<p>The use of lower airspace for air taxi and cargo applications opens up exciting prospects for futuristic Unmanned Aircraft Systems (UAS). However, ensuring the safety and security of these UAS within densely populated urban areas presents significant challenges. Most modern aircraft systems, whether unmanned or otherwise, rely on the Global Navigation Satellite System (GNSS) as a primary sensor for navigation. From satellite navigations point of view, the dense urban environment compromises positioning accuracy due to signal interference, multipath effects, etc. Furthermore, civilian GNSS receivers are susceptible to spoofing attacks since they lack encryption capabilities. Therefore, in this thesis, we focus on examining the safety and cybersecurity assurance of UAS in dense urban environments, from both theoretical and experimental perspectives. </p> <p>To facilitate the verification and validation of the UAS, the first part of the thesis focuses on the development of a realistic GNSS sensor emulation using a Gazebo plugin. This plugin is designed to replicate the complex behavior of the GNSS sensor in urban settings, such as multipath reflections, signal blockages, etc. By leveraging the 3D models of the urban environments and the ray-tracing algorithm, the plugin predicts the spatial and temporal patterns of GNSS signals in densely populated urban environments. The efficacy of the plugin is demonstrated for various scenarios including routing, path planning, and UAS cybersecurity. </p> <p>Subsequently, a robust state estimation algorithm for dynamical systems whose states can be represented by Lie Groups (e.g., rigid body motion) is presented. Lie groups provide powerful tools to analyze the complex behavior of non-linear dynamical systems by leveraging their geometrical properties. The algorithm is designed for time-varying uncertainties in both the state dynamics and the measurements using the log-linear property of the Lie groups. When unknown disturbances are present (such as GNSS spoofing, and multipath effects), the log-linearization of the non-linear estimation error dynamics results in a non-linear evolution of the linear error dynamics. The sufficient conditions under which this non-linear evolution of estimation error is bounded are derived, and Lyapunov stability theory is employed to design a robust filter in the presence of an unknown-but-bounded disturbance. </p>

Teleworking has become a significant aspect of working life, especially after the outbreak of the COVID-19 pandemic, which accelerated the trend of teleworking. However, this shift has increased the risk of cyber threats and security risks. Despite organizations' efforts to strengthen cybersecurity, a significant risk remains, with employees posing one of the main security risks in the form of human error and mistakes. Previous research highlights that employees tend to exhibit lower levels of cybersecurity awareness and are more likely to perform riskful actions when working remotely compared to working in the office. However, recent research has shown the opposite, where employees are more conscious of cybersecurity awareness and more likely to apply security-based precaution measures during remote work compared to office work. In light of these research findings, this study focuses on examining how IT-security experts perceive cybersecurity when working remotely compared to working in the office. To explore this, the study has, through qualitative mapping, conducted semi-structured interviews with a theoretical basis in Protection Motivation Theory (PMT). Overall, the study showed that IT- security experts perceive cybersecurity as more manageable when working in the office compared to remote work, with an increased awareness of the importance of the human factor. / Distansarbete har blivit en betydande aspekt av arbetslivet, särskilt efter utbrottet av Covid-19- pandemin, vilket accelererade trenden med distansarbete. Denna omställning har emellertid ökat risken för cyberhot och säkerhetsrisker. Trots organisationers insatser för att stärka cybersäkerheten kvarstår en betydande risk, då anställda utgör en av de främsta säkerhetsriskerna i form av mänskliga fel och misstag. Tidigare forskning framhäver att anställda ofta är mindre säkerhetsmedvetna och mer benägna att utföra riskfyllda handlingar när de arbetar på distans jämfört med arbete på kontoret. Däremot har senare forskning visat motsatsen, där anställda är mer säkerhetsmedvetna och mer benägna att vidta säkerhetsåtgärder under distansarbete jämfört med arbete på kontoret. Mot bakgrund till dessa forskningsresultat, fokuserar denna studie på att undersöka hur IT-säkerhetsexperter uppfattar cybersäkerhet vid distansarbete jämfört med arbete på kontoret. För att utforska detta har studien, genom kvalitativ kartläggning, genomfört semistrukturerade intervjuer med teoretisk grund i Protection Motivation Theory (PMT). Sammantaget visade studien på att IT-säkerhetsexperter uppfattar cybersäkerhet som mer hanterbar vid arbete på kontoret jämfört med distansarbete, med en ökad medvetenhet om den mänskliga faktorns betydelse.

