A New Approach to Mobility Prediction in Small Cell and Femtocell Networks Using Cells' Identities

The paper deals with mobility prediction in small cell and femtocell networks, which is important for automatic resource allocation schemes in 5G/6G. For a long time, the way to approach mobility predictions in cellular networks has been mainly through the coordinate-based trajectory. However, this approach has encountered many drawbacks, especially the drifting effect, which can significantly affect the prediction accuracy. In the cases of small cell networks, where the coverage radius of the cells is very small compared to the micro- and macrocells, such drawbacks can lead to an increase in the next cell's prediction error. To overcome this flaw, this research proposes using cell-based trajectory prediction, a potential solution for mobility prediction, and comparing it with traditional coordinatebased trajectory approaches. The experiment in our contribution shows remarkably better performance and accuracy of the cell-based mobility approach compared to the coordinate-based trajectory approaches.