SRK 2022

POVZETEK
Aplikacije za potrebe industrije so eno izmed najpomembnejših področij, ki jih naslavlja tehnologija 5G, saj funkcionalnosti prejšnjih generacijah mobilnih tehnologij niso ravno najbolje dosegale zahtev industrijskih aplikacij, kot so npr. pametne tovarne, avtonomna mobilnost, gosta senzorska omrežja (npr. pametna mesta), ipd. Vsaka aplikacija oz. storitev v 5G omrežju ima svoje specifike in zahteve, prav tako je tudi samo omrežje 5G dokaj kompleksno, zato si uvajanje novih 5G storitev težko zamislimo brez ustreznega testiranja vse od zasnove storitve do njene produkcijske uporabe. Cilj testiranja je pridobiti čim več relevantnih podatkov o delovanju storitve (in tudi omrežja) v različnih okoliščinah, saj se na podlagi temeljite analize izmerjenih podatkov odločamo o nadaljnjih korakih prilagajanja oz. optimiranja storitve in omrežja s ciljem čim boljše uporabniške izkušnje ob sočasni optimalni izrabi virov. Da bi zagotovili čim bolj relevantne rezultate, v testnih scenarijih ne smemo pozabiti na testiranja “end-to-end” in na realistično generiranje testnega prometa. Poleg izzivov in možnih pristopov k testiranju v 5G bomo v prispevku predstavili tudi nekaj konkretnih primerov iz prakse.

ABSTRACT
Industry related use cases are among most promising for the 5G (e.g., factories of the future, logistics, power grids, etc.), however, since not two use cases have the same requirements, combining technological features to realize required service is a challenge in each single case. Due to the complexity of the 5G, testing plays a critical role since the very first stages of the network integration (e.g., lab testing) and remains during the production stage as monitoring, also serving as an input for the network optimization. It is a good idea to have ability to collect as much as possible parameters related to the service tested, as well, it is also helpful to collect data in various situations, e.g., continuous vs. on-demand testing, fixed locations testing vs. drive/flight testing. Further (AI-based) data analysis may show what data are meaningful in a certain use case. Since services created within the 5G ecosystem tend to tightly couple communications and data processing features, testing should not be limited to network parameters only, but should also include other end-to-end service parameters/KPIs relevant for the customer experience. In the article, we are, besides challenges and possible approaches, also discussing on practical experiences gained in the field.