The banking sector is experiencing the rise of several new types of innovations and trends. For instance, increased use of Artificial Intelligence (AI) to streamline day-to-day activities. These trends are, e.g., influenced by an increased frequency of cyber attacks, the emergence of newly proposed regulations such as DORA and the AI Act, and the improving computational capabilities of AI-driven systems. The full impact these trends will have on the sector is yet to be realized. The sector is diverse and deeply integrated within society, meaning that it is critical to understand how actors mitigate the risks associated with the implementation of AI. This study analyzes how organizations can mitigate the risks involved with this implementation and how it affects the risk management process. To examine the implementation of AI in the banking sector, the study conducted semi-structured interviews with twelve respondents with expertise in AI, security, or the banking sector. The study used two theoretical frameworks to analyze the data. The first framework, the Dynamic Risk Management Framework, was used to analyze changes in the risk management process based on its unique position within society. The second framework, the Multi-Level Perspective, gave the study a holistic understanding of the impact of AI as a driver of a socio-technical shift. The results show that the implementation of AI leads to a set of new risks. These risks are primarily organizational and regulatory and will lead to a revision in how actors classify risks. The constant evolution of AI also means that products must be reviewed periodically, changing how actors view the risk management process. Additionally, the results identify a lack of knowledge regarding both AI and security within the sector. Consequently, the organization will have to change its structure to accommodate interactions between different competencies. To succeed in implementing AI, meet the regulatory demands and mitigate unintended bias when developing AI, the study concludes that these competencies must create a shared terminology to communicate efficiently. In conclusion, the study contributes to a growing field regarding business applications of AI by creating a holistic understanding of aspects impacting the risk management process in banking. The findings result in a series of recommended actions for organizations that aim to implement AI in their businesses. Further research is recommended to understand the long-term effects of these actions. Future in depth analyses could validate the results of this study and further investigate the development of AI as a business tool. / Banksektorn upplever en uppåtgående trend när det kommer till användandet av innovation. Ett exempel på detta är användningen av artificiell intelligens (AI) för att effektivisera bankens dagliga aktiviteter. Denna trend beror på flertalet olika faktorer, bland annat den ökade frekvensen av cyberattacker mot bankaktörer, de nya föreslagna förordningarna DORA och AI Act, och att AI-drivna systems kapacitet förbättras. Däremot har inte effekten av AI på sektorn ännu realiserats till fullo. Banksektorn har en unik position i samhället och dess aktörer har många olika utmaningar, vilket innebär att det är avgörande att förstå hur aktörerna hanterar de risker som uppstår i samband med implementeringen av AI. Denna studie analyserar hur organisationer kan minska riskerna med denna implementering och hur AI påverkar riskhanteringsprocessen. För att undersöka implementeringen har studien genomfört semistrukturerade intervjuer med tolv intervjuobjekt med expertis inom AI, säkerhet eller banksektorn. För att analysera den framtagna datan har studien använt två teoretiska ramverk. Det första ramverket, som kallas Dynamic Risk Management Framework, användes för att analysera förändringar i riskhanteringsprocessen med tanke på banksektorns unika position i samhället. Det andra ramverket, som kallas Multi-Level Perspective, undersökte AI som en drivkraft mot ett sociotekniskt skifte och gav därmed studien en helhetsbildav effekten av AI. Resultaten visar att implementeringen av AI leder till en rad nya risker. Dessa risker är i första hand organisatoriska och regulatoriska. Då dessa är relativt nya, måste organisationer se över hur de klassificerar risker. AI utvecklas kontinuerligt, vilket innebär att produkterna och deras effekt måste ses över regelbundet. Dessutom identifierar resultaten en brist på kunskap om både AI och säkerhet inom banksektorn. För att tillgodose nya kompetenser och underlätta interaktionerna mot existerande kompetenser kommer organisationer behöva struktureras om. Studien drar slutsatsen att en lyckad AI implementering, där de regulatoriska kraven möts och utveckling av AI är fri från oavsiktliga fördomar och diskriminering, kräver en rad förändringar. Organisationen måste kunna kommunicera effektivt, vilket kräver att alla pratar samma språk och använder samma terminologi. Sammanfattningsvis bidrar studien till ett växande akademiskt område gällande affärstillämpningar av AI genom att skapa en helhetsbild över vilka aspekter som påverkar riskhanteringsprocessen inom bankverksamhet. Denna summering har resulterat i en rad åtgärder verksamheter som strävar efter att implementera AI rekommenderas att ta. Framtida studier rekommenderas däremot att undersöka de långsiktiga effekterna av dessa åtgärder. Genom att utföra djupgående analyser kanframtida studier inte bara validera denna studies resultat, de kan också förbättra förståelsen för hur AI som ett affärsverktyg kan komma att utvecklas.

