Distributed cell-free processing with dynamic AP clusters

Massive MIMO is a promising technology for current and future wireless networks in which a massive number of antennas per base station serve users. In conventional cellular networks, each base station has a non-overlapping coverage area (a cell) and each user is only served by the one base station that is serving the particular cell in which the user is located. On the other hand, the concept of 'Cell-Free networks' got a lot of traction recently. In this paradigm, antennas are distributed as much as possible in different access points (AP) to exploit spatial diversity. In such a network, the users are no more restricted to one AP. In other words, the boundary between different cells is removed. Furthermore, in serving users the APs can cooperate by the use of a central processing unit (CPU). This change in topology is expected to boost the spectral efficiency of wireless networks substantially. However, this comes at the cost of high complexity. In this thesis, we try to come up with a new algorithm for signal processing that exploits AI methods to do better than linear combining and precoding which is already present in literature.