For surveillance of high-security zones, seismic sensors have received considerable attention in numerous civilian and military applications. Since seismic sensors are highly sensitive to the Earth’s surface vibration, outdoor surveillance is difficult during seasonal variations. We acquire empirical seismic signals of various events during normal and rainy weather over a period of three months and analyze the potential challenges in the classification of intrusion with existing solutions. In this work, we propose to categorize events occurring during rain as a separate class to improve upon classification accuracies for human, vehicle, animal movements, and no-disturbance in normal as well as rainy weather. Next, considering the similarities in speech and seismic signals, we propose Gaussian mixture modeling of Mel-frequency cepstral coefficients (MFCCs) and their first derivatives ($\Delta $-MFCCs) of seismic signals followed by maximum likelihood estimation of the corresponding Gaussian mixture model (GMM) of the true event. We conduct extensive experiments with a setup in an outdoor environment and demonstrate that consideration of rain as a separate event greatly enhances the efficacy of the classifier. We provide a comparative performance analysis and show that the proposed classifier with an overall accuracy of 97.7% in normal and rainy weather outperforms existing solutions, popular machine learning (ML), and deep learning (DL) models.

The detection of a weak signal in additive Cauchy noise is of great importance in many applications. A locally optimum detector (LOD) exists for such a scenario; however, it is non-linear in nature. In general, implementation of non-linear detectors is difficult in practice, and linear detectors with good properties, such as high asymptotic relative efficiency (ARE) with respect to the LOD, are often desirable. In this paper, we propose a linear detector for a weak signal in additive Cauchy noise. The proposed test statistic is a linear combination of order statistics. For the special case of a constant signal in additive Cauchy noise, we prove the asymptotic normality of the trimmed linear detector, and show that the ARE of the trimmed linear detector with respect to the LOD is unity. Extensive simulation results are provided to demonstrate that the loss in the performance of the linear detector is very small compared with the non-linear LOD. We also discuss the hardware complexities of the LOD and the linear detector, and demonstrate the advantages of the linear detector over the LOD, in terms of hardware implementation.