No description available.

No description available.

With the exponential proliferation of data, there has been a surge in data generation fromdiverse sources, including social media platforms, websites, mobile devices, and sensors.However, not all data is readily visible or accessible to the public, leading to the emergence ofthe concept known as "dark data." This type of data can exist in structured or unstructuredformats and can be stored in various repositories, such as databases, log files, and backups.The reasons behind data being classified as "dark" can vary, encompassing factors such as limited awareness, insufficient resources or tools for data analysis, or a perception ofirrelevance to current business operations. This research employs a qualitative research methodology incorporating audio/videorecordings and personal interviews to gather data, aiming to gain insights into individuals'understanding of the risks associated with dark data and their behaviors concerning thesharing of personal information online. Through the thematic analysis of the collected data,patterns and trends in individuals' risk perceptions regarding dark data become evident. The findings of this study illuminate the multiple dimensions of individuals' risk perceptions andt heir influence on attitudes towards sharing personal information in online contexts. Theseinsights provide valuable understanding of the factors that shape individuals' decisionsconcerning data privacy and security in the digital era. By contributing to the existing body ofknowledge, this research offers a deeper comprehension of the interplay between dark datarisks, individuals' perceptions, and their behaviors pertaining to online information sharing.The implications of this study can inform the development of strategies and interventionsaimed at fostering informed decision-making and ensuring personal safety in an increasinglydata-centric world

<p>Can a critical industrial system, such as a nuclear reactor, be made self-aware and cognizant of its operational history? Can it alert authorities covertly to malicious intrusion without exposing its  defense  mechanisms?  What  if  the  intruders  are  highly  knowledgeable  adversaries,  or  even  insiders that may have designed the system? This thesis addresses these research questions through a novel physical process defense called Covert Cognizance (C2). </p> <p>C2  serves  as  a  last  line  of  defense  to  industrial  systems  when  existing  information  and  operational technology defenses have been breached by advanced persistent threat (APT) actors or insiders. It is an active form of defense that may be embedded in an existing system to induce intelligence,  i.e.,  self-awareness,  and  make  various subsystems  aware  of  each  other.  It  interacts with the system at the process level and provides an additional layer of security to the process data therein without the need of a human in the loop. </p> <p>The C2 paradigm is  founded on two core requirements – zero-impact and zero-observability. Departing from contemporary active defenses, zero-impact requires a successful implementationto leave no footprint on the system ensuring identical operation while zero-observability requires that the embedding is immune to pattern-discovery algorithms.  In other words, a third-party such as  a  malicious  intruder  must  be  unable  to  detect  the  presence  of  the  C2  defense  based  on  observation of the process data, even when augmented by machine learning tools that are adept at pattern discovery. </p> <p>In the present work, nuclear reactor simulations are embedded with the C2 defense to induce awareness across subsystems and defend them against highly knowledgeable adversaries that have bypassed existing safeguards such as model-based defenses.  Specifically, the subsystems are made aware  of  each  other  by  embedding  critical information from  the  process  variables  of  one sub-module  along  the  noise of  the  process  variables  of  another,  thus  rendering  the  implementation  covert and  immune  to  pattern  discovery.   The  implementation  is  validated  using  generative adversarial  nets,  representing  a  state-of-the-art  machine  learning  tool,  and  statistical  analysis  of  the  reactor  states,  control  inputs,  outputs  etc. The  work  is  also  extended  to  data  masking  applications  via  the  deceptive  infusion  of  data  (DIOD)  paradigm.  Future  work  focuses  on  the  development of automated C2 modules for “plug ‘n’ play” deployment onto critical infrastructure and/or their digital twins.</p>

