition, EGNOS
provides Integrity, which can be suitable for safety
critical applications in the maritime sector.
The EC (European Commission) and EUSPA
(European Union Agency for the Space Programme)
are working on the future EGNOS L1 Maritime
Service with target service declaration by end of 2023
or beginning of 2024. This service will be defined in a
Service Definition Document and notifications of
predicted EGNOS unavailability will be provided to
mariners through Maritime Safety Information (MSI)
service. At the same time, IEC TC80 (International
Electrotechnical Commission Technical Committee) is
developing a new standard for SBAS L1 receiver
equipment (IEC 61108 7, [1]), which will include the
performance requirements and methods of testing
required for maritime navigation. This standard IEC
61108 7 plans to be published by end of 2023. From
that moment, vessels will start using type-approved
SBAS L1 receivers (compliant with IEC 61108 7) in
order to ensure the required operational performance
for the maritime community.
This article presents the EGNOS (European
Geostationary Navigation Overlay Service)
performance observed along several maritime
campaigns carried out in European waters.
These results in real maritime environment aim at
demonstrating that EGNOS L1 service is suitable to
support navigation in ocean waters, coastal waters
and harbour entrances/approaches according to the
operational requirements defined in the IMO Res.
A.1046 (27) [2]. Thus, this assessment is an evidence of
the benefits of using EGNOS, encouraging maritime
community to equip their vessels with type-approved
SBAS L1 receivers and use EGNOS L1 service once it
is operational.
EGNOS Performance in Several Maritime Campaigns
E. Lacarra
1
, R. Gonzalez
1
& M. Lopez-Martinez
2
1
EGNOS Satellite Services Provider, Madrid, Spain
2
European Union Agency for the Space Programme, Prague, Czech Republic
ABSTRACT: This article presents the EGNOS (European Geostationary Navigation Overlay Service)
performance observed along several maritime campaigns carried out in European waters with the objective to
demonstrate the availability of corrections and the suitability of the accuracy to support maritime navigation.
The regions selected correspond to those located in the border area of EGNOS coverage and include the
following campaigns: Norwegian coast in 2018, Southwest of Europe in 2018 & 2019, Finnish coast in 2019,
Baltic Sea in 2021 and Irish coast in 2022.
These results in real maritime environment aim at demonstrating that EGNOS L1 service is suitable to support
maritime navigation in ocean waters, coastal waters and harbour entrances/approaches according to the
operational requirements defined in the IMO Res. A.1046 (27), being beneficial for maritime community.
http://www.transnav.eu
the
International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 17
Number 4
December 2023
DOI: 10.12716/1001.17.04.
11
856
2 METHOD OF TESTING
Different GNSS campaigns have been carried out
along different European waters, trying to cover the
border areas of EGNOS coverage: Northwest area
within the Norwegian coast and Irish coast, Northeast
within the Baltic Sea and South west within a route
from South of Iberian Peninsula to Canary Ireland.
Different types of receivers were installed on board of
a vessel to output the SBAS corrected position and the
GNSS observation measurements. Additionally, a SIS
L1 recorder was also installed to be able to replay in
the laboratory the recorded GNSS signals and feed
them into different receivers to compute again
different positioning solutions, such as GPS
standalone.
There are different types of receivers in these
GNSS data campaigns: high-end receivers, maritime
receiver and software (SW) receiver.
High-end receiver: It is able to provide high level
of performance and output a wide variety of GNSS
information useful for post-processing analysis.
High-end receivers are normally used in marine
engineering and surveying operations.
Maritime receiver: A receiver brand that is
commonly used for maritime navigation.
SW receiver: This type of receiver is used in the
laboratory to initially test the performance of a
software implementation. The SW receiver was fed
with the GNSS observation measurements
provided by the other receivers installed on board
the vessel in order to compute SBAS or GPS
position in the laboratory.
Once the standard for SBAS L1 receiver equipment
(IEC 61108 7, [1]) is published, vessels will start being
equipped with type-approved SBAS L1 receivers
(compliant with IEC 61108 7) in order to ensure the
required operational performance (i.e. harbour
entrances/approaches). These types of receivers shall
have been developed according to the specification
and shall have passed successfully the required
methods of testing described in that IEC standard.
