“Fifth-generation cellular networks are designed to cater not only for more stringent bandwidth, latency and scalability requirements, but to accelerate digital transformation across industries with diverse use-cases.
In the near future, LPWAN technologies such as NB-IoT will be able to penetrate underground structures to reach sensors that last 10 years on a single battery charge, and with 5G NR, radio access round trip times lower under 4 ms to enable remote monitoring and control for autonomous vehicles. These networks will bring scenarios from science fiction to our everyday lives, enable vastly different business models and elevate consumer expectations to the next level.
Therefore, at the same time, it will be necessary to embed compute capabilities in the network; not only to leverage diminished access latencies, but to improve economies of certain compute heavy applications like media processing, computer vision and analyzing the volatile data of the massive amount of sensors. This will inevitably increase infrastructure complexity to a level where traditional management tools and methods will not be able to satisfy industry requirements, and gradually introduce machine intelligence into network and workload management.”