Impact of Correlated Failures in 5G Dual Connectivity Architectures for URLLC Applications

TL;DR

This paper examines how correlated failures in 5G dual connectivity architectures significantly affect end-to-end reliability for URLLC applications, highlighting the importance of considering failure correlation in network design.

Contribution

It investigates the impact of realistic failure correlations on 5G dual connectivity architectures, revealing that even small correlations can drastically reduce reliability.

Findings

Correlation increases end-to-end error rates by orders of magnitude

Small failure correlations have significant impact on reliability

Results suggest need for alternative architecture designs

Abstract

Achieving end-to-end ultra-reliability and resiliency in mission critical communications is a major challenge for future wireless networks. Dual connectivity has been proposed by 3GPP as one of the viable solutions to fulfill the reliability requirements. However, the potential correlation in failures occurring over different wireless links is commonly neglected in current network design approaches. In this paper, we investigate the impact of realistic correlation among different wireless links on end-to-end reliability for two selected architectures from 3GPP. In ultra-reliable use-cases, we show that even small values of correlation can increase the end-to-end error rate by orders of magnitude. This may suggest alternative feasible architecture designs and paves the way towards serving ultra-reliable communications in 5G networks.