Causal Modeling-Based Multivariate KPIs Prediction for Parameter Adjustment in Cellular Networks
Zhenyu Zhang, Qi Li, Yong Zhang, Yuemeng Zhang, Tianmu Sha, Xiaolei Hua, Lin Zhu, Junlan Feng · IEEE Transactions on Vehicular Technology · 2025
The rapid development of 5G technology has led to the emergence of various Configuration and Optimization Parameters (COPs), which significantly influence network Key Performance Indicators (KPIs). This paper aims to accurately predict long-term KPI changes for parameter adjustments in cellular networks. Existing studies have employed causal inference to estimate the short-term effects of COP adjustments but often overlook the low-frequency nature of COP adjustments in real-world networks. The conventional time-series causal graph (CTSCG) primarily focuses on the direct causal relationships between COPs and KPIs. Given the infrequent changes in COPs data and the highly complex, dynamic nature of KPIs data, models based on CTSCG struggle to accurately characterize the temporal dynamics of KPIs with limited COPs variations. Furthermore, prior research on multivariate time series (MTS) has primarily relied on time-domain and frequency-domain methods, which struggle to effectively capture the dynamic characteristics of non-stationary data. To address these limitations, Wave-CMNet (Wavelet-domain Causal Modeling Network) is proposed, an innovative causal modeling framework for predicting multivariate KPIs. This framework models the effects of COP adjustments on KPIs as Individual Treatment Effects (ITEs) within a point treatment setting in causal inference. To mitigate the challenges posed by low-frequency COP adjustments, the framework incorporates pre-treatment outcomes as covariates into the causal graph, thereby uncovering intrinsic dynamic patterns within time series data and improving longterm prediction performance. Wave-CMNet leverages the timefrequency localization properties and multi-resolution analysis capabilities of the wavelet domain to effectively capture both frequency-domain and time-domain features. Subsequently, the model represents time-series data in the wavelet domain and decomposes the MTS representation into time-dependent and inter-variable features to facilitate targeted learning. Extensive evaluations on real-world cellular network datasets confirm the effectiveness of Wave-CMNet.