MYRIOPOS Project

The MYRIOPOS project ensures operational security by introducing and proposing new methods targeting both at the hardware and the software levels with focus on early detection and prevention of intrusions and illegal cyber content.

The mitigation of potential intrusions and cyber risks includes the analysis and indexing of information collected over networking infrastructures, dark web sites and marketplaces, as well as fast offloading of data upon detection.

At the level of early detection of illegal cyber content and illegal cyber-activities over the dark web, it includes the distributed data collection (i.e., data mining and web crawling, distributed data centers, etc.) over geographically distributed assets (i.e., network devices, dark web marketplaces and forums) and the analysis of such data with advanced deep learning algorithms.

The generated reports, alerts, figures and controls guarantee the assurance of operational security and the prompt enforcement of security policies.​​

MYRIOPOS Vision

The MYRIOPOS project differentiates from the state-of-the-art as it introduces innovative services, algorithms and solutions for:

  • scalable large-scale distributed data management, analysis and visualization,
  • efficient and real-time network and data traffic analysis through deep learning applicable to the early detection of anomalies, behavioural and pattern analysis of social networking media, the dark and the world wide web,
  • the application of innovative mitigation methods for fast networking and data offloading according to critical events finding, and
  • advanced mining and web intelligence leveraging artificial intelligence and big data to provide real-time detection, analytics and insights to both specialized security scientists, infrastructure owners and users.

MYRIOPOS Project members & Consortium

MYRIOPOS Objectives

  • Objective 1

    To develop a new infrastructure for scalable network security services to securely configure, deploy, and operate distributed detection and protection services across heterogeneous collections of hardware and software.

  • Objective 2

    To develop innovative real-time anomaly detection methods that extend the current state-of-the-art, web intelligence and semantic knowledge extraction by leveraging parallelized deep learning techniques and efficient networking mechanisms that offer distributed processing, fast offloading of "suspicious information" by exploiting programmable data paths.

  • Objective 3

    To develop a real-time risk assessment engine that facilitates timely threat handling, combined with a collection of security, resilience and operational protection policies that are enforced to mitigate the identified or reported risks by taking into account the system / assets, environmental and temporal characteristics.

  • Objective 4

    To demonstrate the developed technologies of the integrated MYRIOPOS platform in a Smart Building Infrastructure use case.

MYRIOPOS Pilot

The MYRIOPOS integrated platform is validated and demonstrated in a Smart Building use case which includes a diverse infrastructure with edge and cloud services, organised in a cloud cluster by means of a service graph. Different attack scenarios affected by the heterogeneity and complexity of this system of systems use case showcase that the prompt detection, traffic classification and prediction, findings over the dark web about massive calls for DDoS attacks, etc. and the enforcement of security policies are capable of mitigating and eliminating the cyber risk.

MYRIOPOS Outcomes

Deliverables

Publications

  • Anastasiou, T., Karagiorgou, S., Petrou, P., Papamartzivanos, D., Giannetsos, T., Tsirigotaki, G., & Keizer, J. (2022). Towards Robustifying Image Classifiers against the Perils of Adversarial Attacks on Artificial Intelligence Systems. Sensors, 22(18), 6905.

MYRIOPOS Open-Source

Datasets & Code

MYRIOPOS Contact us

    MYRIOPOS News

    Copyright ©2023 - MYRIOPOS Project