This thesis investigates Sub-6 GHz multiband Wi-Fi sensing, focusing on coherent combination of distributed frequency subbands to emulate ultra-wideband operation. The central idea is to aggregate Channel Frequency Response (CFR) measurements across multiple non-contiguous Wi-Fi channels, thereby synthesizing a larger effective bandwidth. This approach enhances delay resolution and channel reconstruction accuracy without requiring specialized hardware beyond commodity Wi-Fi systems. The problem is formulated within a compressed sensing framework, leveraging the inherent sparsity of wireless channels in the delay domain. By employing sparse recovery techniques such as Orthogonal Matching Pursuit (OMP) and refining estimates with DBSCAN-based clustering, multipath components are identified with high accuracy. To overcome limitations of existing datasets, a synthetic CSI dataset was developed, covering multiple subbands across the Sub-6 GHz spectrum under controlled conditions. Experimental evaluation demonstrates that coherent subband combination substantially improves multipath resolvability compared to single-band sensing. Furthermore, analysis of frequency-subset selection shows a logarithmic improvement in ambiguity performance as more bands are aggregated. However, diminishing returns appear beyond a certain threshold, with optimal performance achieved when approximately 35–40% of available frequencies are utilized. Overall, the results confirm that Sub-6 GHz Wi-Fi can be transformed into an effective sensing modality through intelligent frequency aggregation, offering a practical and efficient path toward high-resolution wireless sensing.
Questa tesi analizza il sensing multibanda Wi-Fi nella banda Sub-6 GHz, con particolare attenzione alla combinazione coerente di sottobande di frequenza distribuite per emulare un funzionamento a banda ultra-larga. L’idea centrale è quella di aggregare le misure della Channel Frequency Response (CFR) provenienti da più canali Wi-Fi non contigui, sintetizzando così una larghezza di banda effettiva maggiore. Questo approccio migliora la risoluzione in ritardo e l’accuratezza nella ricostruzione del canale, senza richiedere hardware specializzato oltre ai sistemi Wi-Fi commerciali. Il problema è formulato all’interno di un quadro di compressive sensing, sfruttando la naturale sparsità dei canali wireless nel dominio del ritardo. Attraverso tecniche di ricostruzione sparsa come l’Orthogonal Matching Pursuit (OMP) e affinando le stime con clustering basato su DBSCAN, i componenti multipath vengono identificati con elevata precisione. Per superare i limiti dei dataset esistenti, è stato sviluppato un dataset sintetico di CSI, che copre più sottobande nello spettro Sub-6 GHz in condizioni controllate. La valutazione sperimentale dimostra che la combinazione coerente delle sottobande migliora sensibilmente la risoluzione dei cammini multipath rispetto al sensing a banda singola. Inoltre, l’analisi della selezione dei sottoinsiemi di frequenze mostra un miglioramento logaritmico delle prestazioni di ambiguità man mano che vengono aggregate più bande. Tuttavia, i guadagni tendono a saturarsi oltre una certa soglia, con prestazioni ottimali raggiunte quando viene utilizzato circa il 35–40% delle frequenze disponibili. Nel complesso, i risultati confermano che il Wi-Fi Sub-6 GHz può essere trasformato in una modalità di sensing efficace attraverso un’aggregazione intelligente delle frequenze, offrendo un percorso pratico ed efficiente verso un sensing wireless ad alta risoluzione.
coherent multiband ranging using sub 6GHZ wifi signals
MOHAMMAD AMRI, AMIR HOSSEIN
2024/2025
Abstract
This thesis investigates Sub-6 GHz multiband Wi-Fi sensing, focusing on coherent combination of distributed frequency subbands to emulate ultra-wideband operation. The central idea is to aggregate Channel Frequency Response (CFR) measurements across multiple non-contiguous Wi-Fi channels, thereby synthesizing a larger effective bandwidth. This approach enhances delay resolution and channel reconstruction accuracy without requiring specialized hardware beyond commodity Wi-Fi systems. The problem is formulated within a compressed sensing framework, leveraging the inherent sparsity of wireless channels in the delay domain. By employing sparse recovery techniques such as Orthogonal Matching Pursuit (OMP) and refining estimates with DBSCAN-based clustering, multipath components are identified with high accuracy. To overcome limitations of existing datasets, a synthetic CSI dataset was developed, covering multiple subbands across the Sub-6 GHz spectrum under controlled conditions. Experimental evaluation demonstrates that coherent subband combination substantially improves multipath resolvability compared to single-band sensing. Furthermore, analysis of frequency-subset selection shows a logarithmic improvement in ambiguity performance as more bands are aggregated. However, diminishing returns appear beyond a certain threshold, with optimal performance achieved when approximately 35–40% of available frequencies are utilized. Overall, the results confirm that Sub-6 GHz Wi-Fi can be transformed into an effective sensing modality through intelligent frequency aggregation, offering a practical and efficient path toward high-resolution wireless sensing.| File | Dimensione | Formato | |
|---|---|---|---|
|
MohammadAmri_AmirHossein.pdf
accesso aperto
Dimensione
3.85 MB
Formato
Adobe PDF
|
3.85 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/94134