Carrier China Telecom and Chinese vendor ZTE have built an adaptive spatio-temporal awareness network that's easy for them to use.
Based on ZTE's Radio Composer, the Adaptive Spatio-Temporal Awareness Network is designed to help maintain connectivity in "high-capacity scenarios" - which can mean a lot of people trying to access the network simultaneously in a small area, like a soccer field.
Adaptive Networks
I want to synchronize the spatio-temporal gubbins of network resources and traffic distribution more precisely and efficiently through "on-demand elastic coverage of a two-tier network". the network intelligently predicts traffic distribution and can react accordingly, while the LSTM (Long Short Term Memory) algorithm analyzes and predicts traffic distribution at the physical grid level.
What is the point of that? After making predictions, the spatio-temporal cognitive network implements a power scaling function to "enable flexible coverage adjustment". Basically, this means that it provides more power to cover high-load areas and less power to low-load areas, which it says "precisely" matches radio resources to demand. Apparently, the network also builds a knowledge base of the base station's AI logic grid to further balance network efficiency.
There aren't even the usual canned quotes explaining how revolutionary the technology is and how proud they are of it, etc. This sometimes results in empty statements, but it does at least add a little humanity to the very technical announcement. Instead, China Telecom and ZTE seem happy to throw up their hands and tell us they've completed a nice new project and shelved it, which is fair enough.
So there you have it, an AI-driven adaptive network. It's not exactly "self-aware", but we're getting there ......
