Abstract
The use of opportunistic multicast scheduling (OMS) with transmit precoding is examined in this work for multicell MIMO systems with multiple multicast groups. In multicast systems, data transmitted by the base-stations (BSs) must be received by all intended receivers and, thus, the transmission rate is limited by the the worst intended receiver. OMS traditionally has been used to optimize the tradeoff between unicast and broadcast, but can have an even larger impact in systems with multiple multicast groups since users can be further selected in this case to avoid interference. Here, we assume that the base-stations (BSs) have knowledge of the data intended for all groups as well as the channel state information (CSI) of all links. The OMS and transmit precoder are designed based on two optimization criteria: the worst per-user throughput and the weighted sum throughput. Given the user selection, the problem of finding the optimal transmit precoder under both criteria are non-convex and, thus, are alternatively formulated as convex feasibility problems, which are solved by bisection search. Then, an outer optimization is performed to solve for the optimal user selection. However, since the problem is combinatorial in nature, the complexity of the problem can be intractable for systems with large number of users. Hence, iterative user selection (IUS) algorithms are proposed to reduce the complexity of the search. The effectiveness of the proposed schemes are demonstrated through computer simulations.