Galileo Optimization: Single and Dual Frequency Testing Results European Space Week - User Consultation Platform – Mass Market Session –Consumer Platforms: Smartphones and Wearables
Tomasz Lewandowski
02.12.2020
Agenda
➢ GSA testing activities
➢ Galileo development and implementation
➢ Smartphones Testing• Test Plan• Test Cases • Test Setup• Devices Under Test• Location• Results
➢ Wearables Testing• Motivation• Test Plan• Test Cases Execution• Results
➢ Conclusions
GSA testing activities in mass market segment
➢ GSA (MKD) runs testing campaigns in various market segments
➢ GSA manages and defines the testing; moreover GSA acts as an interface for the
receiver/chipset manufacturers
➢ Testing related with mass market segment is performed by Airbus, which supports GSA
in different activities in the framework of the Galileo Service and System Support
contract (WP1x - W02)
➢ There are testing campaigns for the smartphones and wearables which are on-going or which
were performed recently; some of the outcomes are presented on the following slides
Galileo development and implementation
Development of the Galileo constellation and improvement of the Galileo implementation in the mass market devices.
➢ Test date: February 2017, Huawei Mate 9 (SF), Galileo in-view satellites: 6 – 7, Galileo tracked satellites: 3
➢ Test date: July 2020, Samsung Galaxy S10+ (SF), Galileo in-view satellites: 7– 12, Galileo tracked satellites: 7 – 10
Galileo development and implementation
Development of the Galileo constellation and improvement of the Galileo implementation in the mass market devices.
➢ Test date: July 2020, Xiaomi Mi 9 (DF), Galileo in-view satellites: 7 – 12, Galileo (E1) tracked satellites: 4 – 9, Galileo (E5a)
tracked satellites: 3 – 9
Smartphones Testing – Test Plan
➢ Objective• Evaluate the accuracy improvement of dual-frequency smartphones compared to single-frequency
ones
• Evaluate the Galileo standalone performance
➢ Testing Approach
• Live signals: Airbus carried out the live signal tests in real user conditions
• RFCS signals: Joint Research Centre (JRC) carried out the synthetic signal tests
➢ Receivers under test• 3 dual-frequency
• 3 single-frequency
Smartphones Testing – Test Cases
Test ID Test case Condition Test configuration Dynamic Duration Responsible
T-A-SIS-01 Open Sky Assisted Signal in Space Static 4 h Airbus
T-A-SIS-02 Open Sky Assisted Signal in Space Static 24 h Airbus
T-A-SIS-03 Open Sky Unassisted Signal in Space Static 24 h Airbus
T-A-SIS-04 Urban Static Assisted Signal in Space Static 4 h Airbus
T-A-SIS-05 Urban Pedestrian Assisted Signal in Space Dynamic 4 h Airbus
T-A-SIS-06 Suburban Pedestrian Assisted Signal in Space Dynamic 4 h Airbus
T-A-SIS-07 Highway Assisted Signal in Space Dynamic 4 h Airbus
T-U-LAB-08 Open Sky E1/E5 Unassisted RFCS Static 3 h JRC
T-U-LAB-09 Open Sky E1 Unassisted RFCS Static 3 h JRC
T-U-LAB-10 Open Sky E5 Unassisted RFCS Static 3 h JRC
T-U-LAB-11 Open Sky Scintillation Unassisted RFCS Static 3 h JRC
T-U-LAB-12 Urban Pedestrian E1 Unassisted RFCS Dynamic 0.5 h JRC
T-U-LAB-13 Urban Pedestrian E1/E5 Unassisted RFCS Dynamic 0.5 h JRC
Car Setup
Live test cases have been carried out by Airbus in the Munich area. NovAtel SPAN (GNSS + IMU) system was used to compute the reference position.
Pedestrian Setup
Open Sky Setup
Smartphones Testing – Airbus Test Setup
Smartphones Testing – JRC Test Setup
Test cases with the RFCS have been carried out in the JRC facilities. Spirent GSS9000 simulator and
Anechoic Chamber have been used.
