As networks continually evolve, it becomes difficult for human operators to continually keep up-to-date with the latest anomalies. The unique algorithm enables networks to continually identify new anomalies and avoid becoming an obsolete tool.
Cybercrime is the greatest threat to virtually every company in the world; the damage that malware and viruses cause is predicted to cost the world $6 trillion annually by 2021.
Currently, networks can be trained to detect known malware and viruses—but with a constant flow of new data traffic, and a proliferation of new devices, networks struggle to determine new threats. Be able to determine what is ‘normal traffic’ becomes difficult—particularly when ‘normal’ is constantly changing.
Victoria University of Wellington researchers may have the answer to the problem with a novel machine learning technique they have developed to detect unknown incoming threats to IT networks. By being able to inform networks to constantly train and learn what is ‘not normal’, the tool aims to help IT network operators identify major anomalies before they create significant damage.
Features and benefits
The machine learning tool enables operators to identify anomalies at point of entry, allowing them to enact preventative actions to protect the network.
Meets industry needs
Industry engagement and feedback throughout the invention process has validated the technology and ensured it is tailored to meet industry needs.
The team is currently doing some preliminary market feedback and looking to build a working proof-of-concept in a real network environment. At the end of this field trial, the team aims to have a hardware/ software product that can potentially be integrated into existing networks as a ‘plug and play’ solution.
We are now seeking industry partners i.e. enterprise level firms, network hardware providers and telecommunications operators, to help trial the technology within their existing networks.