Considering that the standard will be published by
end of 2023, there were no receivers in line with
potential IEC 61108 7, [1] by the time of the data
campaigns. However, the software receiver used
follows the guidelines developed for SBAS maritime
receivers, implementing the requirements defined as
the baseline for the standard under development.
Details about the methodology of computing can
be found in [3]. Receiver was configured to compute
navigation data and store the associated outputs
every 1 second (1 Hz frequency). Once the vessel had
concluded his route and the equipment was collected,
receiver outputs were analysed to obtain the
following:
Real position. The real position considered as
reference for the position error estimation was
computed in post-processing using Precise Point
Positioning (PPP) with the GNSS observations data
obtained by a high-end receiver.
SBAS Navigation solution. The SBAS Navigation
solution using EGNOS was computed by the
receiver and reported in the output files.
GPS-only solution. GNSS radio frequency L1
signal recorded during the maritime dynamic data
campaign was replayed in the receiver to obtain
GPS standalone solution. GPS-only solution is
compared with SBAS solution to show the
improvements of using SBAS.
Based on this, the following performances were
obtained:
Accuracy: Horizontal Navigation System Error
(HNSE) or Horizontal Position Error (HPE)
achieved 95% of the time. Horizontal Navigation
System Error is the 2D radial error between the
instantaneous position solution (SBAS or GPS)
with respect to the instantaneous real position
(computed using PPP in post-processing). It is
noted that HNSE / HPE was computed every 1
second.
Position Availability: Percentage of time that the
receiver is providing the corresponding position
(EGNOS or GPS).
Considering that EGNOS L1 maritime service
plans to support the navigation in ocean/coastal
waters and harbour entrances/approaches according
to the operational requirements defined in the IMO
Res. A.1046 (27) [2], these performances will be
compared with the accuracy requirement of 10 meters.
3 EGNOS SIS AVAILABILITY
EGNOS Signal in Space was 100% during all the
maritime campaigns, which meets the 99.8% EGNOS
signal in space availability requirement of IMO Res.
A.1046 (27) [2]. The fact that EGNOS system has two
operational geostationary satellites ensured that at
least one satellite was transmitting EGNOS
corrections during the whole time. As an example,
from January 2017 to January 2023, EGNOS SIS
availability was 100%. Figure 1 shows the EGNOS SIS
availability over the last 6 months, and the SIS
availability per SBAS satellite (PRN136 and PRN 123
are the EGNOS GEO satellites in operation since
March 2020).
Figure 1. EGNOS SIS availability trending (last 6 months)
4 MARITIME CAMPAIGNS
4.1 Norwegian coast in 2018
In order to assess the EGNOS performance at user
level in the maritime domain at high latitudes in
857
Europe, a data campaign was carried out along the
Norwegian Coast. In the northbound, the vessel
departed from Trondheim to Kirkenes. The
southbound was from Kirkenes to Bergen.
Figure 2. Vessel trajectory through Norwegian Coast
The Horizontal Navigation System Error (HNSE)
histogram was computed for the EGNOS High-end
receiver (coral), the EGNOS software receiver (blue)
and the GPS standalone receiver (purple) and
presented in Figure 3.
Figure 3. HNSE histogram (left). Daily HNSE 95-th
percentile values (right)
These figures confirm that EGNOS SW receiver
provides the best (the lowest HNSE values) accuracy
performances with values about 1m. This EGNOS SW
was developed implementing the receiver
requirements that were considered important for the
future IEC61108-7 standard [1] for SBAS L1 maritime
receivers. Both, EGNOS SW and EGNOS high-end
receiver (HNSE of 0.95m and 1.26m respectively),
provide enhanced position solution with respect to
GPS (HNSE of 2.03m).
EGNOS position availability was impacted due to
the following reasons:
The coverage area of GEO satellite PRN 120 was
very limited in the Northeast European Area,
which impacted the reception of EGNOS messages
from PRN120 when the vessel was located in the
Northeast area, positions in which the elevation
angle was very low. On 23rd August 2018, PRN
136 was introduced in EGNOS operational
platform replacing PRN 120, which improved the
visibility over the Northeast of Europe.
In northern altitudes, where the elevation of the
SBAS GEO was very low, EGNOS signal were
blocked in areas close to mountains or fiords.
