Fault injection attacks are a powerful technique that intentionally induces faults during computations to leak secret information. This thesis studies the fault injection attack techniques. The thesis first categorizes various fault attack methods by fault model and fault analysis and gives examples of the various fault attacks on symmetric key cryptosystems and public key cryptosystems. The thesis then demonstrates fault injection attacks on RSA-CRT and constant time CSIDH. The fault attack consists of two main components: fault modeling, which examines methods for injecting faults in a target device, and fault analysis, which analyzes the resulting faulty outputs to deduce secrets in each cryptosystem. The thesis aims to provide a comprehensive survey on fault attack research, directions for further study on securing real-world cryptosystems against fault injection attacks, testing fault injection attacks with RSA-CRT, and demonstrate and evaluate fault injection attacks on constant time CSIDH. / Master of Science / Fault injection attacks are attacks where the attackers intentionally induce the fault in the device during the operation to obtain or recover secret information. The induced fault will impact the operation and cause the faulty output, providing the information to attackers. Many cryptographic algorithms and devices have been proven vulnerable to fault injection attacks. Cryptography is essential nowadays, as it is used to secure and protect confidential data. If the cryptosystem is broken, many system today will be compromised. Thus, this thesis focus on the fault injection attacks on the cryptosystems. This thesis introduces the background of fault injection attacks, categorizes them into different types, and provides examples of the attacks on cryptosystems. The thesis studies how the attacks work, including how the attack induces the fault in the device and how the attack analyzes the fault output they obtained. Specifically, I examine how these attacks affect two commonly used encryption methods: symmetric key cryptography and public key cryptography. Additionally, I implement the fault injection attack on RSA-CRT and emph{Commutative Supersingular Isogeny Diffie-Hellman}~(CSIDH). This research aims to understand the potential attack method on different cryptosystems and can explore mitigation or protection in the future.

Cyber Physical System (CPS) is a system where cyber and physical components work in a complex co-ordination to provide better performance. By exploiting the communication infrastructure among the sensors, actuators, and control systems, attackers may compromise the security of a CPS. In this dissertation, security measures for different types of attacks/ faults in two CPSs, water supply system (WSS) and smart grid system, are presented. In this context, I also present my study on energy management in Smart Grid. The techniques for detecting attacks/faults in both WSS and Smart grid system adopt Kalman Filter (KF) and χ2 detector. The χ2 -detector can detect myriad of system fault- s/attacks such as Denial of Service (DoS) attack, short term and long term random attacks. However, the study shows that the χ2 -detector is unable to detect the intelligent False Data Injection attack (FDI). To overcome this limitation, I present a Euclidean detector for smart grid which can effectively detect such injection attacks. Along with detecting attack/faults I also present the isolation of the attacked/faulty nodes for smart grid. For isolation the Gen- eralized Observer Scheme (GOS) implementing Kalman Filter is used. As GOS is effective in isolating attacks/faults on a single sensor, it is unable to isolate simultaneous attacks/faults on multiple sensors. To address this issue, an Iterative Observer Scheme (IOS) is presented which is able to detect attack on multiple sensors. Since network is an integral part of the future CPSs, I also present a scheme for pre- serving privacy in the future Internet architecture, namely MobilityFirst architecture. The proposed scheme, called Anonymity in MobilityFirst (AMF), utilizes the three-tiered ap- proach to effectively exploit the inherent properties of MF Network such as Globally Unique Flat Identifier (GUID) and Global Name Resolution Service (GNRS) to provide anonymity to the users. While employing new proposed schemes in exchanging of keys between different tiers of routers to alleviate trust issues, the proposed scheme uses multiple routers in each tier to avoid collaboration amongst the routers in the three tiers to expose the end users.

Buffer overflow vulnerabilities continue to prevail and the sophistication of attacks targeting these vulnerabilities is continuously increasing. As a successful attack of this type has the potential to completely compromise the integrity of the targeted host, early detection is vital. This thesis examines generic approaches for detecting executable payload attacks, without prior knowledge of the implementation of the attack, in such a way that new and previously unseen attacks are detectable. Executable payloads are analysed in detail for attacks targeting the Linux and Windows operating systems executing on an Intel IA-32 architecture. The execution flow of attack payloads are analysed and a generic model of execution is examined. A novel classification scheme for executable attack payloads is presented which allows for characterisation of executable payloads and facilitates vulnerability and threat assessments, and intrusion detection capability assessments for intrusion detection systems. An intrusion detection capability assessment may be utilised to determine whether or not a deployed system is able to detect a specific attack and to identify requirements for intrusion detection functionality for the development of new detection methods. Two novel detection methods are presented capable of detecting new and previously unseen executable attack payloads. The detection methods are capable of identifying and enumerating the executable payload’s interactions with the operating system on the targeted host at the time of compromise. The detection methods are further validated using real world data including executable payload attacks.

