I joined ONR in January 2010. I was given a new program in Cybersecurity, with an annual budget of ~4.5M (6.1. Basic Research Fund), and was told to go wild by my boss at the time, Dr. Wen Masters. Wild I went, with radical concepts & thinking and grew the program to ~$68M annual budget (6.1, 6.2 & 6.3 funds), including 61 & 6.2 Cybersecurity research programs, 2 MURIs, an FNC (RHIMES), an INP (TPCP), and several SBIRs & STTRs, before I left in the end of June 2019.
Part of this success was that I viewed my ONR role as a venture capital (VC) for the scientific/academic worlds, with the return of investment (ROI) based on new capabilities and breakthrough & cool scientific results and especially practical & novel technologies deploy-able into the Navy’s cyber systems, in lieu of financial gain in VC world.
To maximize scientific ROI, I took on an Anti-Popular strategy, in that I monitor but not invest on popular-research-topics-of-the-day and invest in less-explored and completely-new, unexplored topics. For this cause, I had to perform my own research & analysis on cyber related issues, and often build fundamental concepts for dealing with them. Once a problem has been identified and/or a concept has been built, I would discuss it with the academics and encourage them to pursue the research topics. Occasionally, such as the case with BFT++, I also developed a few solutions.
Anti-Popular strategy makes ONR cyber program portfolio looked significantly different than those of other research institutions within the DoD and beyond. ONR cyber & software program developed many unique (and radical) scientific results which may not have existed today without its scientific leadership.
ONR’s Cyber & Software program (CySoft) initiative timelines:
- 2010 ~ 2011
- ONR’s Autonomic Computing
- Crypto Factory, automating synthesis (mining) for Crypto Primitives
- Orthogonal Malware Detection
- 2011 ~ 2012
- Cybermoat
- Improving Software Robustness and Efficiency
- Single Core Equivalent (SCE) – FAA & EASA blessed
- Compiler and operating systems for Heterogeneous Multi-Processing (Popcorn Linux)
- Late Stage Software Complexity Reduction (debloating & delayering, precursor to FY18 INP TPCP)
- 2012 ~ 2013
- Learn 2 Reason — hybrid of statistical learning and logical reasoning, synergistic w/ Autonomic Computing (Neuro-Symbolic is very much related to Learn2Reason)
- Game of Games (MURI ’16, influenced research community, on deficiency of Game Theory for adversarial cyber interaction)
- 2013 ~ 2014
- Noise Factory (MURI ‘18) synergistic w/ Cybermoat
- Real-Time Tactical Cloud Infrastructure (RT-Xen & ANCHORS)
- 2014 ~ 2015
- BFT++, cyber-attack resilient CPS (a CySoft’s Invention, precursor FY17 FNC RHIMES)
- 2015 ~ 2016 (preoccupied w/ RHIMES & TPCP)
- 2016 ~ 2017 (preoccupied w/ RHIMES & TPCP)
- Live wire THz impedance probing for non-destructive microelectronic trojan circuit detection
- Bottom Up Formal Methods (within TPCP)
- 2017 ~ 2018 (preoccupied w/ RHIMES & TPCP)
- Expanding Late Stage Software Complexity Reduction into Protocol Subsetting & Dialecting (within TPCP)
- Legends:
- ONR influenced research topic
- ONR invented research area & Concept
- ONR invented concept & solutions/methods
Publicly released ONR February 2018 Cyber Discovery & Innovation (D&I) review slides (some slides removed) provides illustration of our work at ONR. → slide k-Cysoft-DnI-review-20180209-e