Cyber Summer School

This session examines the structure of cyberspace and how it creates an imperative to act to advance national interests through exploitation of vulnerabilities. The session uses the United States as a case study of how countries have shifted to this reality and discusses the US doctrine of persistent engagement as a national cybersecurity operational approach.  

Learning outcomes

1. Learn core concepts of cybersecurity strategy and policy
   
2. Understand the unique national security challenges posed by the digital domain
3. Evaluate different forms of cyber strategic competition
4. Assess different approaches to national cyber strategies 
5. Review key assumptions supporting national cyber strategies

This session discusses cyberspace as political domain and its evolution as an arena of international competition.  It will review core challenges like the attribution problem as well as key actors and cases. In particular, this session will delve into how various geopolitical actors have adapted to the cyber domain and endeavored to uses its unique characteristics as tools for geopolitical gain.  

Learning outcomes

1. Learn core concepts of cybersecurity strategy and policy
2. Understand the unique national security challenges posed by the digital domain
3. Evaluate different forms of cyber conflict
4. Assess different approaches to national cyber strategies 
5. Review key cases studies and strategies of international conflict in cyberspace

Even the strongest, most complex security system can easily be bypassed by one often overlooked vulnerability, the human operator. In this session we will discuss the human element of cyberattacks by examining the attack vectors that can be created by human error, manipulation, or a simple lack of understanding. We will then discuss methods to mitigate these risks through implementing security mechanisms, policies, and training.

Learning outcomes

1. Identify commonly used human focused attack vectors.
2. Understand the challenges faced by non-technical computer system user
3. Examine technical security mechanisms that can help mitigate human attack vectors
4. Understand the importance of effective cybersecurity policies and enforcement
5. Assess different training methods for educating a cybermindful workforce

An economic framework to understand the costs and benefits behind blockchain cyberattacks will be introduced. In a blockchain environment, traditional security tools of government coercion through fines and imprisonment are no longer applicable. A secure blockchain environment is enforced whenever the economic costs of a cyber-attack exceed the benefits. We will explore how the consensus protocols of proof-of-work and proof-of-stake may be economically vulnerable to attack. This session will also provide a brief introduction to game theoretical tools in ransomware attacks. We will explore questions that include: whether to negotiate through a third-party in a ransomware attack, and how attackers determine the digital currency and the amount of a ransom.

Learning outcomes

1. Calculate economic costs and benefits of a blockchain attack
2. Apply game theory models to negotiations in ransomware attacks
3. Estimate the economic value that is vulnerable to attack in decentralized finance protocols

In this module, Dr. Wang will introduce the basic concepts of encryption, including symmetric key encryption and public-key encryption. Specifically, Dr. Wang will introduce the details of two common encryption algorithms that we use in practice, including Advanced Encryption Standard (AES) and RSA. In addition, Dr. Wang will introduce side-channel attacks, and utilize AES and RSA as two concrete examples to discuss how attackers can compromise secret keys of encryption algorithms by analyzing power consumption of target devices (such as microcontrollers and FPGAs). Recent research results on deep-learning-based side-channel attacks will also be briefly discussed.  

Learning outcomes

1. Master basic knowledge in encryption
2. Understand concepts of side-channel attacks
3. Learn latest research outcomes in side-channel attacks

These two sessions aim to provide the conceptual understanding of the function of blockchains, design of blockchain, how blockchain is used as an application to cryptocurrencies mining, what is the mathematical puzzle? and how the difficulty levels are setup? what are the methods of securing distributed ledgers and proof of stake vs proof. It also covers the technological and management underpinnings of Blockchain operations as distributed data structures and decision-making systems, their functionality and different architecture types. Finally, we will advocate the application of Blockchain to the Real-Life applications.

Learning outcomes

1. Understand the structure of a Blockchain and why/when it is better than a simple distributed database;
2. Design and Analyze the incentive structure in a Blockchain based system and critically assess its functions, benefits and vulnerabilities;
3. Evaluate the setting where a Blockchain based structure may be applied, its potential and its limitations;
4. Analyze to what extent smart and self-executing contracts can benefit automation, governance, transparency and the Internet of Things (IOT) and Attain awareness of the new challenges that exist in monetizing businesses around Blockchains;
5. Describe and understand the differences between the most prominent Blockchain structures and permissioned Blockchain service providers, as well as rising alliances and networks.

This module is dedicated to a case study of using Machine Learning (ML) in malware detection based on understanding the context in which various program constructs appear. The ML algorithm to be used is introduced, then we will discuss the notion of context useful for our problem. A relevant data set will be discussed. Finally, all these will be integrated into an approach for malware recognition applied to the data set considered.

Learning outcomes

1. Basic approach to supervised learning
2. A particular model: Bayesian Learning
3. An example of application in cybersecurity through a case study of malware classification using Bayesian Learning.

