Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Towards xApp Conflict Evaluation with Explainable Machine Learning and Causal Inference in O-RAN

Pragya Sharma, Shihua Sun, Shachi Deshpande, Angelos Stavrou, Haining Wang

TL;DR

This work proposes a framework for xApp conflict management that combines explainable machine learning and causal inference to evaluate the causal relationships between RAN Control Parameters (RCPs) and Key Performance Indicators (KPIs).

Abstract

The Open Radio Access Network (O-RAN) architecture enables a flexible, vendor-neutral deployment of 5G networks by disaggregating base station components and supporting third-party xApps for near real-time RAN control. However, the concurrent operation of multiple xApps can lead to conflicting control actions, which may cause network performance degradation. In this work, we propose a framework for xApp conflict management that combines explainable machine learning and causal inference to evaluate the causal relationships between RAN Control Parameters (RCPs) and Key Performance Indicators (KPIs). We use model explainability tools such as SHAP to identify RCPs that jointly affect the same KPI, signaling potential conflicts, and represent these interactions as a causal Directed Acyclic Graph (DAG). We then estimate the causal impact of each of these RCPs on their associated KPIs using metrics such as Average Treatment Effect (ATE) and Conditional Average Treatment Effect (CATE). This approach offers network operators guided insights into identifying conflicts and quantifying their impacts, enabling more informed and effective conflict resolution strategies across diverse xApp deployments.

Towards xApp Conflict Evaluation with Explainable Machine Learning and Causal Inference in O-RAN

TL;DR

This work proposes a framework for xApp conflict management that combines explainable machine learning and causal inference to evaluate the causal relationships between RAN Control Parameters (RCPs) and Key Performance Indicators (KPIs).

Abstract

The Open Radio Access Network (O-RAN) architecture enables a flexible, vendor-neutral deployment of 5G networks by disaggregating base station components and supporting third-party xApps for near real-time RAN control. However, the concurrent operation of multiple xApps can lead to conflicting control actions, which may cause network performance degradation. In this work, we propose a framework for xApp conflict management that combines explainable machine learning and causal inference to evaluate the causal relationships between RAN Control Parameters (RCPs) and Key Performance Indicators (KPIs). We use model explainability tools such as SHAP to identify RCPs that jointly affect the same KPI, signaling potential conflicts, and represent these interactions as a causal Directed Acyclic Graph (DAG). We then estimate the causal impact of each of these RCPs on their associated KPIs using metrics such as Average Treatment Effect (ATE) and Conditional Average Treatment Effect (CATE). This approach offers network operators guided insights into identifying conflicts and quantifying their impacts, enabling more informed and effective conflict resolution strategies across diverse xApp deployments.
Paper Structure (17 sections, 6 figures, 4 tables)

This paper contains 17 sections, 6 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. xApp Conflicts Definition and Modeling
  4. Conflict Detection and Impact Estimation
  5. Motivation and Overview of Proposed Approach
  6. Conflict Detection
  7. Data Requirements
  8. ML Regression and Model Explainability
  9. Conflict Impact Estimation
  10. Causal Directed Acyclic Graph (DAG) Construction.
  11. Treatment Effect Estimation
  12. Evaluations
  13. Network Simulation
  14. Conflict Identification via Explainable ML Regression
  15. Conflict Effect Evaluation via Causal Inference
  16. ...and 2 more sections

Figures (6)

  • Figure 1: Example of potential conflicting interactions between xApps with different objectives.
  • Figure 2: Directed graph with edges representing direct, indirect and implicit conflicts among xApps.
  • Figure 3: Proposed approach towards conflict evaluation
  • Figure 4: SHAP value plots indicating the feature importance of each RCP on KPIs (a) Throughput, (b) Spectral Efficiency, and (c) Block Error Rate (BLER)
  • Figure 5: Causal Directed Acyclic Graph (DAG)
  • ...and 1 more figures