Anechoic Chamber
Control Room Spirent GSS9000 simulator
and SimGen interface
Transmitting antenna, camera and smartphones under test
Setup to manage the camera
Transmitting antenna, camera and smartphones under test
Smartphones Testing – Devices Under Test
DeviceSingle-/Dual-
FrequencyChipset Release
Sony Xperia XZ Premium SF Qalcomm Snapdragon 835 June 2017
Samsung Galaxy S8 SF Broadcom BCM4774 April 2017
Samsung Galaxy S10+ SF Broadcom BCM47752 March 2019
Huawei Mate 20 Pro DF HiSilicon Kirin 980 October 2018
Xiaomi Mi 8 DF Broadcom BCM47755 May 2018
Xiaomi Mi 9 DF Qualcomm Snapdragon 855 March 2019
Smartphones Testing – Location
Positions and trajectories of the tests; from the top left to the bottom right: static positions,
highway, suburban pedestrian, urban mobile simulation, urban pedestrian routes
Single-frequencyDual-frequency
Smartphones Testing – Results
Better multipath and noise rejection with L5/E5a
signals (higher bandwidth, narrower correlation)
Slight accuracy improvement
with dual-frequency
All benefit of the code error is not reflected in the
PVT solution
Single-frequencyDual-frequency
Smartphones Testing – Results
Horizontal position error [m]Urban Pedestrian
Samsung Galaxy S10+ (SF) Xiaomi Mi 9 (DF)
63.2th percentile 9.0 5.8
95.0th percentile 24.4 16.2
Smartphones Testing – Results
Number of Galileo and GPS tracked satellites by Huawei Mate 20 Pro and Samsung Galaxy S10+.
Samsung Galaxy S10+ is tracking almost the same
number of Galileo and GPS satellites on E1/L1
More Galileo satellites is tracked on E5a than GPS
on L5
Smartphones Testing – Results
Horizontal accuracy from the urban pedestrian test.
Wearables Testing – Motivation
GSA GNSS Market Report, issue 6, 2019:
➢ Following smartphones, wearableshas become the second most soldGNSS device, reaching 70 mlnshipments in 2019
➢ Wearables and tracking devices areboosting the GNSS market, whilstsmartphone shipments are maturing
Source: GSA GNSS Market Report, issue 6, 2019
Source: https://forums.garmin.com/sports-fitness/running-multisport/f/forerunner-245-series/169019/gps-testing-of-245-music
Source: https://besthiking.net/gps-accuracy-suunto-ambit-suunto-9/
Wearables Testing – Test Plan
➢ Objective
• The aim of this campaign is to assess the navigation performance of the
smartwatches and to evaluate Galileo implementation on these devices
➢ Testing Approach
• All the defined test cases are live tests which are based on the signal in space
(SiS)
➢ Devices under test (single-frequency)
• Suunto 9
• Samsung Galaxy Watch Active 2
• Garmin Fenix 6X PRO
Test ID Test caseTest
configurationDynamic Duration [h]
Placement of the watch
Comment
T-SIS-01 Open Sky Live test Static 1 h N/A
Each test will be executed several times in order to
cover all the possible GNSS
receiver configurations.
T-SIS-02 Open Sky Live testDynamic
pedestrian1 h Wrist
T-SIS-03 Open Sky Live testDynamic
bike1 h Wrist
T-SIS-04 Suburban Live testDynamic
pedestrian1 h Wrist
T-SIS-05 Suburban Live testDynamic
bike1 h Wrist
T-SIS-06 Urban Live test Static 1 h N/A
T-SIS-07 Forest Live testDynamic
pedestrian1 h Wrist
T-SIS-08 Forest Live testDynamic
bike1 h Wrist
T-SIS-09 Suburban Live testDynamic
pedestrian1 h Backpack
Wearables Testing – Test Cases Execution
In order to compute the position error a reference position is needed. A NovAtel SPAN (GNSS + IMU) system was
used for that purpose.
UserDistance traveled
in kilometers
Pedestrian 75
Bike 160
Wearables Testing – Results
Comparisons are performed between devices and also for each device between different GNSS receiver configurations.
Device: Suunto 9Horizontal accuracy [m] on a given
percentileSubtest / Configuration 50.0% 63.2% 95.0%T-SIS-03-01 / GPS only 4.20 4.76 7.45
T-SIS-03-02 / GPS + Glonass 4.78 5.25 7.82 T-SIS-03-03 / GPS + Galileo 2.77 3.32 5.87
Device: Suunto 9Delta time [s] on a given
percentilePVT
availability [%]Subtest / Configuration 50.0% 63.2% 95.0%
T-SIS-03-01 / GPS only 1 1 2 87.56T-SIS-03-02 / GPS + Glonass 1 1 3 83.01T-SIS-03-03 / GPS + Galileo 1 1 2 90.41
In m
any
test
ca
ses
GP
S +
Gal
ileo
co
nfi
gura
tio
n
wa
s th
e b
est
perf
orm
ing
on
e
Conclusions
➢ Smartphones testing
• Dual frequency smartphones take advantage of the L5/E5a signals to get better positioning solution
• Mass market receivers which are used in the smartphones are able to provide a PVT with Galileo signals only
➢ Wearables testing
• Most of the smartwatches which are currently available on the market are able to use Galileo signals
• Among tested devices there was no one for which a particular GNSS receiver configuration would be clearly superior
➢ Common for smartphones and wearables
• Receiver manufacturers consider Galileo in a better and better way in their implementations
• According to the expectations a performance degradation can be observed as the environment gets more harsh; accuracy is
best in the open sky, loses performance in the suburban environment and shows lowest performance in the forest