<p>With the fast spread of machine learning techniques, sharing and adopting public deep neural networks become very popular. As deep neural networks are not intuitive for human to understand, malicious behaviors can be injected into deep neural networks undetected. We call it trojan attack or backdoor attack on neural networks. Trojaned models operate normally when regular inputs are provided, and misclassify to a specific output label when the input is stamped with some special pattern called trojan trigger. Deploying trojaned models can cause various severe consequences including endangering human lives (in applications like autonomous driving). Trojan attacks on deep neural networks introduce two challenges. From the attacker's perspective, since the training data or training process is usually not accessible to the attacker, the attacker needs to find a way to carry out the trojan attack without access to training data. From the user's perspective, the user needs to quickly scan the online public deep neural networks and detect trojaned models.</p> <p>We try to address these challenges in this dissertation. For trojan attack without access to training data, We propose to invert the neural network to generate a general trojan trigger, and then retrain the model with reverse-engineered training data to inject malicious behaviors to the model. The malicious behaviors are only activated by inputs stamped with the trojan trigger. To scan and detect trojaned models, we develop a novel technique that analyzes inner neuron behaviors by determining how output activation change when we introduce different levels of stimulation to a neuron. A trojan trigger is then reverse-engineered through an optimization procedure using the stimulation analysis results, to confirm that a neuron is truly compromised. Furthermore, for complex trojan attacks, we propose a novel complex trigger detection method. It leverages a novel symmetric feature differencing method to distinguish features of injected complex triggers from natural features. For trojan attacks on NLP models, we propose a novel backdoor scanning technique. It transforms a subject model to an equivalent but differentiable form. It then inverts a distribution of words denoting their likelihood in the trigger and applies a novel word discriminativity analysis to determine if the subject model is particularly discriminative for the presence of likely trigger words.</p>

Extensive digitization is currently underway in the railway sector, which has resulted in several benefits and improvements, but also challenges. The increased use of digital technologies increases the risks of vulnerabilities and susceptibility to cyberattacks. The effects of a cyber attack can have significant consequences on operations such as financial losses and damaged reputations, or in the worst-case scenario, devastating consequences where the lives of passengers are endangered. With the ongoing digitalization of the railways and the growing concern for cybersecurity, stakeholders in the sector have identified the need to systematically understand the risks of digitization related to cybersecurity and safety. Therefore, this study aims to identify and communicate these risks using system dynamics modeling. This study evaluated how actors in the railway sector reason about risks in the railway, how safety and cybersecurity are related, and new risks associated with digitalization and cybersecurity that have not been mentioned in the literature before. A qualitative study was conducted to answer the research question. Ten actors from different parts of the railway value chain were interviewed, and secondary data was collected from articles and reports within the area of cybersecurity and the railways. The results revealed a connection between cybersecurity and safety which could be seen through the chain of consequences that can arise from a cyberattack and in the event of loss of data availability and integrity. Based on this, core elements of the system and the relationships between them could be identified, from which the causal loop diagram (CLD) was constructed. New risks that were identified were the safety culture that permeates the railway industry, unclear areas of responsibility that are a result of deregulation in the Swedish railway sector, and competitiveness. Insights from the system dynamic model identified a reinforcing loop telling a causal story that shows that low cybersecurity priorities could lead to decreased safety on the railway. This further demonstrates the usefulness of identifying and communicating risks using system dynamics modeling. / En omfattande digitalisering pågår just nu inom järnvägssektorn vilket dels har resulterat i en mängd fördelar och förbättringar, men också utmaningar. Den ökade användningen av digitala teknologier ökar risker för sårbarheter samt mottagligheten av cyberattacker. Effekterna av en cyberattack kan ha betydande konsekvenser på verksamheten så som ekonomiska förluster och skadat rykte, eller också i värsta fall förödande konsekvenser där passagerarnas liv sätts i fara. I och med den pågående digitaliseringen av järnvägen och den ökade oron för cybersäkerhet har intressenter inom sektorn identifierat behovet av att på ett systematisk sätt förstå riskerna med digitalisering relaterade till cybersäkerhet och säkerhet (safety). Denna studie syftar därför till att identifiera och kommunicera dessa risker genom att använda systemdynamisk modellering. Studien utvärderade hur aktörer i järnvägens värdekedja resonerade kring risker i järnvägen, hur säkerhet och cybersäkerhet var relaterat, samt vad det finns för nya risker relaterade till digitaliseringen och cybersäkerhet som inte nämnts i litteraturen tidigare. En kvalitativ studie genomfördes för att svara på frågeställningen. Tio aktörer från olika delar av järnvägssektorns värdekedja intervjuades, och sekundärdata samlades in från artiklar och rapporter inom cybersäkerhet och järnvägen. Resultaten visade att det finns en koppling mellan cybersäkerhet och säkerhet (safety) som syns i den kedja av konsekvenser som kan uppstå från en cyberattack vid förlust av datas tillgänglighet och integritet. Utifrån detta kunde nyckelelement i systemet samt relationerna mellan dessa identifieras, och baserat på detta konstruerades vidare ett causal loop diagram (CLD). Nya risker som identifierades var säkerhetskulturen som genomsyrar järnvägsbranschen, oklara ansvarsområden som är ett resultat av den svenska järnvägens avreglering, samt konkurrenskraft. Insikter från den systemdynamiska modellen identifierade en förstärkande loop som berättar en orsakshistoria som visar att låg prioritering av cybersäkerhet kan leda till minskad säkerhet på järnvägen. Vidare demonstrerar detta nyttan av att identifiera och kommunicera risker med hjälp av systemdynamsik modellering.

