Ultra-reliable low latency services: 5G architecture and operational alternatives with cost analysis

Our earlier study shows that ultra-reliable low latency (uRLL) services in 5G are economically difficult if existing architectures and operational models are followed due to the very high infrastructure costs. This is mainly driven by the IT admin cost of distributed and attended datacenters necessary to host such services. Therefore, in this paper we analyze potential uRLL use cases and their traffic profiles in more detail, check alternative architecture solutions, and examine the cost saving potential of different datacenter operational models. The results demonstrate that both the alternative operational models and the new network architecture approaches that avoid the tunneling of user plane packets provide significant cost savings compared to the solution based on evolved packet core with a standard operational model.