Data-driven Cybersecurity Analytics, Forensics and Protection (DCAFP)

Project Title: Data-driven Cybersecurity Analytics, Forensics and Protection (DCAFP)

Project Period: 10/01/2019 - 09/31/2022

Amount: $3,000,000.00 (total for all three universities HU UoA and NTU)

PI: Danda B. Rawat @ Howard University

Funding Agency: US DoE's National Nuclear Security Administration (NNSA)

Project description: Our research strategy is to facilitate research collaboration among the consortium academic organizations and Argonne National Laboratory, in the following five thrust research areas:

    • Data Analytics and Machine Learning Science

    • Cyber Security Detection, Protection and Forensic Analysis

    • Wireless Cybersecurity Modeling and Analysis

    • Federated Cybersecurity Testbeds for experimentation, training and validation and

    • Cybersecurity Educational and Training Programs.

Emerging smart city systems and applications are so complex and diverse (different quality of service, delay, computing capabilities, etc.) that traditional approaches for cybersecurity, performance prediction, measurement and management are not applicable in a straightforward manner. Thus, we proposed to use federated framework for data analytics and decision making is depicted in the figure below, where individual or group of domain specific devices by forming clusters can process the data for a real-time response offload their data to the edge for near real-time processing and get the response back or use the cloud computing for off-line processing and data warehousing. We are in the process of developing a federated testbeds that can integrate many different cyber physical testbeds to study the inter-dependencies among different smart city applications (e.g. the impact of power failures on transportations, financial networks, hospitals, etc.); and to use the federated testbed for experimentation, validation and training on the cybersecurity detection and protection tools to be developed in this project.