As technology and Internet use continue to grow globally, cybersecurity has become an increasingly important topic to ensure the safety of individual consumers online. It is a growing cause for concern that cyber attacks are on the rise, as has been the case during the global pandemic. Cybersecurity awareness among the general population is of utmost importance to mitigate and prevent cyber attacks. One of the primary purposes of this study was to identify the existence and measure the extent of knowledge gaps between experts and non-experts, within the field of cybersecurity. This was done by evaluating how experts and non-experts rate the severity of different vulnerabilities. Through extensive collaboration with a technology startup within the f ield of digital design and e-commerce, four vulnerabilities were identified in the startup’s IT environment. These vulnerabilities were simplified and described in four separate scenarios, for the sake of being digestible for a nonexpert audience. The scenarios were used to construct an extensive survey, which asked participants how they would rate the severity of each scenario, using a 1-5 Likert scale. Severity was measured using three vulnerability metrics, including Attack Complexity (AC), Remediation Level (RL) and Confidentiality (C). In many cases, experts and non-experts rated the severity of the studied vulnerabilities similarly. However, the results of this study primarily showed that there doesexiststatistically significant differences intheperceivedseverity between the groups. Using expert responses as a baseline, it was possible to identify for which metrics and in which contexts the lack of cybersecurity awareness existed among non-experts. This lack of awareness was equated with the existence of a knowledge gap. For future work, it would be of great interest to further analyze the extent of and how to reduce this knowledge gap. / I takt med att den internetbaserade teknologiska användningen fortsätter att öka på global skala blir cybersäkerhet desto viktigare för att skydda konsumenter från cyberangrepp online. Under pandemins gång har antalet cyberattacker ökat markant och detta är högst oroväckande. Allmänhetens medvetenhet kring cybersäkerhet är av yttersta vikt för att mitigera och förhindra cyberattackerna. Ett av dem främsta syftena med denna studie var att undersöka eventuella kunskapsskillnader mellan experter och ickeexperter inom cybersäkerhetsområdet. Detta undersöktes genom att evaluera hur experter och icke-experter bedömer allvaret i olika cybersäkerhetsbrister. Genom ett gediget samarbete med en startup inom digital design och ehandel, identifierades fyra separata säkerhetsbrister i företagets IT system. Dessa brister förklarades på en översiktlig nivå i olika scenarier, för att underlätta förståelsen bland icke-experter. Dessa scenarier användes därefter till att utveckla en omfattande undersökningsenkät, vars syfte var att undersöka hur deltagare bedömer allvaret i scenarierna utifrån en 1-5 Likertskala. Allvaret beräknades genom att deltagarna bedömde tre olika sårbarhetsaspekter för varje scenario: cyberattackens komplexitet (AC), åtgärdhetens komplexitet (RL) och konfidentialitet (C). I många fall gjorde experter och icke-experter en likvärdig bedömning kring allvaret i dem undersökta cybersäkerhetsbristerna. Emellertid fanns det i andra sammanhang uppenbara och statistiskt säkerställda skillnader i bedömningarna mellan grupperna. Genom att använda experternas bedömning som måttstock, visade studien i vilka sammanhang och för vilka sårbarhetsaspekter icke-experter saknade en gedigen medvetenhet kring. Denna bristfälliga medvetenhet likställdes med att en kunskapslucka existerade mellan grupperna. Inför framtida forskning är det av stort intresse att undersöka omfattningen av och metoder att reducera denna kunskapslucka.