A Many-to-Many Matching Approach for Access Point Selection in Cell-free Massive MIMO with Altruistic Players

Cell-free massive MIMO is a promising technology to meet the requirements of future wireless networks. It uses multiple access points (APs) spread over a large coverage area to serve multiple User Equipment (UEs) simultaneously. In order to make it practical and ensure good performance, it is important to carefully determine which APs should serve particular users. In this work, we propose a method to efficiently select these clusters using matching theory - by modeling the problem as a many-to-many matching with externalities. We consider both the UEs and APs to be altruistic, and therefore select AP clusters with the goal of improving the network sum spectral efficiency, rather than their own rate. Simulation results show that we can improve the outage probability and the average energy efficiency compared to existing methods.