AI & HPC Research Center

AI and Cyber Security

Area Lead: Prof. Padmalochan Bera

Due to the increasing use of high volume and heterogeneous digital data and deployment of cyber physical systems across various applications such as e-commerce, social media, mass healthcare, transportation, smart grid, utility services, and its communication demand with diverse QoS metrics, a significant transformation has happened in the network architecture and implementation paradigm in form of Software defined networks, sensor networks and cloud computing, etc. Unfortunately, there is an inherent danger in this almost inevitable transformation – the very capabilities that allow for flexible control, sharing and optimized resource utilization also provide conduits for information leakage (leading to privacy violations), inadvertent misconfigurations, and deliberate attacks by both outsiders and insiders (leading to malfunctioning and endangering property and lives), and interdependencies that make the systems more vulnerable. AHRC will develop innovative systems and protocols to Investigate attack scenarios whose detection/prevention is difficult in resource constrained environments. It will focus on design and development of feasible and scalable AI systems for attacks exploring various techniques like Failure Detection and Diagnosis (FDD) of Discrete Event system (DES), Risk Analysis, Stateful filtering and Access control and Intrusion detection system (IDS). In particular, we will investigate: false alarm minimization for signature based IDS to address known attacks; robust AI guided anomaly based IDS for new attacks; and event based IDS for attacks not detected by signature or anomaly. In addition, we will work on risk analysis and game theory guided trust management scheme for detection of suspicious nodes and analysis of threat impact and generation of threat report with necessary remediation plan to counter the cyber-attacks.

Faculty members and researchers at IIT Bhubaneswar are working with DRDO, Bharat Electronics, Cisco, ORSAC, to develop AI, game theory and cryptography guided solutions for detection and prevention of security violations in cyber physical and IoT systems. Faculty members also work on verification and validation of embedded systems, load balancing and openflow network protocols and control. AHRC is working with other leading institutes to develop innovative solutions towards design and development of software and systems for CPS.