Iftekhar Salam

The incorporation of federated learning (FL) and blockchain has appeared as a transformative approach for privacy-preserving, decentralised healthcare data sharing. This literature review examines recent advancements in FL-blockchain frameworks applied to healthcare, focusing on critical challenges including data privacy, security, collaboration barriers, centralisation issues and scalability concerns. The findings reveal that blockchain enhances data integrity, access control and trust, while FL enables collaborative model training without sharing raw patient data. Despite these advantages, significant challenges persist such as vulnerabilities to adversarial attacks, regulatory compliance gaps with General Data Protection Regulation (GDPR) and Health Insurance Portability and Accountability Act (HIPAA), handling heterogeneous, non-independent, and identically distributed (non-IID) medical datasets that degrade model performance. The review highlights emerging solutions including secure multi-party computation, smart contract-based aggregation, and incentive mechanisms, while emphasizing the need for future research to develop regulatory-aligned, scalable, and anomaly-resilient FL-blockchain architectures specific to healthcare. Addressing these challenges is essential to establish a trustworthy, efficient, and legally compliant AI-driven healthcare systems.

The number of smart devices connected to the Internet has been constantly increasing, and as a result, lightweight cryptography (LWC) has become more important in the past decade. The Lightweight Cryptography (LWC) Project is an initiative taken by the National Institute of Standards and Technology (NIST) to standardize such LWC algorithms. Grain128-AEAD, which was submitted to the NIST LWC project, is an encryption algorithm that provides both confidentiality and integrity assurance. Third-party security analysis of the submitted ciphers is an important aspect of the evaluation of the submission to the NIST LWC project. Although several pieces of existing research, such as the bit-flipping attack, random fault attack, and deterministic random fault attack, have examined the security of Grain128-AEAD, there is still room for improvement in the fault attack models of these studies. This work aims to fill this research gap by analyzing the security margin of Grain128-AEAD against a series of improved differential fault attacks. In this study, we developed a probabilistic random fault attack and applied it to Grain128-AEAD. As an improvement of the existing research, a probabilistic approach can be applied to a more relaxed moderate control attack model. The existing moderate control model assumes the fault to be injected within any bit of a given byte, whereas the faults in our improved approach can be injected within any bits of a two-byte/four-byte segment, thereby relaxing the fault precision. The results indicate that the improved moderate control requires 388 keystreams for the two-byte model and 279 for the four-byte model to identify the target fault locations for implementing a state recovery attack. The relaxed fault attack models presented in this work are more practical to implement; hence, the findings of this research have improved the existing studies and narrowed the current research gap on the fault attack models of Grain128-AEAD. % To maintain consistency in terminology, "Grain-128AEAD" has been revised to "Grain128-AEAD" in both the abstract and the main text. Please confirm this revision.

We developed a blockchain-based multi-keyword searchable encryption scheme for securing COVID-19 contact tracing data. In this scheme, we used AES-GCM to encrypt the contact tracing data, guaranteeing that only authorized users can perform decryption. Our scheme employed a Ciphertext-Policy Attribute-based searchable encryption to encrypt the search index, ensuring only users with appropriate attributes can perform search operations. The scheme supports dynamic updates of the search index. The blockchain-based storage with smart contract ensures immutability and non-repudiation of storage and retrieval. Overall, the evaluation of the scheme shows that it works efficiently without compromising the security goals. This is one of the first works to implement a solution for secure storage and search of contact tracing data with blockchain-based searchable encryption. Compared to the existing searchable contact tracing schemes, it provides more features and maintains efficiency even if a large search index is used.

The resource-constrained IoT devices have limited resources such as processing power, memory, and battery capacity. Therefore it is challenging to adopt traditional cryptographic algorithms on them. In order to find a solution, the National Institute of Standards and Technology (NIST) initiated the Lightweight Cryptography (LWC) competition to standardize cryptographic algorithms for resource-constrained devices. The primary aim of our work is to implement and analyse selected finalist algorithms from the NIST competition using modern cryptanalysis techniques, with a focus on statistical fault attacks. Traditional analysis methods, such as linear and differential analysis, were not prioritized as most finalist algorithms have established defences against these methods. We implemented six of the selected finalist algorithms from the competition: Ascon, Elephant, GIFT-COFB, ISAP, TinyJambu, and Xoodyak. We chose TinyJumbu for statistical fault analysis because of its attractiveness, compact block size, and provision of a more lightweight keyed permutation.

Cybersecurity has become one of the focuses of organisations. The number of cyberattacks keeps increasing as Internet usage continues to grow. As new types of cyberattacks continue to emerge, researchers focus on developing machine learning (ML)-based intrusion detection systems (IDS) to detect zero-day attacks. They usually remove some or all attack samples from the training dataset and only include them in the testing dataset when evaluating the performance. This method may detect unknown attacks; however, it does not reflect the long-term performance of the IDS as it only shows the changes in the type of attacks. In this work, we focused on evaluating the long-term performance of ML-based IDS. To achieve this goal, we proposed evaluating the ML-based IDS using a dataset created later than the training dataset. The proposed method can better assess the long-term performance as the testing dataset reflects the changes in the attack type and network infrastructure changes over time. We have implemented six of the most popular ML models, including decision tree (DT), random forest (RF), support vector machine (SVM), naïve Bayes (NB), artificial neural network (ANN), and deep neural network (DNN). These models are trained and tested with a pair of datasets with symmetrical classes. Our experiments using the CIC-IDS2017 and the CSE-CIC-IDS2018 datasets show that SVM and ANN are most resistant to overfitting. Our experiments also indicate that DT and RF suffer the most from overfitting, although they perform well on the training dataset. On the other hand, our experiments using the LUFlow dataset have shown that all models can perform well when the difference between the training and testing datasets is small.