Seismic footstep signal characterization

Abstract

Seismic footstep detection based systems for homeland security applications are an important additional layer to perimeter protection and other security systems. This article reports seismic footstep signal characterization for different signal to noise ratios. Various footstep signal spectra are analyzed for different distances between a walking person and a seismic sensor. We also investigated kurtosis of the real footstep signals under various environmental and modeled noises. We also report on the results of seismic signal summation from separate geophones. A seismic signal sum spectrum obtained was broader than that obtained from a single sensor. The peak of the seismic signal sum was broader than that from the footstep signal of the single sensor. The signal and noise spectra have a greater overlap for a seismic signal sum than that from a single sensor. Generally, it is more difficult to filter out the noise from the sum of the seismic signals. We show that the use of the traditional approach of spectrum technology and/or the statistical characteristics of signal to noise of reliable footstep detection systems is not practical.