Location awareness is crucial for several use cases of next-generation (xG) networks. On the one hand, location-aware xG networks require algorithms for position inference. For example, the exploitation of soft information (SI) has been shown to provide significant performance gains compared to conventional approaches. On the other hand, such networks require strategies to achieve accurate localization with an e!cient resource utilization. While the 3rd Generation Partnership Project (3GPP) has defined the functionalities for performing localization in xG networks, strategies and guidelines to manage the network resources for e!ciently fulfilling service-level requirements are required. This thesis develops strategies for the e!cient operation of location-aware xG networks. Specifically, we focus on functionalities for: (i) node prioritization, which allocates transmitting resources for inter-node measurements; (ii) node activation, which selects and coordinates the active nodes; and (iii) node deployment, which determines adequate node positions for providing reliable measurements. The goals of this thesis are to develop a framework and establish guidelines for the e!cient operation of location-aware xG networks in complex wireless environments. First, we investigate data-driven network operation and develop strategies for node prioritization and node activation based on theoretical limits as well as on the position estimation error. Such strategies employ neural networks as parametric approximation architectures that replace conventional iterative solvers, allowing to e!ciently adapt to complex wireless environments. Then, we introduce the concept of assisting node and investigate deployment strategies for network localization with assisting nodes. We develop methods based on convex optimization and dynamic programming, and establish a bound on the localization performance that can be achieved by deploying assisting nodes. Finally, we explore data collection for SI-based localization and investigate 5G device-free localization via network experimentation. Case studies considering standardized signals and reference scenarios show the potential benefits that can be reaped by employing the proposed strategies. This thesis establishes a framework for the design of resource optimization strategies, provides guidelines for the e!cient operation of location-aware xG networks, and paves the way for achieving stringent localization performance requirements.
La conoscenza della posizione `e un elemento fondamentale per diversi casi d’uso nelle reti di prossima generazione (xG). Le reti xG richiedono algoritmi per stimare la posizione dei nodi con alta accuratezza, ad esempio, algoritmi basati su soft information (SI) che forniscono un guadagno significativo di prestazioni rispetto agli approcci convenzionali. Inoltre, queste reti richiedono strategie per fornire localizzazione accurata con un utilizzo efficiente delle risorse. Il 3rd Generation Partnership Project (3GPP) ha definito le funzionalit`a per eseguire la localizzazione nelle reti xG. Tuttavia sono necessarie strategie e linee guida per gestire le risorse di rete in modo da soddisfare i requisiti dei diversi servizi. Questa tesi sviluppa metodi per l’ottimizzazione di risorse nelle reti di localizzazione xG. In particolare, vengono trattate le seguenti funzionalit`a: (i) node prioritization, per assegnare le risorse per l’esecuzione di misure; (ii) node activation, per selezionare i nodi attivi; e (iii) node deployment, per determinare le posizioni adeguate dei nodi al fine di fornire misure affidabili. Pertanto, gli obiettivi sono lo sviluppo di strategie e la definizione di linee guida per il funzionamento e!ciente di reti di localizzazione xG in ambienti wireless complessi. Inizialmente vengono sviluppate tecniche data-driven per node prioritization e node activation basate su limiti teorici e sull’errore di stima della posizione. Tali strategie sfruttano reti neurali come architetture parametriche di approssimazione che sostituiscono i solver iterativi convenzionali in modo che le strategie possano adattarsi ad ambienti wireless complessi. Viene poi introdotto il concetto di assisting node e vengono sviluppate strategie di node deployment per la localizzazione basate su ottimizzazione convessa e dynamic programming. Si studia inoltre un limite sulle prestazioni di localizzazione che possono essere ottenute con assisting nodes. Infine viene esplorata la raccolta di dati per la localizzazione basata su SI e viene analizzata la localizzazione device-free con segnali 5G, anche attraverso sperimentazione. Casi di studio con segnali e scenari di riferimento standardizzati mostrano i potenziali vantaggi che si possono ottenere impiegando le strategie proposte. Questa tesi sviluppa un framework per la progettazione di strategie di ottimizzazione delle risorse e il funzionamento e!ciente delle reti xG location-aware, fornendo metodi per soddisfare i requisiti stringenti di localizzazione previsti negli standard.
Resource Optimization Strategies for Location-aware Next-generation Networks
GOMEZ VEGA, CARLOS ANTONIO
2025
Abstract
Location awareness is crucial for several use cases of next-generation (xG) networks. On the one hand, location-aware xG networks require algorithms for position inference. For example, the exploitation of soft information (SI) has been shown to provide significant performance gains compared to conventional approaches. On the other hand, such networks require strategies to achieve accurate localization with an e!cient resource utilization. While the 3rd Generation Partnership Project (3GPP) has defined the functionalities for performing localization in xG networks, strategies and guidelines to manage the network resources for e!ciently fulfilling service-level requirements are required. This thesis develops strategies for the e!cient operation of location-aware xG networks. Specifically, we focus on functionalities for: (i) node prioritization, which allocates transmitting resources for inter-node measurements; (ii) node activation, which selects and coordinates the active nodes; and (iii) node deployment, which determines adequate node positions for providing reliable measurements. The goals of this thesis are to develop a framework and establish guidelines for the e!cient operation of location-aware xG networks in complex wireless environments. First, we investigate data-driven network operation and develop strategies for node prioritization and node activation based on theoretical limits as well as on the position estimation error. Such strategies employ neural networks as parametric approximation architectures that replace conventional iterative solvers, allowing to e!ciently adapt to complex wireless environments. Then, we introduce the concept of assisting node and investigate deployment strategies for network localization with assisting nodes. We develop methods based on convex optimization and dynamic programming, and establish a bound on the localization performance that can be achieved by deploying assisting nodes. Finally, we explore data collection for SI-based localization and investigate 5G device-free localization via network experimentation. Case studies considering standardized signals and reference scenarios show the potential benefits that can be reaped by employing the proposed strategies. This thesis establishes a framework for the design of resource optimization strategies, provides guidelines for the e!cient operation of location-aware xG networks, and paves the way for achieving stringent localization performance requirements.File | Dimensione | Formato | |
---|---|---|---|
GomezVega-Thesis-PhD-2025.pdf
accesso aperto
Dimensione
7.2 MB
Formato
Adobe PDF
|
7.2 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/20.500.14242/218635
URN:NBN:IT:UNIFE-218635