arXiv subjects

Vasileios Kouliaridis

Publications and source records attributed to Vasileios Kouliaridis.

6 recordsLinked to original sources

arXiv2026

Protecting On-Device AI Inference: A Systematic Review of Attacks and Defence Mechanisms

The need for secure and private Artificial Intelligence (AI) and Machine Learning (ML) on edge and mobile devices has increased the necessity of protecting the architecture of these systems from threats to both security and privacy. With an ever-increasing number of pre-trained AI models being used on mobile platforms for client-side inference, there are rising concerns about the risks associated with the theft/extraction of AI models, adversarial attacks on AI models, and data breaches. As a result of this trend, a variety of defence mechanisms have been proposed to protect against these threats. These include Trusted Execution Environments (TEEs), homomorphic encryption, obfuscation, and differential privacy, among others. However, current surveys largely focus on edge intelligence, which includes distributed training, and thus overlook security and privacy issues that are specific to on-device AI inference. To the best of our knowledge, this paper presents the first comprehensive review of threats and corresponding defence mechanisms targeting on-device inference. Our results show that the attack and defence literature are unbalanced: approximately one quarter of the surveyed attack papers focus on Intellectual Property (IP) attacks, whereas half of the defence solutions tackle the same issue. More importantly, some attack categories have no defence paper associated to them, such as adversarial attacks that account for roughly one third of the attack literature. This asymmetry between known attacks and available mitigations highlights clear opportunities for future research on securing on-device AI inference.

cs.CR↗

arXiv2024

Assessing the Effectiveness of LLMs in Android Application Vulnerability Analysis

The increasing frequency of attacks on Android applications coupled with the recent popularity of large language models (LLMs) necessitates a comprehensive understanding of the capabilities of the latter in identifying potential vulnerabilities, which is key to mitigate the overall risk. To this end, the work at hand compares the ability of nine state-of-the-art LLMs to detect Android code vulnerabilities listed in the latest Open Worldwide Application Security Project (OWASP) Mobile Top 10. Each LLM was evaluated against an open dataset of over 100 vulnerable code samples, including obfuscated ones, assessing each model's ability to identify key vulnerabilities. Our analysis reveals the strengths and weaknesses of each LLM, identifying important factors that contribute to their performance. Additionally, we offer insights into context augmentation with retrieval-augmented generation (RAG) for detecting Android code vulnerabilities, which in turn may propel secure application development. Finally, while the reported findings regarding code vulnerability analysis show promise, they also reveal significant discrepancies among the different LLMs.

cs.CR↗

arXiv2023

Intrusion Detection based on Federated Learning: a systematic review

The evolution of cybersecurity is undoubtedly associated and intertwined with the development and improvement of artificial intelligence (AI). As a key tool for realizing more cybersecure ecosystems, Intrusion Detection Systems (IDSs) have evolved tremendously in recent years by integrating machine learning (ML) techniques for the detection of increasingly sophisticated cybersecurity attacks hidden in big data. However, these approaches have traditionally been based on centralized learning architectures, in which data from end nodes are shared with data centers for analysis. Recently, the application of federated learning (FL) in this context has attracted great interest to come up with collaborative intrusion detection approaches where data does not need to be shared. Due to the recent rise of this field, this work presents a complete, contemporary taxonomy for FL-enabled IDS approaches that stems from a comprehensive survey of the literature in the time span from 2018 to 2022. Precisely, our discussion includes an analysis of the main ML models, datasets, aggregation functions, as well as implementation libraries, which are employed by the proposed FL-enabled IDS approaches. On top of everything else, we provide a critical view of the current state of the research around this topic, and describe the main challenges and future directions based on the analysis of the literature and our own experience in this area.

cs.CR↗

arXiv2022

A hands-on gaze on HTTP/3 security through the lens of HTTP/2 and a public dataset

Following QUIC protocol ratification on May 2021, the third major version of the Hypertext Transfer Protocol, namely HTTP/3, was published around one year later in RFC 9114. In light of these consequential advancements, the current work aspires to provide a full-blown coverage of the following issues, which to our knowledge have received feeble or no attention in the literature so far. First, we provide a complete review of attacks against HTTP/2, and elaborate on if and in which way they can be migrated to HTTP/3. Second, through the creation of a testbed comprising the at present six most popular HTTP/3-enabled servers, we examine the effectiveness of a quartet of attacks, either stemming directly from the HTTP/2 relevant literature or being entirely new. This scrutiny led to the assignment of at least one CVE ID with a critical base score by MITRE. No less important, by capitalizing on a realistic, abundant in devices testbed, we compiled a voluminous, labeled corpus containing traces of ten diverse attacks against HTTP and QUIC services. An initial evaluation of the dataset mainly by means of machine learning techniques is included as well. Given that the 30 GB dataset is made available in both pcap and CSV formats, forthcoming research can easily take advantage of any subset of features, contingent upon the specific network topology and configuration.

cs.CR↗

arXiv2021

Dissecting contact tracing apps in the Android platform

Contact tracing has historically been used to retard the spread of infectious diseases, but if it is exercised by hand in large-scale, it is known to be a resource-intensive and quite deficient process. Nowadays, digital contact tracing has promptly emerged as an indispensable asset in the global fight against the coronavirus pandemic. The work at hand offers a meticulous study of all the official Android contact tracing apps deployed hitherto by European countries. Each app is closely scrutinized both statically and dynamically by means of dynamic instrumentation. Depending on the level of examination, static analysis results are grouped in two axes. The first encompasses permissions, API calls, and possible connections to external URLs, while the second concentrates on potential security weaknesses and vulnerabilities, including the use of trackers, in-depth manifest analysis, shared software analysis, and taint analysis. Dynamic analysis on the other hand collects data pertaining to Java classes and network traffic. The results demonstrate that while overall these apps are well-engineered, they are not free of weaknesses, vulnerabilities, and misconfigurations that may ultimately put the user security and privacy at risk.

cs.CR↗

arXiv2020

Feature importance in mobile malware detection

The topic of mobile malware detection on the Android platform has attracted significant attention over the last several years. However, while much research has been conducted toward mobile malware detection techniques, little attention has been devoted to feature selection and feature importance. That is, which app feature matters more when it comes to machine learning classification. After succinctly surveying all major, dated from 2012 to 2020, datasets used by state-of-the-art malware detection works in the literature, we analyse a critical mass of apps from the most contemporary and prevailing datasets, namely Drebin, VirusShare, and AndroZoo. Next, we rank the importance of app classification features pertaining to permissions and intents using the Information Gain algorithm for all the three above-mentioned datasets.

cs.CR↗