Learning-based angular localization paper published at IEEE ICT 2026
The paper “Efficient Angular Localization in RIS-Assisted mmWave Network via Learning-Based Estimations” was published following its presentation at IEEE ICT 2026. It explores efficient learning-based angular estimation for user localization in a RIS-assisted millimetre-wave network.
Paper results
- KNN achieved 0.79° mean absolute error and an R² score of 0.99 using six probing beams.
- Reduced beam-probing overhead by more than 93% compared with exhaustive search.
- The resulting 1-bit RIS radiation patterns remained within 1°–2° of the ground-truth direction.
Research experience
The project strengthened my experience in building and evaluating a learning-based estimation pipeline, analysing angular localization behaviour, and presenting the contribution within the wider context of intelligent wireless environments.
Next work
Future work will test the estimation approach with measured data and more complex propagation conditions to assess its reliability beyond the original evaluation setting.
Publication citation
M. T. Hassan, D. Zelenchuk, M. A. B. Abbasi, and A. Ullah, “Efficient Angular Localization in RIS-Assisted mmWave Network via Learning-Based Estimations,” in Proc. 2026 32nd International Conference on Telecommunications (ICT), Thessaloniki, Greece, 2026.
