Aliyu Ahmed Abubakar, Jinshuo Liu, Ezekia Gilliard
Abstract Intrusion Detection System (IDS) is a critical cybersecurity task that involves monitoring network traffic for malicious activity and taking appropriate action to stop it. However, insufficient training data or improperly chosen thresholds often limit the accuracy of such systems, resulting in high false‐positive rates. To improve the accuracy of an IDS, blockchain technology can be used as it provides a secure, decentralized, immutable ledger that can track suspicious activity over time and also identify intrusions globally. In this paper, the authors propose a novel methodology to improve the accuracy of blockchain‐based IDS. The approach combines different intrusion detection algorithms using a blockchain‐integrated architecture. It is based on the fusion principle and weighted votes, which the authors used to determine their results. The authors tested the system on DARPA 99 and MIT‐Lincoln Labs datasets using accuracy and false‐positive rate as their two metrics. The system achieved 92.6% accuracy and 7.4% false‐positive rates, indicating that the proposed system significantly increases the accuracy while reducing the false‐positive rate, opening up new opportunities for the development of highly accurate networks.
O. (Oleksandr) Shmatko, O. (Oleksandr) Matvieiev, I. (Iryna) Mykhailenko
Smart contracts are software programmes that are maintained in a decentralised ledger and perform predefined actions based on specific transactions directed towards them. Smart contracts are often coded using the Solidity programming language, which adheres to a certain structure and grammar. The programming language was specifically designed and implemented for the Ethereum platform. Languages with a well-defined structure are susceptible to certain vulnerabilities, the exploitation of which might result in significant financial ramifications. This study employs a Deep Learning (DL) model for the purpose of identifying vulnerabilities. By employing a selected methodology and accurately defining the structure of input data, it becomes feasible to identify intricate interdependencies among programme variables that harbour vulnerabilities and problems. Through the utilisation of carefully designed experiments, this methodology was examined in order to get a deeper comprehension of the model and enhance its overall efficacy. The model that was constructed utilised the Solidity corpus of smart-contracts as input data to classify vulnerabilities at the string level. The utilisation of the deep learning model enables the detection of vulnerabilities with different levels of complexity in smart contracts.
Crypto malware has become a major threat to the security of cryptocurrency holders and exchanges. As the popularity of cryptocurrency continues to rise, so too does the number and sophistication of crypto malware attacks. This paper leverages machine learning techniques to understand the evolution, impact, and detection of cryptocurrency-related threats. We analyse the different types of crypto malware, including ransomware, crypto jacking, and supply chain attacks, and explore the use of machine learning algorithms for detecting and preventing these threats. Our research highlights the importance of using machine learning for detecting crypto malware and compares the effectiveness of traditional methods with deep learning techniques. Through this analysis, we aim to provide insights into the growing threat of crypto malware and the potential benefits of using machine learning in combating these attacks.
The recent surge in the attention garnered by blockchain technology, an immutable ledger enabling decentralized transactions, is noteworthy.However, the security of blockchain remains susceptible to various attacks, including distributed denial-of-service (DDoS) attacks, which have increasingly targeted Bitcoin services.In response, deep learning algorithms have emerged as a potent solution to complex problems within the realm of information science.This study proposes a novel approach, utilizing these algorithms within hybrid frameworks, to address intricate cybersecurity issues.The methodologies were implemented and fine-tuned within a Python environment.Initially, a technique known as data augmentation was applied to an experimental domain aimed at verifying efficiency and boosting precision in complex datasets.Data augmentation, a method of generating new data points from existing ones, artificially enhances the volume of data.A Conditional Table Generative Adversarial Network (CTGAN) approach was adopted for the creation of tabular synthetic data.The utilization of synthetic data was found to enhance the model's performance and robustness compared to the exclusive use of original data.Subsequently, a binary classification hybrid deep learning model, incorporating Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) algorithms, was proposed for the detection of DDoS attacks within cryptocurrency networks.The proposed model was then validated using actual instances of DDoS attacks within the Bitcoin service dataset.The validation process incorporated a test set comprising 20% of the augmented data.Evidently, the proposed model outperformed standard deep learning implementations, achieving an impressive accuracy of approximately 95.84%.This study, therefore, presents a promising approach to mitigating DDoS attacks within the Bitcoin ecosystem.
