NEC, Japan's NICT Use AI-based Services Classification to Optimize Allocation of Network Resources

NEC announced the successful demonstration of technologies that use AI to automatically classify services, such as video data and Web browsing, based on the characteristics of network traffic.

This demonstration, jointly conducted with Japan's National Institute of Information and Communications Technology (NICT) in February 2018, contributes to the optimal allocation of network resources by automatically classifying services based on their traffic characteristics and visualizing their status of use, said NEC.

This demonstration utilized JGN and RISE, NICT's test bed networks for research and development, Keysight Technologies' network packet broker, and NEC' s "Context-aware Service Controller".

The Context-aware Service Controller automatically classifies services by extracting traffic characteristics based on the chronological data of the traffic and visualizes the status of network use for services by utilizing "NEC Advanced Analytics – RAPID machine learning," which is software that features deep learning technology and belongs to NEC's advanced AI technology series, "NEC the WISE." This reduces the prior setup work that is required of network operators in order to classify services. 

For this demonstration, NEC succeeded in automatically classifying services based on their traffic characteristics. The characteristics of traffic from three different services were processed and learned by NEC's Context-aware Service Controller.

The three services included video data from professional baseball training, video data captured by an IP camera, and lifestyle traffic (such as Web browsing).

This week, NEC and MegaFon also announced the successful completion of AI technology field tests in MegaFon's live network in order to improve the efficiency of planning and maintenance of transport network resources.