Millimeter-Wave Channel Estimation Using Coalitional Game

In millimeter-wave (mm-wave) massive MIMO systems, the channel estimation (CE) is a crucial component to set the mm-wave links. Unfortunately, acquiring channel knowledge is a source of training over-head. In this paper, we propose a CE method leveraging measurements at sub 6-Ghz frequencies in order to reduce the training overhead. This solution extracts spatial information from a sub 6-Ghz channel using a virtual channel transformation, such as the searching space is reduced to the information provided by the low frequency channel. In a second stage, a multicell system and its interference between cells is analyzed, proposing a coalitional game to deal with the intercell interference. In the single cell case, we analyze the proposed method in different SNR scenarios, the computational complexity and over user equipment (UE) mobility environment. Finally, we analyze how the coalitional game improves the throughput and its performance over UE in mobility cases.