abstract: Wireless sensor networks (WSN) and the communication and the security therein have been gaining further prominence in the tech-industry recently, with the emergence of the so called Internet of Things (IoT). The steps from acquiring data and making a reactive decision base on the acquired sensor measurements are complex and requires careful execution of several steps. In many of these steps there are still technological gaps to fill that are due to the fact that several primitives that are desirable in a sensor network environment are bolt on the networks as application layer functionalities, rather than built in them. For several important functionalities that are at the core of IoT architectures we have developed a solution that is analyzed and discussed in the following chapters. The chain of steps from the acquisition of sensor samples until these samples reach a control center or the cloud where the data analytics are performed, starts with the acquisition of the sensor measurements at the correct time and, importantly, synchronously among all sensors deployed. This synchronization has to be network wide, including both the wired core network as well as the wireless edge devices. This thesis studies a decentralized and lightweight solution to synchronize and schedule IoT devices over wireless and wired networks adaptively, with very simple local signaling. Furthermore, measurement results have to be transported and aggregated over the same interface, requiring clever coordination among all nodes, as network resources are shared, keeping scalability and fail-safe operation in mind. Furthermore ensuring the integrity of measurements is a complicated task. On the one hand Cryptography can shield the network from outside attackers and therefore is the first step to take, but due to the volume of sensors must rely on an automated key distribution mechanism. On the other hand cryptography does not protect against exposed keys or inside attackers. One however can exploit statistical properties to detect and identify nodes that send false information and exclude these attacker nodes from the network to avoid data manipulation. Furthermore, if data is supplied by a third party, one can apply automated trust metric for each individual data source to define which data to accept and consider for mentioned statistical tests in the first place. Monitoring the cyber and physical activities of an IoT infrastructure in concert is another topic that is investigated in this thesis. / Dissertation/Thesis / Doctoral Dissertation Electrical Engineering 2017

The back-end database provides accessible and structured storage for each web application's big data internet web traffic exchanges stemming from cloud-hosted web applications to the Internet of Things (IoT) smart devices in emerging computing. Structured Query Language Injection Attack (SQLIA) remains an intruder's exploit of choice to steal confidential information from the database of vulnerable front-end web applications with potentially damaging security ramifications. Existing solutions to SQLIA still follows the on-premise web applications server hosting concept which were primarily developed before the recent challenges of the big data mining and as such lack the functionality and ability to cope with new attack signatures concealed in a large volume of web requests. Also, most organisations' databases and services infrastructure no longer reside on-premise as internet cloud-hosted applications and services are increasingly used which limit existing Structured Query Language Injection (SQLI) detection and prevention approaches that rely on source code scanning. A bio-inspired approach such as Machine Learning (ML) predictive analytics provides functional and scalable mining for big data in the detection and prevention of SQLI in intercepting large volumes of web requests. Unfortunately, lack of availability of robust ready-made data set with patterns and historical data items to train a classifier are issues well known in SQLIA research applying ML in the field of Artificial Intelligence (AI). The purpose-built competition-driven test case data sets are antiquated and not pattern-driven to train a classifier for real-world application. Also, the web application types are so diverse to have an all-purpose generic data set for ML SQLIA mitigation. This thesis addresses the lack of pattern-driven data set by deriving one to predict SQLIA of any size and proposing a technique to obtain a data set on the fly and break the circle of relying on few outdated competitions-driven data sets which exist are not meant to benchmark real-world SQLIA mitigation. The thesis in its contributions derived pattern-driven data set of related member strings that are used in training a supervised learning model with validation through Receiver Operating Characteristic (ROC) curve and Confusion Matrix (CM) with results of low false positives and negatives. We further the evaluations with cross-validation to have obtained a low variance in accuracy that indicates of a successful trained model using the derived pattern-driven data set capable of generalisation of unknown data in the real-world with reduced biases. Also, we demonstrated a proof of concept with a test application by implementing an ML Predictive Analytics to SQLIA detection and prevention using this pattern-driven data set in a test web application. We observed in the experiments carried out in the course of this thesis, a data set of related member strings can be generated from a web expected input data and SQL tokens, including known SQLI signatures. The data set extraction ontology proposed in this thesis for applied ML in SQLIA mitigation in the context of emerging computing of big data internet, and cloud-hosted services set our proposal apart from existing approaches that were mostly on-premise source code scanning and queries structure comparisons of some sort.

