Decentralized learning allows data owners to collaboratively train machine learning models without relying on a central server, making it attractive for privacy sensitive and distributed environments. However, despite keeping data on-premises, model updates exchanged between peers can still leak private information about the underlying dataset. In particular, Membership Inference Attacks (MIAs) allow an adversary to determine whether a specific data sample was used in the training process, posing a significant privacy risk. Differential privacy (DP) is a common defense against this leakage, which injects carefully calibrated noise into model updates, but this inevitably hurts utility. Recent studies have shown that chunking, where model updates are split into chunks and only a subset of chunks is shared to neighboring nodes, can also mitigate leakage. Existing approaches include topology-aware chunking, where the number of chunks for a specific node is dependent on the communication topology, and topology-independent fixed-𝐾 chunking, where a fixed number of chunks 𝐾 is used for all nodes. However, it remains unclear how the underlying topology influences the effectiveness of such defenses.

This thesis investigates whether topology-aware chunking can improve the privacy-utility tradeoff compared to topology-independent chunking strategies. We study decentralized image classification on CIFAR-100 across several communication topologies, including ring, star, grid, fully connected, 𝑑-regular, and Erdős-Rényi graphs. Privacy leakage is measured through the accuracy of the MIA (Area Under the Curve), while utility is measured by global test accuracy. The results show that the effectiveness of topology-aware chunking is strongly influenced by the underlying
communication graph. Without defenses, MIA AUC remains high across all graph families (around 0.97-0.99). Topology-aware chunking reduces leakage significantly in dense graphs, for example, lowering AUC to 0.61 in the fully connected graph, but introduces uneven protection for sparse or heterogeneous topologies, where low-degree nodes remain vulnerable.

Compared to topology-aware chunking, topology-independent fixed-𝐾 chunking proves to be a stronger and more uniform graph-independent baseline. It often achieves equal or better privacy-utility tradeoffs, especially in utility-focused settings. To address the key limitation of topology-aware chunking, we propose ChunkDP, a defense that combines topology-aware chunking with degree-scaled DP noise. ChunkDP improves over DP-only by recovering a portion of the lost accuracy while keeping leakage close to random guessing performance (AUC 0.53). We show that ChunkDP can outperform fixed-𝐾 chunking in balanced privacy-utility settings.

Overall, the results show that topology-awareness alone does not guarantee a better privacy-utility tradeoff. Its effectiveness depends on graph density, node degree, and the desired privacy-utility balance. Fixed-𝐾 remains a robust defense, while the topology-aware ChunkDP can be useful in balanced privacy-utility scenarios. ... Read more

This work addresses the challenge of performing expressive, multi-dimensional range queries directly over encrypted data while balancing query efficiency against privacy leakage. Existing searchable encryption and encrypted multi-map (EMM) schemes either reveal access or volume patterns or incur substantial overhead by fully hiding all leakage (e.g. via ORAM) making them unpractical. Building on the static, multi-dimensional EMM framework of Falzon et al., we introduce a family of five EMM variants that provide tunable leakage profiles, spanning from full access- and volume-pattern exposure to near-complete concealment through adaptive padding and dummy-access techniques. The empirical evaluation of our
five EMM variants reveals a clear, quantifiable spectrum of privacy-performance trade-offs. On large-range workloads, the access-hiding schemes offer the best overall balance, with measured average latency slopes of ≈ 0.012 ms/label. For workloads dominated by small result sets, a volume hiding scheme excels, achieving an even lower slope of 0.0032 ms/label by tuning its padding to realistic occupancy bounds. In contrast, fully padded schemes like incur substantially higher overheads, up to two orders of magnitude greater, making them suitable only when maximal leakage resilience is required. These results allow cloud providers with quantitative guidance to deploy encrypted range search that meets both privacy requirements and performance expectations in real-world, multi-attribute database services. ... Read more

