Today we rely largely on satellite navigation to determine our position and navigate. GNSS have been the prevalent positioning, navigation, and timing technology over the past few decades. However, GNSS signals suffer from some limitations when we are in harsh environments with obstructions that degrade the signal, like for example urban canyons. Our smartphones quietly blend GNSS with other data sources to optimize the accuracy of their results, for example nowadays in every smartphone GNSS is integrated with an Inertial Navigation System (INS). With 5G this trend of ’hybrid positioning’ will accelerate: in fact, 5G cells networks can provide additional corrections to enhance the GNSS localisation accuracy and to complement GNSS when satellites are not visible. In this framework, my thesis focuses on exploring the potential of 5G positioning by exploiting the data from a recent experiment funded by the European Space Agency (ESA). The study took place between March 2022 and August 2022 as part of an internship organized by the Directorate of Navigation of European Space Agency (ESA) and hosted at ESTEC site, in Noordwijk (The Netherlands). In the six months spent in ESTEC we processed data obtained from ESA’s GINTO5G project, whose objective was to organise and collect field measurements with 5G-like positioning signals and to evaluate the potential of fusing GNSS and 5G into one positioning solution. For the execution of this experiment, ESA joined forces with a number of European companies and selected a test area in Germany able to accommodate the equipment needed for the planned measurement campaign. The field campaign consists in estimating the coordinates of a moving trolley in different trajectories, both indoor and outdoor, using three different positioning methods: GNSS (only the outdoor trajectories), 5G and Total Station (used as ground truth). In this thesis 5G data are analyzed and 7 compared with Total Station. The trolley travelled several indoor and outdoor trajectories, each one lasting an average of two minutes. The main objective of this work is to study and implement algorithms for positioning by 5G observations, to asses the potential accuracy of 5G positioning in different environments and to research new techniques to mitigate measurement errors in radio based positioning. To achieve this objective a Python library has been implemented to perform point positioning based on Least Squares approach and using as input Time Difference of Arrivals (TDoAs) generated from differentiating pairs of Time of Arrivals (ToAs) measured on 5G signals. A crucial point in the computation of TDoA from ToA is the choice of the reference Base Station to be used among all the available Base Stations, as the selection could influence the final accuracy of the estimated location. For the analysis and the optimal choice of the correct reference station we used different approaches including the Signal to Noise Ratio (SNR) analysis. To further smoothen the noise of the resulting trajectories, an a posteriori Kalman filter was also implemented. All the results obtained have been analysed both graphically, using interactive maps, and statistically.
Oggi ci affidiamo in larga misura alla navigazione satellitare per determinare la nostra posizione e navigare. Il GNSS è stata la tecnologia di posizionamento, navigazione e temporizzazione prevalente negli ultimi decenni. Tuttavia, i segnali GNSS soffrono di alcune limitazioni quando ci si trova in ambienti difficili, con ostruzioni che degradano il segnale, come ad esempio i canyon urbani. I nostri smartphone mescolano tranquillamente il GNSS con altre fonti di dati per ottimizzare l'accuratezza dei risultati, ad esempio oggi in ogni smartphone il GNSS è integrato con un sistema di navigazione inerziale (INS). Con il 5G questa tendenza al "posizionamento ibrido" si accelererà: infatti, le reti 5G possono fornire correzioni aggiuntive per migliorare l'accuratezza della localizzazione GNSS e per integrare il GNSS quando i satelliti non sono visibili. In questo contesto, la mia tesi si concentra sull'esplorazione del potenziale del posizionamento 5G sfruttando i dati di un recente esperimento finanziato dall'Agenzia Spaziale Europea (ESA). Lo studio si è svolto tra marzo 2022 e agosto 2022 nell'ambito di uno stage organizzato dalla Direzione della Navigazione dell'Agenzia Spaziale Europea (ESA) e ospitato presso il sito ESTEC, a Noordwijk (Paesi Bassi). Nei sei mesi trascorsi in ESTEC abbiamo elaborato i dati ottenuti dal progetto ESA GINTO5G, il cui obiettivo era organizzare e raccogliere misure sul campo con segnali di posizionamento 5G. L'obiettivo principale di questo lavoro è quello di studiare e implementare algoritmi di posizionamento utilizzando osservazioni 5G, valutare la potenziale accuratezza del posizionamento 5G in diversi ambienti e ricercare nuove tecniche per mitigare gli errori di misura .
5G positioning : preliminary analysis of early data sets
PILEGGI, CHIARA
2021/2022
Abstract
Today we rely largely on satellite navigation to determine our position and navigate. GNSS have been the prevalent positioning, navigation, and timing technology over the past few decades. However, GNSS signals suffer from some limitations when we are in harsh environments with obstructions that degrade the signal, like for example urban canyons. Our smartphones quietly blend GNSS with other data sources to optimize the accuracy of their results, for example nowadays in every smartphone GNSS is integrated with an Inertial Navigation System (INS). With 5G this trend of ’hybrid positioning’ will accelerate: in fact, 5G cells networks can provide additional corrections to enhance the GNSS localisation accuracy and to complement GNSS when satellites are not visible. In this framework, my thesis focuses on exploring the potential of 5G positioning by exploiting the data from a recent experiment funded by the European Space Agency (ESA). The study took place between March 2022 and August 2022 as part of an internship organized by the Directorate of Navigation of European Space Agency (ESA) and hosted at ESTEC site, in Noordwijk (The Netherlands). In the six months spent in ESTEC we processed data obtained from ESA’s GINTO5G project, whose objective was to organise and collect field measurements with 5G-like positioning signals and to evaluate the potential of fusing GNSS and 5G into one positioning solution. For the execution of this experiment, ESA joined forces with a number of European companies and selected a test area in Germany able to accommodate the equipment needed for the planned measurement campaign. The field campaign consists in estimating the coordinates of a moving trolley in different trajectories, both indoor and outdoor, using three different positioning methods: GNSS (only the outdoor trajectories), 5G and Total Station (used as ground truth). In this thesis 5G data are analyzed and 7 compared with Total Station. The trolley travelled several indoor and outdoor trajectories, each one lasting an average of two minutes. The main objective of this work is to study and implement algorithms for positioning by 5G observations, to asses the potential accuracy of 5G positioning in different environments and to research new techniques to mitigate measurement errors in radio based positioning. To achieve this objective a Python library has been implemented to perform point positioning based on Least Squares approach and using as input Time Difference of Arrivals (TDoAs) generated from differentiating pairs of Time of Arrivals (ToAs) measured on 5G signals. A crucial point in the computation of TDoA from ToA is the choice of the reference Base Station to be used among all the available Base Stations, as the selection could influence the final accuracy of the estimated location. For the analysis and the optimal choice of the correct reference station we used different approaches including the Signal to Noise Ratio (SNR) analysis. To further smoothen the noise of the resulting trajectories, an a posteriori Kalman filter was also implemented. All the results obtained have been analysed both graphically, using interactive maps, and statistically.| File | Dimensione | Formato | |
|---|---|---|---|
|
CHIARA PILEGGI THESIS.pdf
accessibile in internet per tutti
Dimensione
7.64 MB
Formato
Adobe PDF
|
7.64 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/197104