The Ponzi scheme, an old-fashioned fraud, is now popular on the Ethereum blockchain, causing considerable financial losses to many crypto investors. A few Ponzi detection methods have been proposed in the literature, most of which detect a Ponzi scheme based on its smart contract source code. This contract-code-based approach, while achieving very high accuracy, is not robust because a Ponzi developer can fool a detection model by obfuscating the opcode or inventing a new profit distribution logic that cannot be detected. On the contrary, a transaction-based approach could improve the robustness of detection because transactions, unlike smart contracts, are harder to be manipulated. However, the current transaction-based detection models achieve fairly low accuracy. In this paper, we aim to improve the accuracy of the transaction-based models by employing time-series features, which turn out to be crucial in capturing the life-time behaviour a Ponzi application but were completely overlooked in previous works. We propose a new set of 85 features (22 known account-based and 63 new time-series features), which allows off-the-shelf machine learning algorithms to achieve up to 30% higher F1-scores compared to existing works.
Tran Viet Khoa, Do Hai Son, Chi-Hieu Nguyen, Dinh Thai Hoang · 11 authors
With the escalating prevalence of malicious activities exploiting vulnerabilities in blockchain systems, there is an urgent requirement for robust attack detection mechanisms. To address this challenge, this paper presents a novel collaborative learning framework designed to detect attacks in blockchain transactions and smart contracts by analyzing transaction features. Our framework exhibits the capability to classify various types of blockchain attacks, including intricate attacks at the machine code level (e.g., injecting malicious codes to withdraw coins from users unlawfully), which typically necessitate significant time and security expertise to detect. To achieve that, the proposed framework incorporates a unique tool that transforms transaction features into visual representations, facilitating efficient analysis and classification of low-level machine codes. Furthermore, we propose an advanced collaborative learning model to enable real-time detection of diverse attack types at distributed mining nodes. Our model can efficiently detect attacks in smart contracts and transactions for blockchain systems without the need to gather all data from mining nodes into a centralized server. In order to evaluate the performance of our proposed framework, we deploy a pilot system based on a private Ethereum network and conduct multiple attack scenarios to generate a novel dataset. To the best of our knowledge, our dataset is the most comprehensive and diverse collection of transactions and smart contracts synthesized in a laboratory for cyberattack detection in blockchain systems. Our framework achieves a detection accuracy of approximately 94% through extensive simulations and 91% in real-time experiments with a throughput of over 2,150 transactions per second.
Communication using Emails, SMS, and various social media platforms have become an important part of life when dealing professionally and socially. Due to their large uses, these platforms also have gained the attention of hackers to carry out sophisticated attacks by sending out messages that contains Malicious content (Generally URL). Huge numbers of Phishing emails are sent daily. Over the past few years, various Machine Learning based techniques were introduced to identify Phishing Attacks still, the count and loss due to Phishing is increasing daily. Blockchain technology is gaining a lot of attention in Security Domain over the past few years, in this state of art authors have proposed a blockchain-based system integrated with a Machine Learning model to identify and prevent Phishing content that is sent out on various messaging platforms. For implementation purposes Ethereum Based platform and from a Machine Learning perspective Classification based Gradient Boost Algorithm and Support Vector Machine Algorithm was used. The proposed systems talk about, sender while sending out a message to a recipient should pay some ETH (cryptocurrency) and if after content validation through ML models the message is found to be legitimate, the ETH value that the sender had paid will be refunded back, else if ML models identify the content as Phishing, then there will be no refund. Even though after validation through the ML model if the recipient marks the email as SPAM or reports the same as Suspicious then also there will be no refund. This eventually will reduce the traffic and will make hackers rethink when sending Phishing Messages. Authors were able to Create a PoC illustrating the same. API endpoints that were created in this state-of-art executed within 1000 ms & the accuracy obtained from the Machine Learning model was also high.
