Non-intrusion Events Research Articles

Run-time failure detection via non-intrusive event analysis in a large-scale cloud computing platform

Cloud computing systems fail in complex and unforeseen ways due to unexpected combinations of events and interactions among hardware and software components. These failures are especially problematic when they are silent, i.e., not accompanied by any explicit failure notification, hindering the timely detection and recovery.In this work, we propose an approach to run-time failure detection tailored for monitoring multi-tenant and concurrent cloud computing systems. The approach uses a non-intrusive form of event tracing, without manual changes to the system’s internals to propagate session identifiers (IDs), and builds a set of lightweight monitoring rules from fault-free executions. We evaluated the effectiveness of the approach in detecting failures in the context of the OpenStack cloud computing platform, a complex and “off-the-shelf” distributed system, by executing a campaign of fault injection experiments in a multi-tenant scenario. Our experiments show that the approach detects the failure with an F1 score (0.85) and accuracy (0.77) higher than the ones provided by the OpenStack failure logging mechanisms (0.53 and 0.50) and two non-session-aware run-time verification approaches (both lower than 0.15). Moreover, the approach significantly decreases the average time to detect failures at run-time (∼114 seconds) compared to the OpenStack logging mechanisms.

High-performance railway perimeter security system based on the inline time-division multiplexing fiber Fabry-Perot interferometric sensor array

This paper proposes a high-performance railway perimeter security system based on the inline time-division multiplexing fiber Fabry-Perot interferometric (TDM-FFPI) sensor array. A rectangular-pulse binary phase modulation (RPBPM) method is used for phase demodulation and a field programmable gate array (FPGA)-based acquisition system is adopted for receiving and processing the signals with the advantages of low complexity and high real-time capability in this system. Besides, a three-polarizer polarization diversity receiver (tr-PDR) is used to overcome the polarization-induced fading elimination. For the demodulated phase signals, a real-time and accurate two-level pattern recognition method based on the support vector machines (SVM) is employed for the recognition of different events (non-intrusion events including environment, rain and wind, and four intrusion events including shaking, cutting, climbing and knocking). The feasibility of the present system is verified experimentally. The average recognition accuracy is up to about 94.5% with the integrating time less than 0.2 s, which satisfies the high accuracy and high real-time requirements.

Surface intrusion event identification for subway tunnels using ultra-weak FBG array based fiber sensing.

A scheme is proposed for the identification of surface intrusion events, from signals detected by an ultra-weak fiber Bragg grating array in a subway tunnel. The spectral subtraction and the root mean square of the power spectral density are combined to extract event signals. The local characteristics-scale decomposition and the multi-scale permutation entropy are employed subsequently for feature extraction, which can improve the event recognition rate from the perspective of multi-scale analysis. Experimental demonstration verifies that the proposed scheme can identify four common events. Among the events, the discrete pulse construction and the continuous pulse construction on the ground surface are intrusion events, the subway train traveling in the tunnel and the lorry passing on the ground surface are non-intrusion events. The average recognition rate of 96.57% is achieved, which can satisfy actual application requirements.

Seasonal and interannual changes of the Kuroshio intrusion onto the East China Sea Shelf

AbstractAn advanced artificial neural network classification algorithm is applied to 20 years of multisatellite geostrophic velocity data to study the Kuroshio intrusion onto the southern shelf of the East China Sea. The results suggest that the on‐shelf intrusion may occur year round, but winter intrusion events are more frequent than summer events. Both stronger intrusion (which is closely correlated to surface heat flux gradient) and weaker intrusion (which is correlated to wind forcing) occur during wintertime, but the former dominates in late winter while the latter prevails in early winter. There is a significant year‐to‐year variation of the winter on‐shelf intrusion. Although on‐shelf intrusion is the major characteristic of the region during wintertime, seldom intrusion events have been identified in the winters of 1997–1998 and 2002–2003. Two conditions are responsible for the cause of the nonintrusion events. During the two nonintrusion winters, the upstream Kuroshio transport anomaly is insignificant (the Kuroshio is not weakened) and no significant winter cooling off northeast Taiwan. Thus, the Kuroshio tends to flow along the shelf break following the 200 m isobath and on‐shelf intrusion ceases. Those two nonintrusion events take place during the winters when both the Niño 3.4 index and the PDO index are large and positive.

Real-time Distributed Fiber Optic Sensor for Security Systems: Performance, Event Classification and Nuisance Mitigation

The success of any perimeter intrusion detection system depends on three important performance parameters: the probability of detection (POD), the nuisance alarm rate (NAR), and the false alarm rate (FAR). The most fundamental parameter, POD, is normally related to a number of factors such as the event of interest, the sensitivity of the sensor, the installation quality of the system, and the reliability of the sensing equipment. The suppression of nuisance alarms without degrading sensitivity in fiber optic intrusion detection systems is key to maintaining acceptable performance. Signal processing algorithms that maintain the POD and eliminate nuisance alarms are crucial for achieving this. In this paper, a robust event classification system using supervised neural networks together with a level crossings (LCs) based feature extraction algorithm is presented for the detection and recognition of intrusion and non-intrusion events in a fence-based fiber-optic intrusion detection system. A level crossings algorithm is also used with a dynamic threshold to suppress torrential rain-induced nuisance alarms in a fence system. Results show that rain-induced nuisance alarms can be suppressed for rainfall rates in excess of 100 mm/hr with the simultaneous detection of intrusion events. The use of a level crossing based detection and novel classification algorithm is also presented for a buried pipeline fiber optic intrusion detection system for the suppression of nuisance events and discrimination of intrusion events. The sensor employed for both types of systems is a distributed bidirectional fiber-optic Mach-Zehnder (MZ) interferometer.

Identification Technique for the Intrusion of Airport Enclosure Based on Double Mach-Zehnder Interferometer

In view of the high rate of false alarm problems on the distributed optical fiber sensing technique applied in enclosure, a sensor is proposed based on double Mach-Zehnder interferometer and a method is needed urgently to identify intrusion accurately. This involves eliminating non-intrusion events caused by operating environments from intrusion events. To address the issue, this paper states the tensile strain principle of optical fiber and combines autocorrelation with frequency-energy distribution of signals to extract vibration characteristics. Finally, an identification technique is presented based on vibration theory to distinguish events. The results of simulation show that the algorithm can eliminate vibration characteristic of non-intrusion events so as to minimize the nuisance alarm rate and make prepare for the following recognition of intrusion.