A new publication was recently published in the domain of intelligent data management in IoT networks:
Kylafas, C., Kolomvatsos, K., ‘Adaptive Multi-Level Clustering with Dynamic Representative Selection for Data Redundancy Reduction in IoT Networks’, Computing, Springer, vol. 108, art. no. 183, 2026. https://doi.org/10.1007/s00607-026-01777-z
The paper proposes an adaptive clustering algorithm for IoT sensor networks. It groups nodes into dynamic micro-, meso- and macro-clusters based on their measured values and spatial properties. Base-station-assisted online learning and dynamic representative selection remove most redundant transmissions while keeping reconstruction error low, and the approach outperforms send-on-delta prediction, TEEN and compressed sensing at the same communication cost.
