No abstract

No abstract

No abstract

No abstract

Accurate and fast beam-alignment is important to cope with the fast-varying environment in millimeter-wave communications. A data-driven approach is a promising solution to reduce the training overhead by leveraging side information and on the field measurements. In this work, a two-stage tensor completion algorithm is proposed to predict the received power on a set of possible users' positions, given received power measurements on a small subset of positions; based on these predictions, a small subset of beams is recommended to reduce the training overhead of beam-alignment, based on positional side information. The proposed method is evaluated with the DeepMIMO dataset, generated from Wireless Insite, which provides parameterized channels for the experiment. The numerical results demonstrate that, with high probability, the proposed algorithm recommends a small set of beams which contain the best beam, thus achieving correct alignment with small training overhead. Given power measurements on only 8 random positions out of 25, our algorithm attains a probability of incorrect alignment of only 2%, with only 2.7% of trained beams. To the best of our knowledge, this is the first work to consider the beam recommendation problem with only the knowledge at neighboring positions but none at the user's position.