A Socially Aware Many-to-Many Matching Approach for Access Point Selection in Cell-free Massive MIMO

Cell-free massive MIMO has emerged as a key technology that is envisioned to play a central role in future wireless networks. It leverages the benefits of massive MIMO by utilizing a large number of access points (APs) distributed over a large coverage area to concurrently serve multiple user equipment (UEs). In its canonical form, each user is served by all the APs which is impractical. It is therefore important to carefully select groups of APs that will participate in serving each UE. In this work, we propose a method to form efficient UE-AP association clusters using matching theory, by modeling the problem as a many-to-many matching with externalities. We consider that the UEs exhibit partially altruistic behavior and therefore select APs in an empathetic way, aiming to improve their neighbor’s rate in addition to their own. Simulation results show that we can improve the system sum spectral efficiency, outage probability and the average energy efficiency compared to existing methods.