Accurate cellular traffic forecasting requires modeling both local topology-driven interactions and long-range spatio-temporal dependencies. This paper proposes ST-GAT, a dual-branch framework that combines Chebyshev graph convolution with multi-head spatial self-attention and a Temporal Transformer. An Adaptive Spatio-Topological Gating Mechanism performs input-dependent fusion of the local graph representation and the global attention representation. On the Milan Internet-traffic test set, ST-GAT with adaptive gating achieves an MAE of 77.7318, an RMSE of 136.8921, and an $$R^{2}$$ R 2 of 0.9683, outperforming the considered baselines under the evaluated experimental setting. Additional experiments on SMS and call traffic show that the same architecture remains competitive across different cellular service types, although the magnitude and consistency of the improvement are service-dependent. A stratified analysis of public holidays, weekends, and non-holiday weekdays further shows that the reported prediction metrics vary across calendar conditions. These results support the applicability of the proposed architecture within the evaluated offline forecasting setting, while continual adaptation, cross-city transfer, and evaluation under broader temporal conditions remain directions for future work.
Karimi Kashani, Alireza, Soleimani, & Hossein (2026). ST-GAT: a spatial–temporal graph-augmented transformer for cellular network traffic forecasting. Telecommunication Systems. https://doi.org/10.1007/s11235-026-01516-x
Karimi Kashani, et al. “ST-GAT: a spatial–temporal graph-augmented transformer for cellular network traffic forecasting.” Telecommunication Systems, 2026. doi:10.1007/s11235-026-01516-x.
@article{Kashani2026,
title = {ST-GAT: a spatial–temporal graph-augmented transformer for cellular network traffic forecasting},
author = {Kashani, Karimi and Alireza and Soleimani and Hossein},
journal = {Telecommunication Systems},
year = {2026},
doi = {10.1007/s11235-026-01516-x},
url = {http://link.springer.com/openurl/fulltext?id=doi:10.1007/s11235-026-01516-x}
}
In disaster relief scenarios, traditional communication infrastructures are often destroyed, leaving rescue personnel with limited computation capability and energy on their end devices. These devices generate continuous information and...
Cell-free massive multi-input multi-output (MIMO) is one of the architectures that look highly favorable for 6G networks as geographically distributed access points (AP) are able to together serve users without any set cell boundaries....
The blockchain technology based on the existing stack of the Internet (i.e., TCP/IP) faces significant challenges regarding scalability, throughput, and latency, which are critical to its operation. This paper proposes shifting from an...
Recent advances in self-supervised vision transformers—particularly DINOv2—have shown impressive capabilities in learning rich visual representations from natural images. However, their potential for building robust and deployable...
Free-space optical communication provide high bandwidth and flexible deployment, however its exposure to the atmosphere makes it vulnerable to environmental degradation and security threats. To address these challenges, this paper...