The impact of the orography in the availability of
the EGNOS signal received by the receiver is greater
according to the elevation angle of the SBAS GEO
satellite as shown in Figure 4. It is noted that the
EGNOS position lower than 100% in locations lower
than 62° latitude degrees was because the vessel was
sailing in narrow fiords with poor SBAS satellite
visibility from the vessel.
Figure 4. EGNOS position availability in function of latitude
obtained from EGNOS Rx.
4.2 Southwest of Europe in 2018 and 2019
The oceanographic Hespérides vessel sailed from
Cartagena (Spain) to Antarctica on 19th December
2018 and returned back on 25th June 2019, covering 48
days inside the EGNOS MT27 service area (longitude
from -40° to 40°, latitude from 20° to 72°). In the return
trip, the ship performed some inspections close to
Canary Islands, being possible to test EGNOS
performance in the border coverage area.
Figure 5 shows the route followed by the vessel
along the mission and the number of days analysed.
Additionally, this figure shows in yellow the vessel
trajectory in which the EGNOS navigation solution
was available with no alarms, after passing the
recommended plausibility checks defined hereafter
and which are included in draft IEC 61108-7 [1]. Then,
a navigation position is considered as valid with no
alarms when the following three conditions
(plausibility checks) are met:
4. SBAS position: The navigation solution using
EGNOS messages have been computed.
5. HDOP check: The Horizontal Dilution of Precision
(HDOP) is equal or lower than 4 (HDOP 4).
6. Integrity using RAIM: The receiver autonomous
integrity monitoring (RAIM) status is ok, with a
selected accuracy of 10m at the 95% confidence
level, which corresponds to harbour
entrances/approaches and coastal waters [2].
Figure 5.Vessel trajectory through Southwest Europe and
EGNOS navigation solution
As it is observed in yellow, EGNOS navigation
solution was available with no alarms (SBAS position
computed, HDOP check passed successfully and
RAIM integrity algorithm provides a safe status) in
the coastal waters in Europe, covering Canary Islands,
Madeira and South of Spain. As expected, EGNOS
navigation solution (with no alarms) was not available
when getting close to the MT27 area as expected,
especially when located down to 25 degrees North.
Then, the potential SBAS L1 maritime service area for
harbour entrances/approaches and coastal waters
858
could cover in the Southwest of Europe the Canary
and Madeira Coasts.
When computing the percentage of time that the
EGNOS position was OK at receiver level with no
alarms (the three conditions met) for those days in
which the latitudes were greater than 25°, it resulted
in 99.5% for the Maritime SW receiver and 99.7% for
the SBAS receiver. It is noted that these figures are
very good, considering that they were affected by
receiver local effects, such as problems in signal
tracking, gaps in processing or other issues in pseudo-
range measurements.
Figure 6 shows the Horizontal Navigation System
Errors (HNSE) for EGNOS and GPS computed for the
receivers: SBAS High-end Rx and Maritime SW.
EGNOS HNSEs are lower with respect to GPS in all
the cases, meeting with quite margin the 10 meters
requirement for ocean/coastal waters and harbour
entrances/approaches [2].
Figure 6. EGNOS accuracy: CDF figures (left: SBAS High-
end Rx, right: SW Maritime)
Checking the performance per specific days, there
are situations in which position errors are especially
better in EGNOS with respect to GPS. Especially for
areas close to equator in which the ionospheric
activity is high, where EGNOS can monitor the
ionosphere in real time and broadcast up-to-date
corrections, providing more stable accuracy results.
Below an example is presented, in which GPS solution
reached instantaneously the 10-meter error, whereas
with SBAS, the position errors are more stable and
always lower than 4 meters (0.925 meters percentile at
95%).
Figure 7. EGNOS vs GPS: 22/12/2018
4.3 Finnish coast in 2019
The vessel Mastera departed from Porvoo (Finland) to
Primorsk (Russia) then until Naantali (Finland).
Afterwards, the vessel went from Naantali (Finland)
to Primorsk (Russia) being two days stopped close to
Uusimaa waiting for orders (adrift) and finally
arrived at Porvoo (Finland). The total trajectory was
from 1st November 2019 to 14th November 2019.
