The rapid proliferation of Internet of Things (IoT) devices has transformed the smart home into a complex distributed system, where the choice of wireless communication protocol is critical for ensuring reliability, efficiency, and interoperability. This thesis conducts a rigorous, empirical performance comparison between two leading technologies: the established Zigbee protocol and the modern, IP based Matter over Thread stack. Through a series of controlled experiments on a hardware testbed, four key performance areas were quantitatively measured: protocol overhead and scalability, latency and throughput, mesh network resilience, and the feasibility of machine learning-based device classification. The findings reveal a clear engineering trade-off. Zigbee exhibits superior performance in terms of low single-hop latency and exceptionally rapid route recovery, rendering it highly agile for smaller static networks. However, the reactive routing protocol employed by the system has been observed to result in network instability and suboptimal scalability in deep multi-hop topologies. Conversely, Matter over Thread exhibits excellent scalability, lower network overhead in larger configurations, and robust multi-hop throughput, establishing it as a more stable foundation for large, interconnected smart homes. This stability, however, is accompanied by a significant reduction in the network's capacity for self-healing in the event of node failures. The findings of this research suggest that neither of the aforementioned protocols is universally superior. Thus, the selection of one over the other signifies a fundamental compromise between Zigbee's reactivity and Thread's stability. By expanding beyond the confines of technical specifications to encompass real-world data, this work provides a critical, data-driven understanding that can assist engineers and developers in the design of the next generation of automated and connected environments.
La rapida diffusione dei dispositivi Internet of Things (IoT) ha trasformato la casa intelligente in un sistema distribuito complesso, in cui la scelta del protocollo di comunicazione wireless è fondamentale per garantire affidabilità, efficienza e interoperabilità. Questa tesi conduce un rigoroso confronto empirico delle prestazioni tra due tecnologie leader: il consolidato protocollo Zigbee e il moderno stack Matter over Thread basato su IP. Attraverso una serie di esperimenti controllati su un banco di prova hardware, sono state misurate quantitativamente quattro aree chiave di prestazioni: overhead e scalabilità del protocollo, latenza e throughput, resilienza della rete mesh e fattibilità della classificazione dei dispositivi basata sul Machine Learning. I risultati rivelano un chiaro compromesso ingegneristico. Zigbee mostra prestazioni superiori in termini di bassa latenza single-hop e recupero del percorso eccezionalmente rapido, rendendolo altamente agile per reti statiche più piccole. Tuttavia, è stato osservato che il protocollo di routing reattivo utilizzato dal sistema comporta instabilità della rete e scalabilità non ottimale in topologie multi-hop profonde. Al contrario, Matter over Thread mostra un'eccellente scalabilità, un overhead di rete inferiore in configurazioni più grandi e un throughput multi-hop robusto, affermandosi come una base più stabile per le grandi case intelligenti interconnesse. Questa stabilità, tuttavia, è accompagnata da una significativa riduzione della capacità di autoriparazione della rete in caso di guasti dei nodi. I risultati di questa ricerca suggeriscono che nessuno dei due protocolli sopra menzionati sia universalmente superiore. Pertanto, la scelta dell'uno rispetto all'altro rappresenta un compromesso fondamentale tra la reattività di Zigbee e la stabilità di Thread. Espandendosi oltre i confini delle specifiche tecniche per includere dati reali, questo lavoro fornisce una comprensione critica e basata sui dati che può aiutare ingegneri e sviluppatori nella progettazione della prossima generazione di ambienti automatizzati e connessi.
Experimental comparison of matter over thread and Zigbee
Nobile, Massimo
2024/2025
Abstract
The rapid proliferation of Internet of Things (IoT) devices has transformed the smart home into a complex distributed system, where the choice of wireless communication protocol is critical for ensuring reliability, efficiency, and interoperability. This thesis conducts a rigorous, empirical performance comparison between two leading technologies: the established Zigbee protocol and the modern, IP based Matter over Thread stack. Through a series of controlled experiments on a hardware testbed, four key performance areas were quantitatively measured: protocol overhead and scalability, latency and throughput, mesh network resilience, and the feasibility of machine learning-based device classification. The findings reveal a clear engineering trade-off. Zigbee exhibits superior performance in terms of low single-hop latency and exceptionally rapid route recovery, rendering it highly agile for smaller static networks. However, the reactive routing protocol employed by the system has been observed to result in network instability and suboptimal scalability in deep multi-hop topologies. Conversely, Matter over Thread exhibits excellent scalability, lower network overhead in larger configurations, and robust multi-hop throughput, establishing it as a more stable foundation for large, interconnected smart homes. This stability, however, is accompanied by a significant reduction in the network's capacity for self-healing in the event of node failures. The findings of this research suggest that neither of the aforementioned protocols is universally superior. Thus, the selection of one over the other signifies a fundamental compromise between Zigbee's reactivity and Thread's stability. By expanding beyond the confines of technical specifications to encompass real-world data, this work provides a critical, data-driven understanding that can assist engineers and developers in the design of the next generation of automated and connected environments.| File | Dimensione | Formato | |
|---|---|---|---|
|
2025_07_Nobile_Tesi_01.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Tesi
Dimensione
804.95 kB
Formato
Adobe PDF
|
804.95 kB | Adobe PDF | Visualizza/Apri |
|
2025_07_Nobile_ExecutiveSummary_02.pdf
accessibile in internet per tutti
Descrizione: Executive Summary
Dimensione
435.47 kB
Formato
Adobe PDF
|
435.47 kB | 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/10589/240758