NEC has unveiled new technology that uses artificial intelligence to analyse network data.
Net2Vec analyses customer data from a range of input sources, including equipment logs, measurement probes, user data and Deep Packet Inspection (DPI) equipment.
The vendor said its platform was backed by “novel deep learning algorithms” based on “patented machine learning technology” that was effective even when data is encrypted (HTTPS) or when cookies are blocked.
The tech is designed to allow operators to identify user needs and improve customer retention, including by offering responsive tariff plans.
Additional security protection, including identifying users infected by malware or with risky behaviour, can also be applied.
The solution is now available for customer trials.
"At present, there are neither any AI solutions to address all of these problems for operators, nor existing solutions optimised for network data," said Dr Saverio Niccolini, General Manager at NEC Laboratories Europe.
"Net2Vec provides an optimised system, tailored to the network, and based on existing deep learning solutions.”
Meanwhile, NEC announced that it had completed joint verification trials in Japan with NTT DOCOMO using Massive MIMO.
The trial achieved spectral efficiency over eight times higher than that of LTE, NEC said.
In the trial, NEC used its active antenna system, which is compatible with the low end of the super-high frequency band. This uses digitised beam control technology, which makes beam forming more precise.
Read more: EE builds a brain as it looks to get back on track with customer experience