Localizing the root cause of network faults is crucial to network operation and maintenance. Significant operational expenses will be saved if the root cause can be identified agilely and accurately. However, this is challenging for human beings due to the complicated wireless environments and network architectures. Resorting to data analysis and machine learning is promising but remains difficult due to various practical difficulties, such as the lack of well-labeled samples, hybrid fault behaviors, missing data, and so on. The overall target of this competition is to help people better understand the root cause localization problem and open up new paths for the next-generation wireless networks empowered by causal inference techniques.
We provide a novel real-world telecommunication dataset collected from live 5G networks, together with the associated causal graph endorsed by human experts. Various practical problems were born with this dataset due to the underlying system limitations, such as how to efficiently handle the short of well-labeled data, how to handle the missing data issue, how to combine classic time series analysis with new-fashioned learning-based models, how to perform accurate and reliable causal inference, and so on. We hope these interesting problems can arouse the broad attention of scholars as well as practitioners with different backgrounds and inspire new ideas during the competition. Full support on the data and platform will be provided by the organizing team.