Figure 8. Vessel trajectory through Finnish coast
Within this GNSS campaign, EGNOS position
availability was 100% for both SBAS maritime receiver
and high-end receiver.
For the EGNOS position accuracy, Table 1 presents
daily HPE (95%) for EGNOS and GPS standalone,
both for High-end and Maritime receivers. It is
noted that for that the GPS position for the high-end
receiver was only available for 5 days, the duration of
the recorded SIS during the campaign. It is observed
that the EGNOS position was always better than the
GPS position. For example: 4.07 m GPS position for
High-end receiver with respect to the 1.18 m for
EGNOS position on DOY (GPS Day Of the Year) 305.
Table 1. HPE (95%) obtained using EGNOS and using GPS
standalone
________________________________________________
EGNOS HPE GPS standalone HPE
95% percentile [m] 95% percentile [m]
________________________________________________
DOY High-end Maritime High-end Maritime
receiver receiver receiver receiver
________________________________________________
305 1.179 0.869 4.073 1.444
306 0.935 0.791 3.284 1.809
307 0.970 0.792 1.653 1.222
308 0.970 0.747 2.103 1.638
309 1.023 0.835 1.738 1.307
310 0.967 0.841 2.235 2.000
311 0.950 0.907 1.832 --
312 0.956 0.718 3.818 --
313 0.873 0.762 3.312 --
314 1.000 0.786 3.527 --
315 1.057 0.808 3.860 --
316 0.911 0.784 1.208 --
317 0.870 0.792 1.228 --
318 1.049 0.872 1.688 --
________________________________________________
TOTAL 0.971 0.815 3.149 1.551
________________________________________________
Additionally, Figure 9 shows the HNSE and ECDF
(Empirical Cumulative Distribution Function) for the
first 5 days where there were results for GPS position.
As it can be observed, EGNOS position errors were
considerably reduced with respect to GPS and the
EGNOS accuracy met with quite margin the 10 meters
requirement for ocean/coastal waters and harbour
entrances/approaches [2].
859
Figure 9. HNSE Histogram for Maritime Rx (left) and ECDF
for both receivers (right)
4.4 Baltic Sea in 2021
Kiisla is a Finnish Aframax crude oil tanker operated
by Neste Shipping. This icebreaking tanker
transported crude oil year-round along the coast of
the Baltic Sea. OSM Group, the Norwegian Oil
Transport provider, manages this vessel. Kiisla,
departed from Munkebo (Denmark) on September 1st
2021 and navigated up to December 1st 2021. The oil
tanker visited eleven harbours placed in five different
countries as seen in Figure 10: Denmark, Lithuania,
Latvia, Sweden and Finland.
Figure 10. Baltic Sea maritime campaign in 2021
EGNOS position availability was 99.99% for the
SBAS maritime receiver. All days presented 100%
EGNOS position availability but one day that was
99.93% due to a local failure at receiver level.
Table 2 shows main statistics for the horizontal
position accuracy when the maritime receiver was
using EGNOS position solution.
Table 2. Horizontal position accuracy obtained using
EGNOS
________________________________________________
EGNOS horizontal accuracy [m]
Period of time (DOY) Mean 95%-percent. Max.
________________________________________________
244 - 253 0.310 0.715 1.168
254 - 263 0.337 0.844 2.350
264 - 273 0.327 0.749 1.145
274 - 282 0.419 0.925 1.829
283 - 293 0.351 0.826 1.447
294 - 303 0.359 0.881 1.730
304 - 313 0.360 0.955 2.068
314 - 320 0.387 0.920 1.617
321 - 335 0.399 0.939 1.466
________________________________________________
TOTAL 0.361 0.779 2.350
________________________________________________
Table 2 presents daily HNSE (95%) quite close to
1m when using EGNOS solution whereas the global
value for the whole data campaign is 0.779m with a
maximum value of 2.35m.
Instantaneous EGNOS horizontal position error is
drawn in the following picture, where it is observed
that EGNOS position met with quite margin the 10-
meter requirement for ocean/coastal waters and
harbour entrances/approaches [2].
