Abstract
A system for blind topology inference and spatial QoE classification from OTT call transition pair statistics is disclosed. Symmetric transition counts between cell identifier pairs are formed into a weighted graph. A disparity-filtered backbone extraction retains statistically significant edges, and a hybrid classification separates adjacency edges from non-adjacency edges with a numerically stable fallback for large-scale networks. Per-node capture rates are estimated from the inferred adjacency, and QoE metrics are analyzed with spatial autocorrelation. A greedy breadth-first clustering groups cells into operator-actionable geographic clusters, and a log-weighted composite scoring with confidence intervals ranks worst offenders. A two-level comparison across granularities automatically classifies degradation as intra-site or inter-site.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Anonymous, "Blind Topology Inference and Spatial Quality-of-Experience Classification from OTT Call Transition Pair Statistics", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/12526