Toward Zero-Touch Cellular Networks via Next-Generation Crowdsourcing

Abstract

Under the Zero-touch network and Service Management (ZSM) paradigm, cellular networks are moving to fully automated schemes in order to minimize human intervention and maximize efficiency. Until now, such management has been based on network metrics and complemented with drive tests. However, this approach is clearly insufficient to achieve Zero Touch since the information available from the network side is limited, aggregated from all the users within a cell, and collected with a frequency of minutes. Also, the complexity and high cost of performing drive tests to collect user-side data are among the main drawbacks. Nevertheless, there is an option between the two aforementioned: Next-Generation Crowdsourcing Metrics (gCMs). In this regard, the present work describes a novel framework for ZSM cellular networks that exploits the advantages of integrating crowdsourcing User Equipment (UE) side metrics to advanced network management. We analyzed the suitability of the proposal by using data acquired from commercial cellular networks. The obtained results state that there is a wide range of applications for gCMs that can benefit network operation, including the detection of coverage holes, mobility issues, and interference problems. Finally, the identified open challenges are disclosed based on the achieved results.

Keywords

• Crowdsourcing
• KQI
• ML
• QoE
• RAN
• UE