Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AI-Enhanced Spatial Cellular Traffic Demand Prediction with Contextual Clustering and Error Correction for 5G/6G Planning

Mohamad Alkadamani, Colin Brown, Halim Yanikomeroglu

TL;DR

An AI-driven framework that reduces leakage and improves spatial generalization via a context-aware two-stage splitting strategy with residual spatial error correction is presented, supporting more reliable bandwidth provisioning and evidence-based spectrum planning and sharing assessments.

Abstract

Accurate spatial prediction of cellular traffic demand is essential for 5G NR capacity planning, network densification, and data-driven 6G planning. Although machine learning can fuse heterogeneous geospatial and socio-economic layers to estimate fine-grained demand maps, spatial autocorrelation can cause neighborhood leakage under naive train/test splits, inflating accuracy and weakening planning reliability. This paper presents an AI-driven framework that reduces leakage and improves spatial generalization via a context-aware two-stage splitting strategy with residual spatial error correction. Experiments using crowdsourced usage indicators across five major Canadian cities show consistent mean absolute error (MAE) reductions relative to location-only clustering, supporting more reliable bandwidth provisioning and evidence-based spectrum planning and sharing assessments.

AI-Enhanced Spatial Cellular Traffic Demand Prediction with Contextual Clustering and Error Correction for 5G/6G Planning

TL;DR

An AI-driven framework that reduces leakage and improves spatial generalization via a context-aware two-stage splitting strategy with residual spatial error correction is presented, supporting more reliable bandwidth provisioning and evidence-based spectrum planning and sharing assessments.

Abstract

Accurate spatial prediction of cellular traffic demand is essential for 5G NR capacity planning, network densification, and data-driven 6G planning. Although machine learning can fuse heterogeneous geospatial and socio-economic layers to estimate fine-grained demand maps, spatial autocorrelation can cause neighborhood leakage under naive train/test splits, inflating accuracy and weakening planning reliability. This paper presents an AI-driven framework that reduces leakage and improves spatial generalization via a context-aware two-stage splitting strategy with residual spatial error correction. Experiments using crowdsourced usage indicators across five major Canadian cities show consistent mean absolute error (MAE) reductions relative to location-only clustering, supporting more reliable bandwidth provisioning and evidence-based spectrum planning and sharing assessments.
Paper Structure (19 sections, 10 equations, 8 figures, 2 tables)

This paper contains 19 sections, 10 equations, 8 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Data-driven Spatial Traffic Demand Modeling
  3. Grid-cell representation and study areas
  4. Traffic demand proxy from crowdsourced measurements
  5. Feature layers and grid-cell mapping
  6. Spatial Dependency Characterization
  7. Global Moran's I for spatial autocorrelation
  8. Local Moran's I for spatial clusters and outliers
  9. Two-Stage Data Splitting and Spatial Error Correction
  10. Stage 1: spatial clustering for leakage reduction
  11. Stage 2: context-aware refinement within spatial clusters
  12. Learning model and spatial error correction
  13. Performance Evaluation
  14. Cross-city evaluation
  15. Learning curves
  16. ...and 4 more sections

Figures (8)

  • Figure 1: Modeling pipeline.
  • Figure 2: Moran's I versus distance (grid cells).
  • Figure 3: Local Moran's I clusters across five Canadian cities.
  • Figure 4: Two-stage clustering and spatial error correction framework.
  • Figure 5: Comparison of clustering techniques for Montreal.
  • ...and 3 more figures