Research

Dr. Yang's current research focuses on providing Predictive Cyber Situation Awareness and Anticipatory Cyber Defense through the development of ASSERT (Attack Strategy Synthesis and Ensemble pRedictions of Threats), CASCADES (Cyber Attack SCenArio & network Defense Simulator), and CAPTURE (Cyber Attack Prediction of Threats from Unconventional REsources). ASSERT recognizes multistage attack behaviors in enterprise networks and synthesizes statistical models for these behaviors based on large-scale, incomplete and noisy cyber observables. CASCADES generates cyber attack scenarios based on the interplay between adversary capability/opportunity/intent/preference and network defense configurations. CAPTURE aims at forecasting cyber attacks based on unconventional data sources, sensors, and signals. Dr. Yang and his team are developing novel machine learning, data analytics, and simulation techniques as sub-components of the above systems. The ultimate goal of these systems is to enable anticipatory, proactive cyber defense that are agile and resilient to increasingly sophisticated cyber attacks.

The above ongoing works are built upon the earlier work by Dr. Yang’s team who introduced Fuzzy-Variable Length Markov Models (F-VLMM), Virtual Terrain (VTAC), Attack Social Graphs (ASG), FuSIA, and an attack obfuscation modeling framework using statistical graph models. Students and collaborators who are interested in working with Dr. Yang may contact him via jay.yang@rit.edu.

Students, alums, and collaborators who have contributed to this set of work include:

• Current students:.

• Research Faculty: Ahmet Okutan.

• Alumni: Shao-Hsuan (Steven) Su, To Chen (Robin) Chang, Steve Moskal, Gordon Werner, Ian Perry, Christopher Sweet, Lutzu (Roy) Lu, Fuyuan Cheng, Alex Krall, David Stolze, Alexandra Harrison, Jake Saxton, Biru Cui, Terence O’Brien, Lisa Trova, Neil Wong Hon Chan, Ben Wheeler, Haitao Du, Ryan Rawlings, Steven Strapp, Venkata Mudireddy, Daniel Lu, Chris Murphy, Jordan Bean, Stephen Byers, Daniel Fava, Brian Arguaer, Gilbert Hendry, and Jared Holsopple.

• Collaborators: Michael Kuhl (ISE, RIT), Bill Stackppole (CSEC, RIT), Daryl Johnson (CSEC, RIT), Katie McConky (ISE, RIT), Aunshul Rege (Criminal Justice, Temple University), and Moises Sudit (University of Buffalo).

• Current and past funding support: NSF, IARPA, NSA, AFRL, ONR, ARL, DARPA.

High level fusion for cross-domain decision support

• Aimed at providing timely impact/threat assessment and collection requirements given diverse cross-domain data sources.

• Developed a series of Integer Programming problem for cross-domain C2ISR decision support from plausible futures.

• Visualization of impact/threat for asymmetric warfare.

• Alumni: Michael Nusinov, Khiem Tong.

• Collaborator: Moises Sudit (CUBRC) and Jared Holsopple (CUBRC).

• Funded under AWARE program (AFRL) in collaboration with CUBRC.

Information Fusion via Sensor-Cloud

• Aimed at developing components that can effectively and securely share sensor data, along with simulated and knowledge base data via cloud services

• Utilize mathematical frameworks, such as Markov Random Fields and Information Theory, to analyze the interplays between sensor data as well as between cloud services

• Application domain includes Smart Grid and Surveillance

• Alumni: Jon Szymaniak, Patrick Ciambrone, Corey Beres

Cooperative and autonomous robots (with heterogeneous sensors and wireless communication)

• Robot swarm algorithms for multi-threat containment, regional and cooperative surveillance and patrol

  • Current students: Pat LaRocque

  • Alumni: Michael Ellis, Dieter Laskowski, Daniel Stella, Jessica LaRocque, Bhushan Mehendale and Nate Ransom

§ MAHESHDAS: A simulator for autonomous robot collaborations

• Autonomous shape formation with no reference points, Mark Seidman

• Mobility models for autonomous robots, Sidd Sail

• Comprehensive sensor modeling in generic network simulators (OPNET), Niranjan Krishnamurthi

• Partially funded by the Technology Innovation Center under the Director of Central Intelligence and RIT grants

Micro-scale Haptic Feedback and Tactile Informatics

• Aims at capturing and displaying tactile information of micro-scale objects

• Model-data fusion with large uncertainty

• Require transformation from unknown micro-scale properties to human comprehensible haptic sensation

• Alumni: Daniel, Liu, Adam Weissman and Athena Frazier

• Collaborator: Dr. Yen Wen Lu and Dr. Jijie Xu

Maximum data retrieval problem (subject to finite sensor resource constraints)

• Jointly optimizes scheduling and routing decisions in the dynamic regime

• Accounts for buffer usage and energy spikes

• Leads to energy efficient data forwarding protocol

• Alumni: Cory Cress

• Partially funded by the Technology Innovation Center under the Director of Central Intelligence

• Wireless Ad Hoc Relay Protocol by limited Gossiping (i.e., probabilistic flooding): Julie Yerdon, Zeping Qiu

• Collaborative networked PTZ cameras for maximal dynamic surveillance coverage: John Ruppert

  • Leveraging computational geometry and application specific dynamic camera coverage models

• OPNET Projects: Sungho Maeung, Mark Seidman, Niranjan Krishnamurthi,

  • Modular Topology Control and Energy Model for Wireless Ad Hoc Sensor Networks

  • Mobile Wireless Sensor Formation Algorithm Development using OPNET

  • Modeling and Simulation of Gossip-Based Relayed Network for Ubiquitous and Expanded Access

• Ubiquitous roaming over secured ad hoc networks (UROSAN)

• User perceived performance oriented best effort networks