Description:
Incipient fault detection in low signal-to-noise ratio (SNR) conditions requires robust features for accurate condition-based machine health monitoring. Accurate fault classification is positively linked to the quality of features of the faults. Therefore, there is a need to enhance the quality of the features before classification. This paper presents a novel vibration spectrum imaging (VSI) feature enhancement procedure for low SNR conditions. An artificial neural network (ANN) has been used as a fault classifier using these enhanced features of the faults. The normalized amplitudes of spectral contents of the quasi-stationary time vibration signals are transformed into spectral images. A 2-D averaging filter and binary image conversion, with appropriate threshold selection, are used to filter and enhance the images for the training and testing of the ANN classifier. The proposed novel VSI augments and provides the visual representation of the characteristic vibration spectral features in an image form. This provides enhanced spectral images for ANN training and thus leads to a highly robust fault classifier.
Description:
Industrial Internet of Things (IIoT) deploys edge devices to act as intermediaries between sensors and actuators and application servers or cloud services. Machine learning models have been widely used to thwart malware attacks in such edge devices. However, these models are vulnerable to adversarial attacks where attackers craft adversarial samples by introducing small perturbations to malware samples to fool a classifier to misclassify them as benign applications. Literature on deep learning networks proposes adversarial retraining as a defense mechanism where adversarial samples are combined with legitimate samples to retrain the classifier. However, existing works select such adversarial samples in a random fashion which degrades the classifier's performance. This work proposes two novel approaches for selecting adversarial samples to retrain a classifier. One, based on the distance from malware cluster center, and the other, based on a probability measure derived from a kernel based learning (KBL). Our experiments show that both of our sample selection methods outperform the random selection method and the KBL selection method improves detection accuracy by 6%. Also, while existing works focus on deep neural networks with respect to adversarial retraining, we additionally assess the impact of such adversarial samples on other classifiers and our proposed selective adversarial retraining approaches show similar performance improvement for these classifiers as well. The outcomes from the study can assist in designing robust security systems for IIoT applications.
Description:
Industrial Internet of Things (IIoT) deploys edge devices to act as intermediaries between sensors and actuators and application servers or cloud services. Machine learning models have been widely used to thwart malware attacks in such edge devices. However, these models are vulnerable to adversarial attacks where attackers craft adversarial samples by introducing small perturbations to malware samples to fool a classifier to misclassify them as benign applications. Literature on deep learning networks proposes adversarial retraining as a defense mechanism where adversarial samples are combined with legitimate samples to retrain the classifier. However, existing works select such adversarial samples in a random fashion which degrades the classifier's performance. This work proposes two novel approaches for selecting adversarial samples to retrain a classifier. One, based on the distance from malware cluster center, and the other, based on a probability measure derived from a kernel based learning (KBL). Our experiments show that both of our sample selection methods outperform the random selection method and the KBL selection method improves detection accuracy by 6%. Also, while existing works focus on deep neural networks with respect to adversarial retraining, we additionally assess the impact of such adversarial samples on other classifiers and our proposed selective adversarial retraining approaches show similar performance improvement for these classifiers as well. The outcomes from the study can assist in designing robust security systems for IIoT applications.
Description:
The Internet of Things (IoT) has facilitated services without human intervention for a wide range of applications, including continuous remote patient monitoring (RPM). However, the complexity of RPM architectures, the size of data sets generated and limited power capacity of devices make RPM challenging. In this paper, we propose a tier-based End to End architecture for continuous patient monitoring that has a patient centric agent (PCA) as its center piece. The PCA manages a blockchain component to preserve privacy when data streaming from body area sensors needs to be stored securely. The PCA based architecture includes a lightweight communication protocol to enforce security of data through different segments of a continuous, real time patient monitoring architecture. The architecture includes the insertion of data into a personal blockchain to facilitate data sharing amongst healthcare professionals and integration into electronic health records while ensuring privacy is maintained. The blockchain is customized for RPM with modifications that include having the PCA select a Miner to reduce computational effort, enabling the PCA to manage multiple blockchains for the same patient, and the modification of each block with a prefix tree to minimize energy consumption and incorporate secure transaction payments. Simulation results demonstrate that security and privacy can be enhanced in RPM with the PCA based End to End architecture.
Description:
The Internet of Things (IoT) has facilitated services without human intervention for a wide range of applications, including continuous remote patient monitoring (RPM). However, the complexity of RPM architectures, the size of data sets generated and limited power capacity of devices make RPM challenging. In this paper, we propose a tier-based End to End architecture for continuous patient monitoring that has a patient centric agent (PCA) as its center piece. The PCA manages a blockchain component to preserve privacy when data streaming from body area sensors needs to be stored securely. The PCA based architecture includes a lightweight communication protocol to enforce security of data through different segments of a continuous, real time patient monitoring architecture. The architecture includes the insertion of data into a personal blockchain to facilitate data sharing amongst healthcare professionals and integration into electronic health records while ensuring privacy is maintained. The blockchain is customized for RPM with modifications that include having the PCA select a Miner to reduce computational effort, enabling the PCA to manage multiple blockchains for the same patient, and the modification of each block with a prefix tree to minimize energy consumption and incorporate secure transaction payments. Simulation results demonstrate that security and privacy can be enhanced in RPM with the PCA based End to End architecture.
Description:
Wireless sensor networks (WSNs) will be an integral part of the future Internet of Things (loT) environment and generate large volumes of data. However, these data would only be of benefit if useful knowledge can be mined from them. A data mining framework for WSNs includes data extraction, storage and mining techniques, and must be efficient and dependable. In this paper, we propose a new type of behavioral pattern mining technique from sensor data called regularly frequent sensor patterns (RFSPs). RFSPs can identify a set of temporally correlated sensors which can reveal significant knowledge from the monitored data. A distributed data extraction model to prepare the data required for mining RFSPs is proposed, as the distributed scheme ensures higher availability through greater redundancy. The tree structure for RFSP is compact requires less memory and can be constructed using only a single scan through the dataset, and the mining technique is efficient with low runtime. Current mining techniques in the literature on sensor data employ a single memory-based sequential approach and hence are not efficient. Moreover, usage of the. MapReduce model for the distributed solution has not been explored extensively. Since MapReduce is becoming the de facto model for computation on large data, we also propose a parallel implementation of the RFSP mining algorithm, called RFSP on Hadoop (RFSP-H), which uses a MapReduce-based framework to gain further efficiency. Experiments conducted to evaluate the compactness and performance of the data extraction model, RFSP-tree and RFSP-H mining show improved results. (C) 2016 Elsevier Inc. All rights reserved.