Smart contracts, programs running on a blockchain, play a crucial role in driving Web 3.0 across a variety of domains, such as digital finance and future networks. However, they currently face significant security vulnerabilities that could result in potential risks and losses. This paper outlines the inherent vulnerabilities of smart contracts, both those typical of their applications and those unique to Web 3.0 applications. We then systematically classify the techniques based on their core approach to detecting vulnerabilities in smart contracts. Using these approaches, we conduct a comparative analysis of existing tools in terms of their vulnerability coverage, detection effectiveness, open-source availability, and integration capabilities. Finally, we present the Co-Governed Sovereignty Multi-Identifier Network (CoG-MIN) as a case study to demonstrate the significance of smart contract application security in establishing a community with a shared future in cyberspace during the Web 3.0 era and anticipate future research directions with challenges. To conclude, this study addresses the gap in integrating existing smart contract security research with the advancement of Web 3.0 development, while also providing recommendations for future research directions.
As our reliance on digital infrastructure and the transmission of sensitive information grows, the importance of effective cybersecurity measures becomes increasingly urgent. This study explores the efficacy of honeypots and the MITRE ATT& CK framework in detecting adversary behaviors in ethereum, smtp, ftp, and ldap attacks. By deploying honeypots, we gathered a diverse range of attack data, revealing prevalent patterns like phishing, scamming, account hijacking in ethereum, bruteforce in ftp, and port scanning in ldap. Mapping this data to the MITRE ATT& CK framework enabled the identification of adversary TTPs (Tactics, Techniques, and Procedures), informing security strategies and mitigating future attacks. Our findings highlight the significance of employing honeypots to identify and analyze adversary tactics, techniques, and procedures (TTPs). This underscores the importance of continual research to further improve our comprehension of evolving cyber threats. –
Digital forensics is crucial in investigating cybercrimes, focusing on collecting, analyzing, and preserving digital evidence. With technology deeply integrated into our lives, digital forensics plays a vital role in uncovering evidence related to criminal activities, cybersecurity incidents, and fraud. However, mobile devices pose unique challenges due to their diverse nature and complex data storage structures. In this context, blockchain technology emerges as a promising solution. Blockchain, a decentralized and immutable digital ledger, provides a secure framework for storing and verifying digital transactions. Its features, such as tamper resistance and cryptographic security, align well with the needs of mobile device forensics. By leveraging Blockchain, digital evidence's integrity and authenticity can be enhanced by securely recording the chain of custody. Additionally, Blockchain addresses data privacy and security concerns by offering secure data storage and access controls. Furthermore, Blockchain enables secure collaboration among investigators through its distributed nature, allowing for shared and auditable repositories of forensic data. This paper explores the intersection of mobile device forensics and blockchain technology, highlighting the challenges faced and the potential benefits of utilizing Blockchain. It proposes a blockchain-based mobile cloud forensic framework to reliably record forensic findings and resolve issues in mobile cloud forensics. Integrating blockchain technology in mobile device forensics offers improved trustworthiness and reliability of digital evidence, overcomes privacy and security concerns, and enables efficient collaboration among forensic teams. This research aims to inspire further advancements in mobile device forensics, enhancing the capabilities of digital investigators and contributing to more effective cybercrime investigations. Finally, we proposed a blockchain-based mobile cloud forensic framework to reliably record the forensic findings and resolve the issues with mobile cloud forensics.
Amit Mittal, Gunawan Widjaja, Renzon Daniel Cosme Pecho, R. Kiruba · 6 authors
As blockchain technology and the Internet of Things (IoT) have advanced, smart contracts have found new uses in many industries. Ethereum has become the go-to system for smart contract development. However, major financial losses have come from the ubiquity of smart contract vulnerabilities. Existing techniques have struggled to make efficient use of the syntax information in contract source code, making the detection of these vulnerabilities a key field of research. The reentrancy flaw in smart contracts is addressed by proposing a new vulnerability detection tool called SC-Defender . The abstract syntax tree representation of the source code for a smart contract written in the Solidity language is the core focus of SCDefender 's deep learning technique. Its efficient vulnerability identification is the result of its use of a tree-based convolutional neural network. In order to remove superfluous branches while keeping the essential ones, we develop a pruning technique for abstract syntax trees. The vulnerability detection findings achieved by SC-Defender are encouraging, with a precision of 81.43 percent, a recall of 92.12 percent, and an F1-score of 86.4 five percent. The results of the abolition experiments show that the abstract syntax tree pruning technique is a crucial part of SC-Defender 's ability to find vulnerabilities. These improvements in smart contract vulnerability detection, especially in the context of IoT-enabled smart contracts, have a lot of potential to make smart contracts safer and reduce the economic risks that come with their broad use.
