Distributed Scheduling Achieves the Optimal Multiuser Diversity Gain for MIMO-Y Channel

A distributed user scheduling scheme is proposed to harvest multiuser diversity (MuD) for the MIMO-Y channel. A three-party information exchange model is considered, where three representatives are selected from three clusters of users to exchange information via a fixed-gain amplify-and-forward (AF) relay. In particular, a novel MIMO-Y transmission scheme is proposed with the reference signal space (RSS), which enables the distributed transmit/receive beamforming and scheduling of the users. It is shown that the proposed distributed scheduling achieves the same MuD gain as the optimal centralized one, while significantly reducing the complexity and feedback. Moreover, the outage probability of the proposed scheme is quantified, and the achievable MuD gain is shown to be the minimum number of candidates in the three clusters. Finally, numerical results validate our theoretical derivations.