Cyber-attacks cost the world economy $6 trillion per year and are increasing by 15 percent per year. By 2025, the world economy will have to pay a heavy price of $10 trillion for such attacks. A major factor contributing to these attacks is inadequate attention paid to computer network architectures. Accordingly, automation of technologies, like firewalls, that keep check of network architectures is desired. An automated firewall will be able to point out the rules missing or changes to rules when required. Along with automation it is also important to keep a human in the loop to overrule incorrect suggestions from the technology. In order to generate firewall rules that are effective, there are several thousands of network flows that need to be analyzed, which is simply beyond a reasonable execution time for a human. Overall, targeted approach is to use a machine learning algorithm in order to derive more complex rules that can simultaneously consider many more inputs.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Kothari, Vinit
• Perez, Jerry
• Iandoli, Matthew J.
• Zhang, Jiaqi

Subject

• Security
• Innovation
• Project-Based Learning
• Computing

Publisher

• Worcester Polytechnic Institute

Identifier

• 22651
• E-project-050621-062450

Mot-clé

• Machine Learning
• JupyterLab
• Generative Adversarial Networks
• User Interface
• Cybersecurity
• Docker

Advisor

• Paffenroth, Randy Clinton
• Mangoubi, Oren Rami
• Trapp, Andrew C

Year

• 2021

Sponsor

• Arco IT GmbH

Date created

• 2021-05-06

Resource type

• Major Qualifying Project

Major

• Mathematical Sciences
• Computer Science

Rights statement

• In Copyright

Dernière modification

• 2021-08-29