Aiming to safeguard a decentralized setup such as smart cities, collaborative intrusion detection system (CIDS) has become a mainstream security mechanism to protect different types of computer networks, especially decentralized computing platforms such as Internet of Things (IoT). The main benefit of CIDS relies on the information sharing process among devices, nodes, software and hardware entities. However, traditional CIDS often requires a trusted third partner, e.g., a centralized computing server, to help build up a trusted communication channel among various entities. Such requirement is not practical in real-world implementation, making the integrity of shared information compromised easily. With the wide adoption, blockchain technology has given a solution to protect the distributed/collaborative detection system. In the current market, blockchain technology has been extensively researched across many detection scenarios, but there is a need to explore how such technology can overall contribute to CIDS and a general distributed detection system. In this work, we introduce a blockchain-assisted security management framework for CIDS, which summarizes and provides an integrated protection given by blockchain. In the case study , we evaluate our proposed framework in both a simulated and a real CIDS setup with challenge-based mechanism. The results demonstrate the promising benefits provided by blockchain in CIDS.
Smart contracts, which provide user-defined logic to blockchains, have become increasingly popular in recent years due to their decentralized system architecture. They are executable programs that can automate transactions on the ethereum blockchain. However, security concerns with certain aspects of smart contracts can be challenging to address. Therefore, this paper proposed a multi-objective approach using a neural network makes it a more scalable and effective tool for detecting smart contract vulnerabilities than traditional approaches. With the increasing complexity of smart contracts and the growing importance of security in the blockchain space, a technique like multi-objective is becoming increasingly necessary to ensure the safety and reliability of decentralized applications. The proposed approach evaluated over 11000 real word ethereum smart contracts and detected two vulnerabilities without expert knowledge. The results showed that it achieved an average F1-score of 86.7 and 84.4 percent for reentrancy and timestamp vulnerability, respectively, indicating that the proposed approach has an impressive level of accuracy in classifying complex smart contracts, which has significant implications in the blockchain security field.
Dongcheng Li, W. Eric Wong, Sean Pan, Liang-Seng Koh · 5 authors
Blockchain and smart contract technologies have emerged as innovative approaches to trustworthy and reliable service computing. With the growing popularity of decentralized applications, the quantity of smart contracts has seen an exponential increase. Research on testing smart contracts has primarily focused on identifying specific vulnerabilities in smart contracts and blockchains. However, generating a robust test suite for smart contracts remains a daunting challenge. The state-of-the-art DynaMOSA algorithm uses many-objective optimization for test case generation, introducing preference sorting and dynamic target selection strategies. Yet, under hyper many-objective conditions, the algorithm faces difficulties including insufficient selection pressure and reduced efficiency. This paper proposes a solution to these issues by introducing a many-objective optimization algorithm with a dimensionality reduction strategy. The goal is to condense an extensive dataset into a smaller, more manageable and effective dataset using specific reduction criteria. Additionally, this paper utilizes 28 open-source Solidity projects from GitHub for testing. The experimental results show that compared to DynaMOSA, the proposed approach achieves higher testing coverage on most projects under test and shows a noticeable improvement in algorithm efficiency for the automatic generation of Solidity Tests.
The Move smart contract (MSC) is designed to enhance the type security of digital assets by utilizing a resource-based structure. However, vulnerabilities may be introduced during the development process. In response to the security threats faced by digital assets in the MSC, we have identified five resource-related vulnerabilities for the first time. Additionally, we have defined two test oracles and proposed a method to detect these vulnerabilities in the MSC using resource-flow analysis. Our method begins by analyzing the resource types present in the MSC. Based on the resource operation information within the contract function, we create a resource-flow graph. Next, the resource-flows are derived from the Resource-Flow Graph using a traversal algorithm, and they are then transformed into test cases. In the final step, the generated test cases are executed, and the vulnerabilities are detected by utilizing the proposed test oracles. Through a comprehensive case study, we showcase the various resource-related vulnerabilities and assess the feasibility of our method in detecting these vulnerabilities. The results demonstrate that our proposed method is effective in identifying resource-related vulnerabilities.
