Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Explainable and Robust Millimeter Wave Beam Alignment for AI-Native 6G Networks

Nasir Khan, Asmaa Abdallah, Abdulkadir Celik, Ahmed M. Eltawil, Sinem Coleri

TL;DR

The paper tackles mmWave beam alignment for AI-native 6G by proposing a CNN-based beam alignment engine that predicts the best narrow beam from RSSI measurements of a small set of wide beams, significantly reducing IA overhead. To address explainability and reliability, it integrates Deep k-Nearest Neighbors (DkNN) to analyze internal representations, providing p-values, confidence, and credibility to detect out-of-distribution inputs and adversarial perturbations. Experimental results on a Boston downtown scenario show a 75% reduction in beam training overhead with near-optimal spectral efficiency (≈98.5% of the 128-DFT benchmark) and up to 5× improvement in outlier detection robustness over softmax baselines. This work demonstrates that combining DL-based beam prediction with model-agnostic explainability can enhance both operational efficiency and trust in AI-native 6G beam management.

Abstract

Integrated artificial intelligence (AI) and communication has been recognized as a key pillar of 6G and beyond networks. In line with AI-native 6G vision, explainability and robustness in AI-driven systems are critical for establishing trust and ensuring reliable performance in diverse and evolving environments. This paper addresses these challenges by developing a robust and explainable deep learning (DL)-based beam alignment engine (BAE) for millimeter-wave (mmWave) multiple-input multiple-output (MIMO) systems. The proposed convolutional neural network (CNN)-based BAE utilizes received signal strength indicator (RSSI) measurements over a set of wide beams to accurately predict the best narrow beam for each UE, significantly reducing the overhead associated with exhaustive codebook-based narrow beam sweeping for initial access (IA) and data transmission. To ensure transparency and resilience, the Deep k-Nearest Neighbors (DkNN) algorithm is employed to assess the internal representations of the network via nearest neighbor approach, providing human-interpretable explanations and confidence metrics for detecting out-of-distribution inputs. Experimental results demonstrate that the proposed DL-based BAE exhibits robustness to measurement noise, reduces beam training overhead by 75% compared to the exhaustive search while maintaining near-optimal performance in terms of spectral efficiency. Moreover, the proposed framework improves outlier detection robustness by up to 5x and offers clearer insights into beam prediction decisions compared to traditional softmax-based classifiers.

Explainable and Robust Millimeter Wave Beam Alignment for AI-Native 6G Networks

TL;DR

The paper tackles mmWave beam alignment for AI-native 6G by proposing a CNN-based beam alignment engine that predicts the best narrow beam from RSSI measurements of a small set of wide beams, significantly reducing IA overhead. To address explainability and reliability, it integrates Deep k-Nearest Neighbors (DkNN) to analyze internal representations, providing p-values, confidence, and credibility to detect out-of-distribution inputs and adversarial perturbations. Experimental results on a Boston downtown scenario show a 75% reduction in beam training overhead with near-optimal spectral efficiency (≈98.5% of the 128-DFT benchmark) and up to 5× improvement in outlier detection robustness over softmax baselines. This work demonstrates that combining DL-based beam prediction with model-agnostic explainability can enhance both operational efficiency and trust in AI-native 6G beam management.

Abstract

Integrated artificial intelligence (AI) and communication has been recognized as a key pillar of 6G and beyond networks. In line with AI-native 6G vision, explainability and robustness in AI-driven systems are critical for establishing trust and ensuring reliable performance in diverse and evolving environments. This paper addresses these challenges by developing a robust and explainable deep learning (DL)-based beam alignment engine (BAE) for millimeter-wave (mmWave) multiple-input multiple-output (MIMO) systems. The proposed convolutional neural network (CNN)-based BAE utilizes received signal strength indicator (RSSI) measurements over a set of wide beams to accurately predict the best narrow beam for each UE, significantly reducing the overhead associated with exhaustive codebook-based narrow beam sweeping for initial access (IA) and data transmission. To ensure transparency and resilience, the Deep k-Nearest Neighbors (DkNN) algorithm is employed to assess the internal representations of the network via nearest neighbor approach, providing human-interpretable explanations and confidence metrics for detecting out-of-distribution inputs. Experimental results demonstrate that the proposed DL-based BAE exhibits robustness to measurement noise, reduces beam training overhead by 75% compared to the exhaustive search while maintaining near-optimal performance in terms of spectral efficiency. Moreover, the proposed framework improves outlier detection robustness by up to 5x and offers clearer insights into beam prediction decisions compared to traditional softmax-based classifiers.

Paper Structure

This paper contains 13 sections, 14 equations, 4 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. System Model and Problem Formulation
  3. System Model
  4. Deep learning framework for beam alignment
  5. DL-based Beam Alignment
  6. Proposed Explainable and Robust Beam Classifier Framework
  7. Simulation Results
  8. Simulation Setup
  9. Deep Learning Model Architecture
  10. Performance Evaluation
  11. Measurement Noise Analysis
  12. Explainability and Robustness Evaluation
  13. Conclusion

Figures (4)

  • Figure 1: DkNN-based credibility assessment of the proposed beam alignment framework.
  • Figure 2: Accuracy versus SNR values (dB) for different schemes.
  • Figure 3: Spectral efficiency versus the SNR for different schemes.
  • Figure 4: Reliability diagrams for the classifiers on the Boston-5G dataset.