Radio resource allocation is a major research field in wireless networks. The main challenge is to find the most suitable user for each time and frequency resource to ensure the Quality of Service (QoS) requirements. The use of Massive Multiple-Input-Multiple-Output (MIMO) makes it possible to simultaneously schedule several users on the same time and frequency resources. This is a new paradigm for resource scheduling. In this context, resource allocation algorithms have to account for the choice of users served on the same time and frequency resource in their decision mechanism. In this paper we analyze the impact of different group sizes on system capacity using several user selection algorithms and we propose a new user selection algorithm which performs closer to the optimal algorithm proposed in the literature.