Figure 10. EGNOS HPE vs time (left) and EGNOS HPE
histogram & CDF (right)
4.5 Irish coast in 2022
The vessel Granuaile navigated from Killybegs (North
West of Ireland) on 17th August 2022 up to Dún
Laoghaire on 15th November 2022 (North East of
Ireland), going between both ports up to Oban (North
England) (see Figure 11). It sailed going in anti-
clockwise direction. The ILV Granuaile is a multi-
functional vessel to maintain the automatic navigation
buoys in Irish waters.
Figure 11. Vessel trajectory through Irish coast
EGNOS position was provided along the whole
trajectory by the maritime receiver during 99.39% of
time. Seven days were identified with EGNOS
receiver position availability lower than 100%, which
were all caused by local issues, such as lack of GNSS
signal, power cut-off and other types of problems at
860
receiver or antenna level. EGNOS system was
available.
Analysing the accuracy results (see Figure 12), it
can be concluded that percentile at 95% of
instantaneous Horizontal Position Errors using
EGNOS is 0.788 meters, meeting with quite margin
the 10 meters requirement for ocean/coastal waters
and harbour entrances/approaches [2].
Figure 12. EGNOS HPE vs time for 23/08/2022 (left) and
EGNOS HPE histogram for all trajectory (right)
5 CONCLUSIONS
This article presents the EGNOS (European
Geostationary Navigation Overlay Service)
performance observed along several maritime
campaigns carried out in European waters with the
objective to demonstrate the availability of corrections
and the suitability of the accuracy to support maritime
navigation. The regions selected correspond to those
located in the border area of EGNOS coverage and
include the following campaigns: Norwegian coast in
2018, Southwest of Europe in 2018 & 2019, Finnish
coast in 2019, Baltic Sea in 2021 and Irish coast in 2022.
Analysing the EGNOS and GPS navigation
position, the main conclusions are the following ones:
1. EGNOS signal in space availability was 100%
during all campaigns meeting the 99.8%
requirement of IMO Res. A.1046 (27) [2]. The fact
that EGNOS system has two operational
geostationary satellites ensures that at least one
satellite is transmitting EGNOS corrections,
providing 100% of EGNOS SIS availability.
2. EGNOS position availability was nominally 100%
in open sky visibility in locations between 25° and
70° degrees latitude.
In narrow fiords in Norway the SBAS visibility
was blocked, especially in latitudes over 68
degrees where EGNOS satellites were observed
with low elevation. As an example, EGNOS
position availability was 98.5% over the North
of Norwegian coast from 68° to 70° latitudes.
EGNOS position availability presented high
value above 25° degrees latitude (99.7% for
SBAS High-end receiver), covering navigation
of Canary Islands waters.
In nominal conditions of open sky visibility
with no issues at receiver or antenna level,
EGNOS position availability was 100%. This
was the case of the Finnish coast campaign.
EGNOS position availability was impacted in
sporadic moments by local issues at receiver or
antenna level, the observed unavailability
events were not caused by EGNOS corrections.
For Baltic Sea campaign, EGNOS position was
99.99% due to a local failure at receiver level.
For Irish coast trajectory, EGNOS position was
99.39% due to several local issues, such as block
of GNSS signal, power cut-off and other types
of problems at receiver or antenna level. It is
noted that the Irish coast campaign lasted 90
days being more prompt to local problems and
not related to EGNOS SiS corrections
availability.
3. EGNOS position accuracy at percentile 95th was
lower than 2 meters in all the campaigns, meeting
with quite margin the 10 meter requirement for
harbour entrances/approaches and coastal waters
established in IMO Res. A.1046 (27) [2].
EGNOS position accuracy presented slightly
higher accuracy values over the North border
area and over the Southwest border area,
whereas the improvement with respect to GPS
was bigger and the position accuracy values
were much more stable.
Apart from navigating close to EGNOS
coverage area, GNSS position performance
was more impacted over those areas due to
ionospheric issues. For example, high-end
receiver presented instantaneous EGNOS
HNSE at percentile 95% during the
Norwegian coast of 1.26m and during the
Southwest Europe campaign of 1.506m,
whereas in Finnish coast was 0.971m.
Additionally, maritime receiver presented
instantaneous EGNOS HNSE at percentile
95% during the Southwest Europe campaign
of 1.061m, whereas in Finnish coast was
0.815m and in Baltic Sea was 0.779m.
