Anomalous Seismic Events Detection by Build-In Sensors on Smart Phones

Abstract

Motion sensor equipped smart phones can be operated as the Community Seismic Network for earthquake detection, as the build-in accelerometers share similar principles of those in high precision seismic stations. However, current smart phones often yield poor monitoring quality due to the limited sensing capability of the off-the-shelf sensors and unpredictable dynamics of daily activities of phone users. In this paper, we present a suite of algorithms towards detecting anomalous seismic events from noise contaminated signals, including lightweight signal pre-processing, a two-phase P-phase pickup and timing scheme at the local smartphones, and employing decision fusion to maximize anomaly detection performance at fusion center while meet the requirements on system false alarm rate. We experimentally evaluate the proposed approach on networked smart phones and shake table. The results, including data sampled by smart phones and the numeric simulation, indicate the effectiveness of our scheme in distinguishing anomalous seismic events from motions due to normal daily manipulation.