Causal Beam Selection for Reliable Initial Access in AI-driven Beam Management

2025-08-22

AI Summary

To address poor interpretability, weak generalization, and high scanning overhead in beam alignment for millimeter-wave (mmWave) MIMO initial access, this paper proposes a causality-aware deep learning framework. Methodologically, it introduces a two-stage causal beam selection algorithm: first, causal discovery identifies the minimal sufficient input set and constructs a Bayesian causal graph; second, a deep classifier guided by causal features performs beam prediction. The key contribution lies in embedding causal inference directly into the beam alignment pipeline, explicitly modeling the causal mechanisms between inputs and outputs. Experiments under representative 6G scenarios demonstrate that the method maintains prediction accuracy while reducing input selection time by 94.4% and beam scanning overhead by 59.4%, significantly enhancing robustness, interpretability, and energy efficiency.

Technology Category

• Machine Learning: Causal Learning
• Search and Optimization: Algorithm Configuration
• Reasoning under Uncertainty: Causality

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Efficient and reliable beam alignment is a critical requirement for mmWave multiple-input multiple-output (MIMO) systems, especially in 6G and beyond, where communication must be fast, adaptive, and resilient to real-world uncertainties. Existing deep learning (DL)-based beam alignment methods often neglect the underlying causal relationships between inputs and outputs, leading to limited interpretability, poor generalization, and unnecessary beam sweeping overhead. In this work, we propose a causally-aware DL framework that integrates causal discovery into beam management pipeline. Particularly, we propose a novel two-stage causal beam selection algorithm to identify a minimal set of relevant inputs for beam prediction. First, causal discovery learns a Bayesian graph capturing dependencies between received power inputs and the optimal beam. Then, this graph guides causal feature selection for the DL-based classifier. Simulation results reveal that the proposed causal beam selection matches the performance of conventional methods while drastically reducing input selection time by 94.4% and beam sweeping overhead by 59.4% by focusing only on causally relevant features.

Problem

• Addressing poor generalization in DL-based beam alignment methods
• Reducing unnecessary beam sweeping overhead in mmWave MIMO systems
• Identifying minimal causally relevant inputs for beam prediction

Innovation

• Causally-aware DL framework integration
• Two-stage causal beam selection algorithm
• Bayesian graph-guided feature selection