Quality of information trade-offs in the detection of transient phenomena

Abstract

In this paper, we present a set of attributes that are being proposed to characterize quality of information (QoI) for sensor-enabled applications in a domain-agnostic manner. We then focus on two important of these attributes, timeliness and data reliability, which capture the quality of detection processes with respect to how fast and how accurately a detection is made. With special emphasis on transient phenomena, i.e., phenomena of limited duration, using traditional Bayesian-based hypothesis testing techniques, we investigate the detection of these phenomena and we analytically derive relationships that capture the QoI of a phenomenon detector as a function of the duration of the observed phenomena and the rate with which observations of the phenomena are collected.