“As a small-business leader, I see daily the importance of bringing these types of game-changing capabilities to market on behalf of the National Labs. With the ever-expanding use of devices and sensors in everyday life, the threat related to the compromising of these devices become ever more daunting. This capability goes a long way in addressing the many concerns people have in placing their information sharing eggs in one basket.” stated Alan Seymour, Cleveland Electric Labs President.
The cyberattack battleground is expected to increase by an order of magnitude between now and 2025, while network attacks have increased by over 50 times between 2015 and 2020.
Unfortunately, machine learning-based classification models for attack detection are difficult to achieve with high classification speed. They remain vulnerable to zero-day attacks, while conventional rule-based classification scalable for increased accuracy. While rule-based network packet security systems have previously been the gold standard for protection, INL’s research anticipated that auxiliary machine learning-based systems will be needed to secure future systems analogous to the systems currently used for credit card fraud detection. Using Artificial Intelligence (AI) and Machine Learning (ML) this capability has not only the ability to defend in ways that have not previously been considered, but also the ability to predict potential attacks and thwart them before they can occur.
Researchers at Idaho National Lab developed a mechanism for detecting network packet anomalies indicative of a network attack using the reconstruction probability from a variational autoencoder. This semi-supervised learning approach is not built around binary classification but around anomaly detection to address previously mentioned shortcomings. Network packet metadata shows significant distribution variability with multiple attack signatures ranging from malicious downloads, brute force attempts, vulnerability scans, and malicious command execution.
Once the variational autoencoder is trained, separate packet metadata can be passed to the autoencoder to compute a constructed probability based on the multivariate normal probability distribution function. This system is orthogonal to rule-based network protection systems, including firewall rule implementations and exceptions, and can be deployed in conjunction with such technologies. Unlike rule-based systems, packets are blocked entirely based on the machine learning-based reconstructed probability score, with the only user-tuned feature being the threshold for anomaly designation.
The benefits are impactful and principally include:
- Allowing an organization to detect zero-day attacks.
- Possessing existing rule-based network protection systems applicability.
SRV was advised on this transaction by The Alchemy Group (Alchemy). Alchemy’s mandate going-forward will include arranging the commercialization capital structure for this technology capability, advising on follow-on acquisitions and the capitalization structure of a globally diversified technology portfolio, and identifying strategic operating and/or re-licensing partners to accelerate the deployment of existing and future capabilities in the private sector.
Alchemy’s broker-dealer affiliate, Alchemy Securities, is serving as lead financial advisor to CEL with respect to corporate finance and capital-raising activities with the aim to package and enhance CEL’s product and serving offerings to existing clientele, and to scale CEL’s entry into new and emerging growth markets.
Alchemy CEO, Henry Huang, stated, “Alchemy is honored to be working in partnership with SRV and CEL to form new portfolio companies to incubate next-generation technology capabilities and to carry out the mission of the INL to innovate the world’s energy future and secure our nation’s critical infrastructure apparatus.” Alchemy is a privately held merchant banking, alternative asset management, and financial services company that specializes in corporate debt/equity, real estate debt/equity and infrastructure on a global basis. The firm specializes and transacts across the following sectors: (1) Real Estate & Infrastructure, (2) Healthcare & Life Sciences, (3) Technology, Media, Telecommunications & Sports (TMT-S), (4) Metals & Mining and (5) Renewables.
The SRV team has decades of experience conveying results stemming from scientific and technological research to the marketplace and to wider society, along with associated skills and procedures. SRV relationships reach deep into industry and finance and are themselves an intrinsic part of the technological innovation process. SRV Managing Partner, Robert Riegle, stated, “Technology transfer is a complex process that involves many non-scientific and non- technological factors, and many different stakeholders. Good or high-quality research results are not enough for successful technology transfer; general awareness and willingness both at the level of organizations and individuals, as well as skills and capacity related to specific aspects, such as access to risk finance and intellectual property (IP) management, are also necessary components.” Sub Rosa provides end-to-end expertise across the entirety of the process.
Parties interested in learning more about this ground-breaking technology are encouraged to reach out to Dr. Riegle directly at: 330 977-1897.
Contact:
Robert Riegle, J.D.
330 977-1897
[email protected]
SOURCE Cleveland Electric Laboratories
