No abstract

Followed by the 9/11 attacks in 2001 and the subsequent events, terrorism and other asymmetrical threat situations became increasingly important for security-related efforts of most western societies. In a similar period, the development of data gathering and analysis techniques especially using the methods of machine learning has made rapid progress. Aiming to utilize this development, this paper employs artificial neural networks (ANN) for longterm time series prediction of terrorist event data. A major focus of the paper lies on the specific use of convolutional neural networks (CNN) for this task, as well as the comparison to the performance of classical methods for (long-term) time series prediction. As the database like Global terrorist database (GTD) and Fraunhofer's terrorist event database (TED) are not extensive enough to train a deep learning method, a simple toy model for the generation of time-series data from one or more terrorist groups with defined properties is established. Metrics for comparison of the different approaches are collected and discussed, and a customized sliding window metric (SWM) is introduced. The study shows the principle applicability of CNNs for this task and offers constraints as well as possible extensions for future studies. Based on these results, continuation and further extension of data collection efforts and ML optimization techniques are encouraged.

Organizational and technical approaches have proven successful in increasing the performance and preventing risks at socio-technical systems at all scales. Nevertheless, damaging events are often unavoidable due to a wide and dynamic threat landscape and enabled by the increasing complexity of modern systems. For overall performance and risk control at the system level, resilience can be a versatile option, in particular for reducing resources needed for system development, maintenance, reuse, or disposal. This paper presents a framework for a resilience assessment and management process that builds on existing risk management practice before, during, and after potential and real events. It leverages tabular and matrix correlation methods similar as standardized in the field of risk analysis to fulfill the step-wise resilience assessment and management for critical functions of complex systems. We present data needs for the method implementation and output generation, in particular regarding the assessment of threats and the effects of counter measures. Also included is a discussion of how the results contribute to the advancement of functional risk control and resilience enhancement at system level as well as related practical implications for its efficient implementation. The approach is applied in the domains telecommunication, gas networks, and indoor localization systems. Results and implications are further discussed.

Time difference of arrival (TDOA) based indoor ultrasound localization systems are prone to multiple disruptions and demand reliable, and resilient position accuracy during operation. In this challenging context, a missing link to evaluate the performance of such systems is a simulation approach to test their robustness in the presence of disruptions. This approach cannot only replace experiments in early phases of development but could also be used to study susceptibility, robustness, response, and recovery in case of disruptions. The paper presents a simulation framework for a TDOA-based indoor ultrasound localization system and ways to introduce different types of disruptions. This framework can be used to test the performance of TDOA-based localization algorithms in the presence of disruptions. Resilience quantification results are presented for representative disruptions. Based on these quantities, it is found that localization with arc-tangent cost function is approximately 30% more resilient than the linear cost function. The simulation approach is shown to apply to resilience engineering and can be used to increase the efficiency and quality of indoor localization methods.