In this module, Dr. Wang will discuss several critical cybersecurity research problems that can be tackled by machine learning. The research problems will include network traffic analysis, malware detection, and wireless device authentication. Specifically, the speaker will explain how these problems can be formulated and tackled by machine learning, especially deep neural networks, discuss the limitations on the robustness and reliability of these AI-based solutions, and more importantly, how we can overcome these limitations based on Dr. Wang’s recent research.   

Learning outcomes

1. Understand the limitations of AI in cybersecurity 
2. Learn advanced knowledge to overcome the limitations  
3. Being able to apply machine learning to various applications in cybersecurity

This module helps the students to explore certain software engineering issues related to cybersecurity, including the tracing and compliance of security policies like HIPAA in real-world (electronic health record) software applications, and the critical analysis of security policies and their breaches. The module will also provide the students some hands-on experience of using static analysis and penetration testing tools to detect software vulnerabilities in industrial-strength codebases.

Learning outcomes

1. Understand the importance and limitation of security policies 
2. Analyze and evaluate traceability approach to regulatory compliance
3. Apply automated techniques to detect software vulnerabilities

In this module, students will learn to use the industry-standard tools for reverse engineering (i.e., Ghidra) and see the connection between it and network exploitation. As attackers become more and more adept at breaking into systems, software developers become more and more adept at keeping them out. This module will also introduce students to the techniques that application and system developers are deploying in order to prevent attacks on their systems (ASLR, NX bits, n-variant systems, control-flow integrity [CFI], etc.)

Learning outcomes

1. Use Ghidra, one of the industry-standard reverse-engineering tools, to disassemble a malicious program;
2. Define shell-code injection, return-to-libc, and (B)ROP attacks and use them to exploit a vulnerable program; 
3. Understand the value of, and be able to employ, fuzzing for finding vulnerabilities in software;
4. Recognize the utility of public-key cryptography in verifying that programs have not been altered between the time of their creation and the time of their execution;
5. Describe the techniques that application developers can use to protect their programs against common attacks; and
6. Explain the techniques that modern operating systems use to defeat common attacks against vulnerable programs.

Testing has been the standard for software and hardware validation for a very long time.  But testing has its problems: most notably it is impractical to test all execution paths and expensive failures in the field emerge, sometimes after a year of deployment, even after extensive testing.  Improved confidence in the safety and security of software, hardware, and systems can be achieved with formal methods.  In particular, "correct by construction" tools can create correct and safe code or VHDL from a trustworthy specification, SMT solvers can check the equivalence of two functions that are implemented differently and over different architectures, and they can also verify correctness of an implementation against a golden specification.  In this session the reason specifications can be trustworthy and some of the tools used to verify correctness from specifications will be demonstrated.  It should be noted that Amazon uses such tools to verify correctness of functions used by its AWS.

Learning outcomes

1. Write a specification in some existing systems in some specification language such as Cryptol
2. Design axioms for sound and complete proof systems
3. Write function and safety properties of hardware or software in some specification language
4. Prove the above properties in existing systems using the Software Analysis Workbench
5. Use a Satisfiability/Satisfiability Modulo Theories, and Interactive Theorem Prover solver - that is, set up a logical expression to solve a given problem
6. Understand the benefits and mechanics of Theorem provers and proof systems
7. Understand soundness and completeness and why these are important for proving correctness

Taking advantage of redundancies in images using traditional image processing methods and/or integrating image representations extracted from pre-trained deep neural networks, an image can be encoded into a cover image or a video clip. In this module, we will cover the basics of image hiding, related evaluation measures, primary challenges, and recent advancements (State-of-the art DNN models) reviewing a case study.

Learning outcomes

1. Students will learn simple image processing concepts
2. Students will understand the difference between Shallow features and deep neural network representations
3. Students will learn basics of image encoding and watermarking for privacy/security improvement 
4. Students will learn how to hide an image in a cover image taking advantage of different type of visual features

Trust in integrated circuit (IC) functionality has become a significant concern due to the emergence of highly distributed, multi-institutional process of IC design and development.  In particular, the use of “third-party” foundries for IC manufacturing, where fab lines are often owned by potentially untrusted entities, necessitates the analysis of possible attack methods and the development of threat models and defense methodologies.  

‘Trust’ in this context encompasses at the three mutually related concerns: (1) Trust in the third-party services such as IP providers and foundries (that what they do and provide is functionally as trustable as what equivalent in-house services would provide). (2) Trust that no one can take unfair advantage of the IC design. (3) Trust that the fabricated IC functions and performs exactly as expected – nothing more and nothing less.  Trust is closely related to, but not the same as, correctness, safety and security. 

Research at the University of Cincinnati (UC) has been focused in various topics central to the development of trustworthy ICs. This talk presents current and emerging research directions in split manufacturing, logic encryption/obfuscation, design camouflaging, Trojan detection, reverse engineering, run-time monitoring and side-channel attacks.  

Learning outcomes

1. Obtain basic knowledge about the vulnerabilities in the integrated circuit design and manufacturing process.   
2. Understand various attacks exploiting those vulnerabilities.
3. Appreciate emerging solutions to mitigate those attacks.