Pregled bibliografske jedinice broj: 1192049
Low-power sensor interfaces for detection of sporadic events
Low-power sensor interfaces for detection of sporadic events, 2022., doktorska disertacija, Fakultet elektrotehnike i računarstva, Zagreb
CROSBI ID: 1192049 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Low-power sensor interfaces for detection of
sporadic events
Autori
Gazivoda, Marko
Vrsta, podvrsta i kategorija rada
Ocjenski radovi, doktorska disertacija
Fakultet
Fakultet elektrotehnike i računarstva
Mjesto
Zagreb
Datum
21.04
Godina
2022
Stranica
178
Mentor
Bilas, Vedran
Ključne riječi
always-on wake-up interfaces, interface architectures, acoustic event detection, acoustic signal features, switched inductor feature extractor, passive electromechanical feature extractor
Sažetak
The focus of this thesis was on exploring and developing wake-up sensor interfaces, used in two- stage sensor nodes for lowering power consumptions to levels allowing years of energy autonomy. The developed interfaces were designed to detect sporadic transient events, generating weak signals (around 10 mV on a passive transducer) in the lower acoustic spectrum (up to 2.5 kHz), lasting from around half a second up to a few seconds. The passing motor vehicle (speedboat) event was chosen as a representative case of this event group. An analysis of acoustic signal features was performed, and a set of criteria was developed and used to determine the feature’s applicability in low-power interfaces for passing speedboat detection. Based on the criteria, the spectro- temporal decomposition and level-crossing rate interface architectures were selected and further explored. Two prototype interfaces were developed and functionally tested. The results showed both architectures were applicable for detection of sporadic transient acoustic events, with true positive event detection rates exceeding 90%. They also showed that the level-crossing rate architecture was functional with two times lower input voltages, required around 40% less area and components to implement and had over three times lower power consumption. This emphasized the impact of feature choice on interface architecture and demonstrated the potential advantages of the level-crossing rate interface. Further examination of the spectro-temporal decomposition architecture pointed to its envelope detector limiting the interface’s detection accuracy. To solve this problem, utilization of a switched inductor was proposed to increase the energy efficiency of conventional envelope detectors. Several envelope detectors were developed and tested, including one using a switched inductor. Following this, two feature extractors were developed, one utilizing an electrically and the other a mechanically switched inductor. The feature extractors utilizing the electrically and mechanically switched inductors achieved 43% and 72% lower power consumptions and had four- and two-times lower minimal input voltage, respectively, compared to the conventional spectro-temporal decomposition interface, demonstrating the applicability and advantages of the switched inductor in low-power sensor interfaces for detection of sporadic acoustic events.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika
POVEZANOST RADA
Ustanove:
Fakultet elektrotehnike i računarstva, Zagreb