In those areas where GNSS performance is
prompted to be impacted by ionospheric
issues, EGNOS enhancement with respect to
GPS is bigger. EGNOS can monitor the
ionosphere in real time and broadcast up-to-
date corrections, providing more stable
accuracy results. An example is presented
for the 22nd Dec 2018, where there were
peaks of GPS position error up to 10 meters,
whereas EGNOS position remained lower
than 4 meters during the whole day. In this
case improvement of EGNOS with respect to
GPS was quite notably considering that the
HNSE (95%) was 0.925 m for EGNOS and
3.153 m for GPS.
EGNOS position accuracy at 95% was
nominally lower than 1 meter for both receivers
(Maritime and High-end) in trajectories that
were not located in the border EGNOS
coverage area, such as Finnish coast, Baltic Sea
and Irish coast.
Position errors using EGNOS were between
30% and 70% better than GPS Standalone
considering the percentile at 95% of the
complete campaign. For specific cases of high
solar activity or ionospheric issues, EGNOS
improvement was considerably important (i.e.
70% for example presented on 22nd Dec 2018)
since EGNOS navigation position was stable
during the whole period whereas GPS
presented important high position errors (even
reaching 10 meters).
861
In consequence, the EGNOS performance observed
on these maritime campaigns indicates that EGNOS
can support “Harbour entrances/approaches and
coastal/ocean waters” according to IMO Res A.1046
(27) [2] over Norwegian coastal waters, Irish coastal
waters, Finnish coastal waters, Canary Ireland coastal
waters and Baltic Sea. Once the IEC 61108-7[1] test
standard on SBAS is published, vessels will start
equipping type-approval receivers using EGNOS
Service and RAIM to ensure a safe navigation in
harbour entrances/approaches and coastal waters.
ACKNOWLEDGMENTS
We would like to express their gratitude to the Norwegian
Coastal Administration for their good collaboration in order
to test the EGNOS performances in the Norwegian coasts.
We would like to thank Hurtigruten Cruises and MS
Finnmarken for letting us use its cruise ship for this GNSS
data campaign.
We would like to express our gratitude to the Spanish Navy
(Armada Española), the Marine Technology Unit, the
Instituto Hidrográfico Nacional (Spanish Navy
Hydrographic Institute), the Programmes Management Vice
Directorate & National Armaments Directorate (Ministry of
Defence of Spain), the Division of Transport Studies and
Technologies & General Secretariat for Transports (Ministry
of Transports of Spain) to allow ESSP to install the receiver
on board this vessel.
We would like to express our gratitude to OSM Group AS
(Norwegian Oil Transport provider) and Väylä (Finnish
Transport Infrastructure Agency) to allow ESSP to perform
this GNSS data campaign in the Gulf of Finland and install
GNSS equipment in Mastera.
We would like to kindly express their gratitude to OSM
Group, the Norwegian Oil Transport provider to allow
ESSP to install the receiver on board this oil tanker, Kiisla, as
well as to Väylä, the Finnish Transport Infrastructure
Agency to invite ESSP to perform this GNSS data campaign
in the Baltic sea and coordinate it with OSM Group.
We would like to thank both Commissioners of Irish Lights
to allow ESSP the installation of GNSS equipment in ILV
Granuaile vessel and GLA (General Lighthouse Authorities)
of the UK and Ireland (Research and Radionavigation
Directorate) for its support in the installation and use of
their MF antenna during the GNSS data campaign around
Ireland.
Finally, the authors would like to acknowledge the efforts
done by EC and EUSPA to work at programme level for the
future provision of EGNOS L1 maritime service.
REFERENCES
[1] Standard under development IEC 61108-7: Maritime
navigation and radio-communication equipment and
systems Global navigation satellite systems (GNSS) -
Part 7: Satellite Based Augmentation System (SBAS) L1
Receiver equipment Performance standards, methods
of testing and required test results
[2] IMO Resolution A.1046 (27) Worldwide Radio-
navigation System, 20 December 2011.
[3] Transnav Journal: EGNOS Performance Along Finnish
Coast. R. González , E. Lacarra , M. López , K. Heikonen.
DOI: 10.12716/1001.15.03.07