RIS near-field localization paper accepted at IEEE GLOBECOM 2026
The paper “RIS-Aided Near-Field mmWave Localization Under Cross-Link Interference via Beam-Domain ML Fingerprinting” was accepted at IEEE GLOBECOM 2026. It develops a beam-domain machine-learning fingerprinting approach for user localization while accounting for cross-link interference in RIS-assisted near-field mmWave networks.
Paper results
- At 28 GHz with a 20 × 20 RIS, KNN achieved 0.37° angle MAE and 4 cm range MAE under clean conditions.
- Under cross-link interference, KNN recorded 1.4° angle MAE and 7.6 cm range MAE.
- The results show that interference affects angle estimation more strongly than range estimation.
Research experience
This research expanded my experience in modelling interference-aware localization, constructing useful beam-domain fingerprints, and evaluating how interacting wireless links affect learning-based position estimation.
Next work
The next step is experimental validation with measured channels and an investigation of joint communication and sensing strategies that remain reliable under varying interference levels.
Publication citation
M. T. Hassan, D. Zelenchuk, and M. A. B. Abbasi, “RIS-Aided Near-Field mmWave Localization Under Cross-Link Interference via Beam-Domain ML Fingerprinting,” in Proc. 2026 IEEE Global Communications Conference (GLOBECOM), Macau, China, 2026.
