Multi-Cell (MC) systems are present in mobile network operations from the first generation to the fifth generation of wireless networks, and considers the signals of all users to a base station (BS) centered in a cell. Cell-Free (CF) systems explore a large number of distributed antennas serving users at the same time. In this context, Multiple-input multiple-output (MIMO) techniques are used in both topologies and result in performance gains and interference reduction. In order to achieve the benefits mentioned, proper precoder design and power allocation techniques are required in the downlink (DL). In general, DL schemes assume perfect channel state information at the transmitter (CSIT), which is not realistic. This master thesis studies MC and CF with MIMO systems equipped with linear precoders in the DL and proposes an adaptive algorithm to allocate power in the presence of imperfect CSIT. The developed robust adaptive power allocation outperforms standard adaptive (APA) and uniform power allocation (UPA). Simulations also compare the performance of both systems frameworks using MMSE precoders with robust adaptive power allocation and adaptive power allocation.