Nada Abdu Alsharif, Shailendra Mishra, Mohammed Alshehri
The rise of IoT devices has brought forth an urgent need for enhanced security and privacy measures, as IoT devices are vulnerable to cyber-attacks that compromise the security and privacy of users. Traditional security measures do not provide adequate protection for such devices. This study aimed to investigate the use of machine learning and blockchain to improve the security and privacy of IoT devices, creating an intrusion detection system powered by machine learning algorithms and using blockchain to encrypt interactions between IoT devices. The performance of the whole system and different machine learning algorithms was evaluated on an IoT network using simulated attack data, achieving a detection accuracy of 99.9% when using Random Forrest, demonstrating its effectiveness in detecting attacks on IoT networks. Furthermore, this study showed that blockchain technology could improve security and privacy by providing a tamper-proof decentralized communication system.
Yuqiang Sun, Daoyuan Wu, Yue Xue, Han Liu · 8 authors
Smart contracts are prone to various vulnerabilities, leading to substantial financial losses over time. Current analysis tools mainly target vulnerabilities with fixed control or data-flow patterns, such as re-entrancy and integer overflow. However, a recent study on Web3 security bugs revealed that about 80% of these bugs cannot be audited by existing tools due to the lack of domain-specific property description and checking. Given recent advances in Large Language Models (LLMs), it is worth exploring how Generative Pre-training Transformer (GPT) could aid in detecting logicc vulnerabilities. In this paper, we propose GPTScan, the first tool combining GPT with static analysis for smart contract logic vulnerability detection. Instead of relying solely on GPT to identify vulnerabilities, which can lead to high false positives and is limited by GPT's pre-trained knowledge, we utilize GPT as a versatile code understanding tool. By breaking down each logic vulnerability type into scenarios and properties, GPTScan matches candidate vulnerabilities with GPT. To enhance accuracy, GPTScan further instructs GPT to intelligently recognize key variables and statements, which are then validated by static confirmation. Evaluation on diverse datasets with around 400 contract projects and 3K Solidity files shows that GPTScan achieves high precision (over 90%) for token contracts and acceptable precision (57.14%) for large projects like Web3Bugs. It effectively detects ground-truth logic vulnerabilities with a recall of over 70%, including 9 new vulnerabilities missed by human auditors. GPTScan is fast and cost-effective, taking an average of 14.39 seconds and 0.01 USD to scan per thousand lines of Solidity code. Moreover, static confirmation helps GPTScan reduce two-thirds of false positives.
Hossein Pourrahmani, Adel Yavarinasab, Amir Mahdi Hosseini Monazzah, Jan Van herle
The current advances in the Internet of Things (IoT) and the solutions being offered by this technology have accounted IoT among the top ten technologies that will transform the global economy by 2030. IoT is a state-of-the-art paradigm that has developed traditional living into a high-tech lifestyle. The current study aims to provide a comprehensive review and analysis of the existing cybersecurity attacks and vulnerabilities in IoT, offering suitable countermeasures with a focus on describing the impact of emerging technologies on IoT devices and protocol layers. The main vulnerabilities across different layers of the IoT reference model are discussed and categorized, and suitable countermeasures (such as separating IT and IoT network traffic, enhancing physical security, implementing encryption and secure messaging protocols, etc.) are suggested. In addition, the hardware, communication, application, web, and cloud vulnerabilities are introduced, then the corresponding safeguards and protections are presented. Furthermore, ia! (ia!) has been deliberately defined and the adoption of the NIST framework and IA model is recommended as a metric to ensure security for IoT solutions considering the five pillars of availability, integrity, authentication, confidentiality, and non-repudiation. Finally, Blockchain technology, known for its use in securing cryptocurrencies, is suggested to facilitate secure data exchange, identification, authentication, and communication for IoT devices by various avenues including ensuring the integrity of sensor data, eliminating the need for intermediaries, reducing costs, and enabling direct addressability of IoT devices.
