Power allocation for indoor distributed antenna system based on path loss prediction at 6 GHz

Due to the difficulty in obtaining the perfect channel state information (CSI), in this paper, the power allocation algorithm is investigated with only large scale information. Based on the measured channel data at 6 GHz, log-distance (LGD) model and linear-distance (LID) model are used to predict the large scale information or path loss (PL) firstly, and both achieve a good prediction precision. Then water-filling algorithm with the predicted PL (WPP) is proposed in this paper for frequency selective fading channel. The proposed algorithm can lead to a 33% increase in average capacity compared with uniform power allocation (UPA) scheme, and outperform the selective transmission (ST) scheme, especially in high transmitted signal to noise ratio (SNR). Besides, the robustness of the proposed algorithm is evaluated, which depends on the difference of PLs among radio links.