Skip to Main Content
Multiple Antenna Downlink Channels: From Theory to Practice The multiple antenna (MIMO) downlink channel has recently been the subject of tremendous interest, primarily because MIMO (spatial multiplexing) benefits can be realized on this channel even if mobiles are not equipped with multiple antennas. Realization of these benefits could have a tremendous impact on a number of wireless systems (e.g., cellular and wireless LAN), but fundamental challenges still exist that make approaching the information theoretic capacity of this channel very difficult in practice. This talk is intended to describe these challenges, introduce some potential solutions, and discuss open issues that remain unanswered, particularly at the interface between the physical and network layers. We first discuss the transceiver complexity required for near optimal operation. Although practical methods for implementing the optimal strategy of dirty-paper coding remain unknown, a low complexity alternative (linear beamforming) exists and we quantify the performance degradation associated with this technique. Next we consider the issue of channel feedback. In order to take full advantage of the potential of the MIMO downlink, it is critical that the transmitter (access point) have accurate channel information. We discuss different channel feedback techniques and the associated resource cost, and quantify the relationship between feedback rate and throughput. Finally we consider the channel from a scheduling perspective. If traffic flows are subject to stringent QoS constraints, scheduling should be performed in a channel- and delay-aware manner on the basis of imperfect channel information.