Open access
Physical Unclonable Functions (PUFs) and Hardware Security
Compiling smart contracts into WebAssembly (WASM) bytecode format for better execution performance is becoming a common choice of many blockchain smart contract platforms, including EOSIO smart contract platform, which has been using WASM as its smart contract compilation target since its release. Similar to Ethereum smart contracts, EOSIO smart contracts have suffered from various vulnerability attacks, causing serious economic losses. Although some security analysis tools for EOSIO smart contracts have been proposed in the academic community, they still have some limitations, such as inadequate coverage of vulnerabilities and a lack of information about the triggered conditions and call stack when a vulnerability is identified. This makes it difficult for developers to utilize these tools. To address these problems, this paper proposes EOSVERIF, an automatic fine-grained analyzer of EOSIO smart contracts. It supports the concatenation of the set of vulnerabilities supported by all previous EOSIO security analysis tools and can provide the triggering conditions and execution paths of detected vulnerabilities. Our experimental results show that EOSVERIF can effectively detect vulnerabilities in EOSIO smart contracts.
Smart contracts are one of the most successful applications of blockchain technology. In order to guarantee the security of smart contracts, researchers have successively introduced various testing methodologies, including static analysis, symbolic execution, and fuzzing, which contribute to a more rigorous and precise evaluation of smart contract vulnerabilities. Deep learning techniques have been widely applied in traditional software vulnerability detection, while the opposite is true in the field of smart contract testing. Consequently, we anticipate that deep learning can be similarly applied to enhance traditional smart contract vulnerability detection tools. However, there is a lack of empirical study on the performance of deep learning applied to smart contract testing. In order to explore how deep neural networks can help with testing tools on smart contracts, we construct a test framework based on SMARTEST. We manage to train deep learning language models using various neural networks including Transformer, GRU, RNN and test the symbolic execution tool SMARTEST framework with the application of these models on the CVE dataset. Upon analyzing the experimental results, we find that deep neural networks did not surpass traditional language models in enhancing smart contract testing. In terms of accuracy, the SMARTEST tool, which utilizes a statistical 3-gram language model, succeeded in detecting the greatest number of vulnerabilities. Specifically, the 3-gram model was able to identify 69.8% of vulnerabilities in the benchmark set within the first 5 seconds. Based on our experimental findings and thorough analysis, we outline the challenges faced in DNN-assisted smart contract testing and suggest potential directions for improvement.
Smart contracts are simply computer programs. These programs are deployed on distributed nodes over the blockchain network. These are executed without the need for third-party authentication. Usually, smart contracts are used for transferring assets so it requires the error-free execution of smart contract code. But, due to computer code pitfalls, it may be the possibility of errors or exceptions that may vulnerable to the security of smart contracts. Thus, this paper surveys the smart contract security issues and smart contract code vulnerabilities that have been investigated and security analysis tools are presented. A series of vulnerable codes is presented that may have the risk of stealing assets and information. The solution to these vulnerabilities has also been discussed. A comparison with existing work has also been presented.
Aydin Abadi, Dan Ristea, Artem Grigor, Steven J. Murdoch
Time-Lock Puzzles (TLPs) enable a client to lock a message such that a server can unlock it only after a specified time. They have diverse applications, such as scheduled payments, secret sharing, and zero-knowledge proofs. In this work, we present a scalable TLP designed for real-world scenarios involving a large number of puzzles, where clients or servers may lack the computational resources to handle high workloads. Our contributions are both theoretical and practical. From a theoretical standpoint, we formally define the concept of a “Delegated Time-Lock Puzzle (D-TLP)”, establish its fundamental properties, and introduce an upper bound for TLPs, addressing a previously overlooked aspect. From a practical standpoint, we introduce the “Efficient Delegated Time-Lock Puzzle” (ED-TLP) protocol, which implements the D-TLP concept. This protocol enables both the client and server to securely outsource their resource-intensive tasks to third-party helpers. It enables realtime verification of solutions and guarantees their delivery within predefined time limits by integrating an upper bound and a fair payment algorithm. ED-TLP allows combining puzzles from different clients, enabling a solver to process them sequentially, significantly reducing computational resources, especially for a large number of puzzles or clients. ED-TLP is the first protocol of its kind. We have implemented ED-TLP and conducted a comprehensive analysis of its performance for up to 10,000 puzzles. The results highlight its significant efficiency in TLP applications, demonstrating that EDTLP securely delegates 99% of the client’s workload and 100% of the server’s workload with minimal overhead.