Autonomous Cyber Technology
This project investigates and applies robust cyber security techniques and applications to advanced communications and networking devices, software, algorithms and protocols utilized within wireless tactical networks to protect against nation state level Cyber and Electromagnetic Activities (CEMA) and maintain Warfighter confidence in network information, resources, identities and mission partners by hardening the blue force attack surface. This project investigates RF-enabled cyber approaches to Disrupt, Deny, Degrade, Destroy and Manipulate (D4M) adversary C2ISR systems and capabilities. Furthermore, in full alignment with ARCYBER "reprogrammability" efforts, this project will also focus on establishing new methodologies for the development of EW / OCO effects that are more readily implemented, upgradable and portable across different Army platforms. Work in this project complements Program Element (PE) 0603276A (Electronic Warfare Cyber Advanced Technology) / Project A80 (Autonomous Cyber Advanced Technology). Work in this project is performed by the Army Research Laboratory (ARL) and Command, Control, Computer, Communications, Cyber, Intelligence, Surveillance, and Reconnaissance (C5ISR) Center.
FY2027 planned work Details provided under separate cover.
FY2026 to FY2027 change Funding decrease in details provided under separate cover.
FY2026 plans — current year Will design, develop and conducts experiments with network design principals that can define logical network enclaves to support the dynamic adaptations necessary to enhance security and trust while continuing to provide optimum network traffic flow and services at the tactical level; design and develop suitable predictive algorithms that autonomously identify, learn, and react to changes in network/cyber threats to ensure end-to-end network communications resiliency against adversarial AI-driven cyberattacks; conduct experiments to integrate artificial intelligence/ machine learning (AI/ML) techniques with micro-segmentation solutions to enable dynamic adjustment of micro-segmentation in…
FY2027 planned work Details provided under separate cover.
FY2026 to FY2027 change Funding decrease in details provided under separate cover.
FY2026 plans — current year Will research non-traditional access and effect vectors against emerging targets of interest based on joint prioritization with Operational stakeholders; develop new EW technique methodologies that expedite the development of new capabilities and allow for greater portability across different hardware architectures; conduct research leveraging synthetic data enrichment allowing for improved counter ISR model generation and effectiveness.
FY2027 planned work The details of work for this project is being transitioned to an alternate location to accommodate higher classification requirements, ensuring compliance with security protocols and safeguarding sensitive information
FY2026 to FY2027 change The details of work for this project is being transitioned to an alternate location to accommodate higher classification requirements, ensuring compliance with security protocols and safeguarding sensitive information
FY2026 plans — current year Will design and develop adversarial resilient Artificial Intelligence/Machine Learning (AI/ML) methods for cyber defense that will be resistant to poisoning attacks; investigate nested ensemble defenses composed of gradient boosted classifiers that reduce the effects of clean label attacks where the attacker does not control the labeling process; investigate transfer of machine learning methods for network traffic detection between different machine learning environments; investigate novel AI/ML algorithms and methodologies for operations in contested and constrained environments; investigate resilience of various novel network traffic classifiers against poisoning.