Federated Space Public Key Infrastructure (PKI) can offer a scalable foundation for secure and interoperable communications in collaborative space missions. Yet, its deployment faces challenges stemming from resource-constrained assets, architectural complexity, and the transition to post-quantum (PQ) cryptography. Current CCSDS space guidelines rely on the Internet X.509 profile, whose extensive feature set—if left unrestricted—can increase implementation complexity, certificate size (especially under PQ algorithms), and the risk of interoperability issues. In parallel, the IETF C509 Certificates draft emerges as a streamlined subset of X.509 with a compact encoding specifically tailored for constrained environments. This paper provides an empirical comparison between X.509 and C509 to inform space mission designers about the associated advantages and costs of each, specifically when PQ cryptography is incorporated into space PKIs. To help pave the way for interoperability in federated space missions, a minimal certificate profile for space PKI is proposed.

In addition, the work introduces the first open-source native C509 toolkit that supports PQ algorithms and evaluates open-source and proprietary certificate parsers. While the IETF C509 draft proposal reports a size reduction of over 50%, our evaluation confirms approximately 40% savings for classical certificates generated according to our proposed minimal certificate profile. For PQ certificates, the savings plateau at around 200 bytes, rendering the size gains negligible. However, revocation lists consistently achieve a 60% reduction for 30,000 entries, independent of the cryptographic scheme (PQ or traditional). To quantify and compare the software implementation complexity of X.509 and C509, we conduct software complexity analysis using well-established heuristic metrics (e.g., cyclomatic complexity, Halstead metrics, logical lines of code). The findings further highlight the relative simplicity of the C509 parser implementation in software. Defining a standardised certificate profile for federated space would advance interoperability; however, adopting C509 requires carefully balancing modest PQ size savings against software simplification and the uncertainties associated with a draft standard. ... Read more

With the ever-growing cloud capacity in our society, being able to harness that immense computational power while keeping our data private has become a strong point of interest. This can be achieved with Structured Encryption, which allows a data structure to be encrypted and stored remotely, but still queryable by the client owning
the encryption key. This report is a literature review of the field of Structured Encryption (StE). We analyze the state-of-the-art technologies and their characteristics on the following aspects: security, efficiency, functionality and usability, and discuss their capabilities and limitations. We then compare StE schemes with other promising technologies in the area of computation on encrypted data: Fully Homomorphic Encryption, Oblivious RAM, Secure Multiparty Computation and Trusted Execution Environments. ... Read more

Secure multi-party computation (MPC) allows parties to compute on their secret inputs, without revealing them to each other. As an area of theoretical interest, many MPC protocol have been developed in the last four decades. They each present different characteristics and are classified under distinct categories depending on their generality, security assumptions, and functionality. More recently, MPC has also become an area of practical interest due to optimizations in performance of the protocols. In this paper, we compare MPC protocols and other techniques for computing on encrypted data, considering how their properties affect security, efficiency, usability, and functionality. We show that there is a trade-off between security and efficiency when different adversarial models are used, as well as a trade-off between efficiency and flexibility in specialized protocols. ... Read more

In the digital economy, individual data sovereignty is a requirement for sustainable and secure development. With an increase in the number of outsourced computations, due to the adoption of web services, artificial intelligence and research-based innovation, privacy became one of the main concerns for individuals and third parties alike. Fully Homomorphic Encryption (FHE) addresses this gap by enabling secure computations on encrypted data. This paper reviews and presents the FHE literature in a structural manner, covering the topics of mathematical constructions, security, efficiency and functionality. In doing so, it highlights the current state-of-the-art techniques and provides a systemic comparison with other privacy enhancing mechanisms. In addition, it discusses its applicability in the practical domain, such as Confidential Machine Learning, Medical Data Analysis and Recommender Systems, mentioning potential impediments and their solutions for mass adoption. ... Read more

Outsourcing data to the cloud can pose serious security threats due to an attacker observing the data access patterns, even though data is encrypted. Ideally, confidentiality should not depend on the server being a trusted party. Oblivious Random Access Machines are tackling this problem by obfuscating data access patterns and ensuring obliviousness of the server towards the data. Thus, this paper studies the evolution of Oblivious Random Access Machines and highlights the steps taken so far to discover a more practical algorithm for hiding data access patterns on an un trusted server. In addition, other approaches for privacy-preserving computation, such as Homomorphic Encryption, Structured Encryption, Multi Party Computation and Trusted Execution Environments are discussed and contrasted to ORAM to assess the costs and benefits they come with, but also the trade-offs in security and usability. ... Read more

