This thesis investigates the estimation and synchronization of temporal parameters in multi-drone systems, with a particular focus on the challenges introduced by dynamic motion and three-dimensional operational environments. Accurate time synchronization is crucial for coordinated operations, reliable communication, and precise positioning among multiple drones, especially in tasks such as search and rescue, environmental monitoring, and autonomous navigation. The proposed synchronization framework relies on Time of Arrival (ToA) estimation, derived from the timestamps of exchanged messages between drones. The ToA values are obtained by measuring the precise transmission and reception instants of packets, allowing the calculation of propagation delays that directly contribute to the estimation of clock skew and offset. This method provides a physically grounded and communication-efficient approach to temporal alignment in distributed UAV systems. The research begins by establishing a foundational framework based on bidirectional message exchange, presenting initial estimation techniques for clock skew and offset. It then extends the analysis to dynamic scenarios, examining uniform and uniformly accelerated rectilinear motions, and demonstrating how drone velocity and acceleration affect the growth of measurement uncertainties. Kalman Filter-based estimation is applied to mitigate the impact of time-varying noise, providing robust synchronization even under rapidly changing conditions. Further, the study expands the model to three-dimensional space, incorporating a Gaussian–Uniform mixture to realistically capture the vertical component of positional uncertainty. This approach improves the robustness and accuracy of skew and offset estimations in real-world operational scenarios, accounting for bounded vertical errors and occasional outliers. Simulation results show that both motion dynamics and communication frequency significantly influence estimation performance, with maximum likelihood estimation consistently outperforming least-squares approaches in terms of accuracy and resilience. Overall, the thesis demonstrates that precise modeling of motion, uncertainty propagation, and measurement noise—together with ToA-based synchronization—is essential for enabling reliable, coordinated, and efficient multi-drone operations, providing a framework that can be applied to advanced aerial systems in practical environments.
Questa tesi analizza l’estimazione e la sincronizzazione dei parametri temporali nei sistemi multi-drone, con particolare attenzione alle sfide introdotte dal moto dinamico e dagli ambienti operativi tridimensionali. Una sincronizzazione temporale accurata è fondamentale per garantire operazioni coordinate, comunicazioni affidabili e posizionamento preciso tra più droni, soprattutto in applicazioni come operazioni di ricerca e soccorso, monitoraggio ambientale e navigazione autonoma. Il modello di sincronizzazione proposto si basa sulla stima del Time of Arrival (ToA), ottenuto dai timestamp dei messaggi scambiati tra i droni. I valori di ToA vengono calcolati misurando con precisione gli istanti di trasmissione e ricezione dei pacchetti, consentendo la determinazione dei ritardi di propagazione che contribuiscono direttamente alla stima dello skew e dell’offset dell’orologio. Questo approccio fornisce un metodo fisicamente fondato ed efficiente in termini di comunicazione per ottenere l’allineamento temporale in sistemi UAV distribuiti. La ricerca inizia definendo un quadro metodologico basato sullo scambio bidirezionale di messaggi, introducendo le tecniche iniziali di stima dello clock skew e dell’offset. Successivamente, l’analisi viene estesa a scenari dinamici, considerando moti rettilinei uniformi e uniformemente accelerati, e dimostrando come la velocità e l’accelerazione dei droni influenzino la crescita delle incertezze di misura. Per ridurre l’impatto del rumore variabile nel tempo, viene applicato un filtro di Kalman, che consente una sincronizzazione robusta anche in condizioni operative altamente dinamiche. Lo studio viene poi ampliato allo spazio tridimensionale, introducendo una distribuzione mista Gaussiana–Uniforme per rappresentare in modo realistico la componente verticale dell’incertezza di posizione. Questo approccio migliora la robustezza e la precisione delle stime di skew e offset in scenari operativi reali, tenendo conto di errori verticali limitati e di possibili valori anomali. I risultati delle simulazioni mostrano che la dinamica del moto e la frequenza di comunicazione influenzano in modo significativo le prestazioni di stima, con il metodo di Massima Verosimiglianza (ML) che supera costantemente l’approccio dei Minimi Quadrati (LS) in termini di accuratezza e stabilità. Nel complesso, la tesi dimostra che una modellazione accurata del moto, della propagazione dell’incertezza e del rumore di misura — insieme alla sincronizzazione basata su ToA — è essenziale per abilitare operazioni multi-drone affidabili, coordinate ed efficienti, fornendo un quadro metodologico applicabile a sistemi aerei avanzati in contesti reali.
Time synchronization and uncertainty modeling for UAV networks
BUSCAINO, MARCO
2024/2025
Abstract
This thesis investigates the estimation and synchronization of temporal parameters in multi-drone systems, with a particular focus on the challenges introduced by dynamic motion and three-dimensional operational environments. Accurate time synchronization is crucial for coordinated operations, reliable communication, and precise positioning among multiple drones, especially in tasks such as search and rescue, environmental monitoring, and autonomous navigation. The proposed synchronization framework relies on Time of Arrival (ToA) estimation, derived from the timestamps of exchanged messages between drones. The ToA values are obtained by measuring the precise transmission and reception instants of packets, allowing the calculation of propagation delays that directly contribute to the estimation of clock skew and offset. This method provides a physically grounded and communication-efficient approach to temporal alignment in distributed UAV systems. The research begins by establishing a foundational framework based on bidirectional message exchange, presenting initial estimation techniques for clock skew and offset. It then extends the analysis to dynamic scenarios, examining uniform and uniformly accelerated rectilinear motions, and demonstrating how drone velocity and acceleration affect the growth of measurement uncertainties. Kalman Filter-based estimation is applied to mitigate the impact of time-varying noise, providing robust synchronization even under rapidly changing conditions. Further, the study expands the model to three-dimensional space, incorporating a Gaussian–Uniform mixture to realistically capture the vertical component of positional uncertainty. This approach improves the robustness and accuracy of skew and offset estimations in real-world operational scenarios, accounting for bounded vertical errors and occasional outliers. Simulation results show that both motion dynamics and communication frequency significantly influence estimation performance, with maximum likelihood estimation consistently outperforming least-squares approaches in terms of accuracy and resilience. Overall, the thesis demonstrates that precise modeling of motion, uncertainty propagation, and measurement noise—together with ToA-based synchronization—is essential for enabling reliable, coordinated, and efficient multi-drone operations, providing a framework that can be applied to advanced aerial systems in practical environments.| File | Dimensione | Formato | |
|---|---|---|---|
|
2025_12_Buscaino.pdf
accessibile in internet per tutti
Descrizione: testo tesi
Dimensione
2.99 MB
Formato
Adobe PDF
|
2.99 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/10589/246289