Location awareness is vital for several use cases in next-generation (xG). However, satisfying the positioning service levels (PSLs) requirements specified by the 3rd Generation Partnership Project (3GPP) is challenging in complex wireless scenarios due to multipath and non-line-of-sight (NLOS) conditions. This calls for the development of innovative localization algorithms able to exploit richer positional information and overcome the limitations of conventional localization approaches. In particular, starting from Release 16, the 3GPP is working towards extending the cellular network localization capabilities with new enablers at di!erent network layers, including the use machine learning (ML)-based algorithms. Exploiting these enablers allows the development of more e!ective localization algorithms, both by increasing the achievable localization accuracy and reducing the number of radio resources needed to satisfy a target PSL. This thesis first summarizes the localization architecture of fifth generation (5G) and beyond networks at the physical layer. Then, it extends the soft information (SI)-based localization framework by introducing techniques to leverage sidelink measurements and to perform e”cient node activation. In addition, the concept of blockage intelligence is proposed to provide a synthetic indicator of wireless channel quality that can be seamlessly integrated into localization algorithms. Case studies in 3GPP-compliant indoor and outdoor scenarios are presented to showcase the e!ectiveness of the proposed techniques. Finally, this thesis introduces xG-Loc, the first open dataset for supporting the development of localization algorithms and location-based services in 3GPP-compliant settings. The results presented in this thesis pave the way for the evolution of xG wireless networks to provide accurate localization in complex wireless environments.
La conoscenza della posizione `e fondamentale per molteplici casi d’uso nelle reti di prossima generazione. Tuttavia, soddisfare i requisiti di servizio per la stima della posizione (PSLs) `e arduo in scenari wireless complessi per via della propagazione multipercorso e delle condizioni di non-linea-di-vista (NLOS). `E quindi necessario sviluppare algoritmi di localizzazione innovativi in grado si sfruttare una pi`u ricca informazione posizionale e superare i limiti degli approcci classici. In particolare, a partire dalla Release 16, il 3rd Generation Partnership Project (3GPP) ha iniziato a estendere le capacit`a delle reti cellulari con nuovi abilitatori a diversi livelli di rete, tra cui l’uso di algoritmi basati su machine learning (ML). Lo sfruttamento di tali abilitatori consente lo sviluppo di algoritmi di localizzazione pi`u e!caci, in grado di aumentare la accuratezza di localizzazione raggiungibile e di ridurre il numero di risorse radio necessarie per soddisfare un PSL obiettivo. In primo luogo, questa tesi richiama l’architettura di localizzazione a livello fisico delle reti di quinta generazione (5G). Poi, estende il framework di localizzazione basato su soft information introducendo tecniche per sfruttare misure ottenute via sidelink e per e”ettuare una fase di node activation efficiente. Inoltre, questa tesi introduce il concetto di blockage intelligence per fornire un indicatore sintetico della qualit`a del canale wireless che possa essere facilmente integrata negli algoritmi di localizzazione. Diversi casi di studio sono presentati in scenari indoor e outdoor 3GPP-conformi per mostrare l’efficacia delle tecniche sviluppate. Le prestazioni degli algoritmi proposti sono valutate tramite simulazioni 3GPP-conformi in scenari indoor e outdoor. Infine, questa tesi introduce xG-Loc, il primo dataset aperto per supportate lo sviluppo di algoritmi di localizzazione e servizi basati sulla posizione negli scenari definiti dal 3GPP. I risultati presentati in questa tesi aprono la strada all’evoluzione delle reti di prossima generazione per fornire accurata localizzazione in ambienti wireless complessi.
Location Awareness in xG Networks
TORSOLI, GIANLUCA
2025
Abstract
Location awareness is vital for several use cases in next-generation (xG). However, satisfying the positioning service levels (PSLs) requirements specified by the 3rd Generation Partnership Project (3GPP) is challenging in complex wireless scenarios due to multipath and non-line-of-sight (NLOS) conditions. This calls for the development of innovative localization algorithms able to exploit richer positional information and overcome the limitations of conventional localization approaches. In particular, starting from Release 16, the 3GPP is working towards extending the cellular network localization capabilities with new enablers at di!erent network layers, including the use machine learning (ML)-based algorithms. Exploiting these enablers allows the development of more e!ective localization algorithms, both by increasing the achievable localization accuracy and reducing the number of radio resources needed to satisfy a target PSL. This thesis first summarizes the localization architecture of fifth generation (5G) and beyond networks at the physical layer. Then, it extends the soft information (SI)-based localization framework by introducing techniques to leverage sidelink measurements and to perform e”cient node activation. In addition, the concept of blockage intelligence is proposed to provide a synthetic indicator of wireless channel quality that can be seamlessly integrated into localization algorithms. Case studies in 3GPP-compliant indoor and outdoor scenarios are presented to showcase the e!ectiveness of the proposed techniques. Finally, this thesis introduces xG-Loc, the first open dataset for supporting the development of localization algorithms and location-based services in 3GPP-compliant settings. The results presented in this thesis pave the way for the evolution of xG wireless networks to provide accurate localization in complex wireless environments.File | Dimensione | Formato | |
---|---|---|---|
Torsoli-Thesis-PhD-2025.pdf
accesso aperto
Dimensione
16.6 MB
Formato
Adobe PDF
|
16.6 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/218626
URN:NBN:IT:UNIFE-218626