<p>U ovoj doktorskoj disertaciji predložena je nova metoda za detekciju malicioznih napada injektiranjem loših merenja na elektroenergetski sistem. Predloženi algoritam baziran je na sinergiji statičke i dinamičke estimacije stanja, i u stanju je da detektuje ovaj tip napada u realnom vremenu, za najkritičniji scenario gde napadač ima potpuno znanje o sistemu, i neograničen pristup resursima.</p> / <p>This PhD thesis proposes a novel method for detection of malicious false data<br />injection attacks on power system. The proposed algorithm is based on<br />synergy between static and dynamic state estimators, and is capable of<br />detecting the forementioned attacks in real time, for the most critical scenarios,<br />where an attacker has complete knowledge about the compromised power<br />system and unlimited resources to stage an attack.</p>

Web applications constitute an essential part of our daily lives, providing us access to significant online services and information. Despite their advantages, they are also vulnerable to security threats, particularly SQL injection attacks. SQL injection is a vulnerability that arises when an attacker inserts malicious SQL queries through user input parameters in a web application. This attack can have severe consequences, such as exposing sensitive information. The purpose of this study is to investigate and analyze the security of web applications against SQL injection attacks. This is achieved by examining SQL injection techniques, their impact on security and integrity, as well as the most common protective solutions. The goal of the study is to enhance and improve the security of applications and protect users from potential security risks. To achieve this, a combination of literature study and practical investigations is conducted. A literature review is performed to identify SQL injection techniques, security risks, and the most prevalent protective solutions. Subsequently, these factors are evaluated and analyzed to determine the effectiveness of the techniques using the penetration testing tool SQLmap. The results of the study indicate that the most common and effective attack techniques are Inband SQL injection and Inferential SQL injection. These techniques can have severe implications for users, businesses, and society at large, such as unauthorized access to protected data, data manipulation in databases, and the compromise of confidentiality and data integrity. To safeguard against such attacks, it is crucial to employ defensive coding practices, including the use of prepared statements with parameterized queries and input validation. However, manual implementation remains challenging. A combination of automated prevention techniques and best coding practices should be employed to ensure a reliable database protected against SQL injections, even concerning stored procedures that are difficult to prevent with existing automated prevention techniques. / Webbapplikationer utgör en viktig del av vårt dagliga liv och ger oss tillgång till betydelsefulla onlinetjänster och information. Trots deras fördelar är de också sårbara för säkerhetshot, särskilt SQL-injektionsattacker. SQL-injektion är en sårbarhet som uppstår när en angripare infogar skadliga SQL-frågor genom användarens inmatningsparametrar i en webbapplikation. Attacken medför allvarliga konsekvenser, såsom exponering av känslig information. Syftet med denna studie är att undersöka och analysera säkerheten i webbapplikationer mot SQL-injektionsattacker. Detta genomförs genom att undersöka SQL-injektionstekniker, deras påverkan på säkerhet och integritet, samt de vanligaste skyddslösningarna. Målet med studien är att öka och förbättra säkerheten hos applikationer samt skydda användarna från potentiella säkerhetsrisker. För att åstadkomma detta genomförs en kombination av litteraturstudier och praktiska undersökningar. En litteraturstudie genomförs för att identifiera SQL-injektionstekniker, säkerhetsrisker och de vanligast förekommande skyddslösningarna. Därefter utvärderas och analyseras dessa faktorer för att kunna fastställa effektiviteten hos teknikerna genom användning av penetrationstestningsverktyget SQLmap. Resultaten av studien visar att de vanligaste och mest effektiva attackteknikerna är Inband SQL-injektion och Inferential SQL-injektion. Dessa tekniker kan få allvarliga konsekvenser för användare, företag och samhället i stort, såsom åtkomst till skyddade data, manipulering av data i databasen och förlust av sekretess och dataintegritet. För att skydda mot sådana attacker är det avgörande att använda defensiva kodningsmetoder, inklusive användning av förberedda satser med parametriserade frågor och indatavalidering. Trots detta utgör manuell implementering en utmaning. En kombination av automatiserade förebyggande tekniker och bästa kod-praxis bör användas för att säkerställa en pålitlig databas som är skyddade mot SQL-injektioner, även när det gäller lagrade procedurer som är svåra att förhindra med befintliga automatiserade förebyggande tekniker.