On the Application of Reliability Theory to Cellular Network Mobility Performance Analysis

TL;DR

This paper models cellular network outages using reliability theory, introducing new KPIs like mean outage time and mean time between outages to better capture the stochastic temporal aspects of network reliability.

Contribution

It extends traditional reliability metrics to cellular networks by defining outage-specific KPIs and demonstrates their application through system-level simulations with user mobility.

Findings

New KPIs effectively quantify cellular network reliability.

Outage metrics correlate with application performance requirements.

Simulation results validate the proposed reliability modeling approach.

Abstract

Achieving connectivity reliability is one of the significant challenges for 5G and beyond 5G cellular networks. The present understanding of reliability in the context of mobile communication does not adequately cover the stochastic temporal aspects of the network, such as the duration and spread of packet errors that an outage session may cause. Rather, it simply confines the definition to the percentage of successful packet delivery. In this letter, we offer an elaborate modeling of the outage for a cellular mobile network by showcasing the different types of outages and their contiguity characteristic. Thereafter, using the outage metrics, we define two new key performance indicators (KPIs), namely mean outage time and mean time between outages as counterparts to akin KPIs that already exist in classical reliability theory, i.e., mean down time and mean time between failures. Using a system-level simulation where user mobility is a crucial component, it is shown that these newly defined KPIs can be used to quantify the reliability requirements of different user applications in cellular services.