While securing data-in-use was assured by well-known encryption algorithms, the industry shifted towards trusting hardware manufacturers in exchange for efficiency speedups through Trusted Execution Environments. However, there are many technologies to choose from, each with its own design and trade-offs. Additionally, no work was conducted to systematically compare Trusted Execution Environments side-by-side with privacy-preserving techniques. Therefore, this literature review analyzes, in the first part, Intel's SGX, Keystone, Intel's TDX, and AMD's SEV from four angles that are strongly tied to data-in-use protection (functionality, efficiency, security, and usability). We observed that even though complex and inherently suffering to hardware-related attacks, TEEs offer a great option for confidential computing. Lastly, this research compares these state-of-the-art technologies to four other privacy-preserving techniques (Fully Homomorphic Encryption, Secure Multi-Party Computation, Oblivious RAM, and Structured Encryption) by drawing common properties and displaying them in equal use cases, showing that TEEs are a great choice for many use cases, but with stronger security issues. ... Read more

The exponential growth of digital data has reshaped the global landscape of information management, posing urgent security, compliance, and sustainability challenges. Cloud computing offers scalable, ubiquitous storage but raises critical concerns about data lifecycle governance, especially the irreversible deletion of data that has outlived its purpose. Digital forgetting becomes the art of purposeful disappearance in a cloud that remembers everything. The thesis addresses this pressing question: \textbf{Can cloud-stored data ever be truly forgotten?}

To answer this, the thesis presents four interrelated contributions
to reinforce digital data forgetting in cloud storage: advancing privacy-preserving forgetting, enabling audience-specific expiration control, supporting collaborative deletion for co-owned data, and ensuring verifiable erasure in untrusted multi-cloud environments.

To address retrospective privacy, we propose Key Decay, a cryptographic scheme where encryption keys degrade irreversibly over time, eliminating reliance on ephemeral storage and enhancing data expiration guarantees.

To support audience-specific data expiration, we propose a Disjunctive Multi-Level Forgetting Scheme that enables distinct user groups to access the same data under tailored validity periods. Smart contracts and decay sensitivity tuning enforce flexible governance across hierarchical access levels.

To manage co-owned data deletion, we introduce a Policy-Based Conjunctive Scheme that accommodates overlapping group memberships and collaborative decision-making. It applies conjunctive thresholds and verifiable key decay that comply with secure forgetting under the EU General Data Protection Regulation (GDPR) Right to Be Forgotten in real-world multi-stakeholder settings.

To ensure verifiable deletion under Byzantine infrastructure, we design a Verifiable Deletion Framework for Multi-Cloud Environments, combining Hardware Security Modules, Secure Enclaves, and dual-layer Merkle hashing to produce cryptographic proofs of deletion across providers both locally and globally.

Together, these contributions form a unified, privacy-preserving framework for managing cloud data from creation to irreversible deletion, reinforcing secure digital forgetting and regulatory compliance. ... Read more

This paper explores the scalability of Graph Neural Networks (GNNs) in the context of traffic forecasting, a critical area for improving urban mobility and reducing congestion. Despite GNNs’ demonstrated effectiveness in handling complex spatiotemporal dependencies in traffic data, scaling them to large road networks remains challenging due to increased computational requirements. This study aims to evaluate how the accuracy and computational cost of a state-of-the-art traffic forecasting GNN, the Decoupled Dynamic Spatio-Temporal Graph Neural Network, change with varying road network sizes and complexities (i.e., sensor density). Using two real-world datasets, three experiments are conducted: scaling map area, scaling graph complexity, and testing the geographic location effect. Findings show that larger graphs generally improve accuracy and GPU efficiency. Moreover, geographic location affects accuracy, whereas sensor density has minimal impact. ... Read more