Multiuser diversity for MIMO-Y channel: Max-min selection and diversity analysis

In this paper, a MIMO-Y channel based three-group information exchange problem is considered, where users from each group would like to exchange information with the other users in another two groups. In particular, a Max-Min user selection is proposed, which aims to harvest the multiuser diversity gain for reliable transmission in the MIMO-Y channel. The impacts of user configurations on the multiuser diversity gains are investigated by theoretical bounds as well as numerical results. It is shown that, by adding a few users in only one or two groups, the overall system performance can be improved significantly. This observation reveals an interesting behavior of MIMO-Y channel, i.e., the local configuration has a global impact. However, we prove that such asymmetrical multiuser diversity gain does not scale with the number of users. Finally, we show that multiuser diversity gain scales with the number of users if we add an equal number of users in all three groups.