Capacity and downlink transmission algorithms for a multi-user MIMO channel

Abstract

Downlink beamforming in a multi-user MIMO channel can provide significant gain in system throughput by allowing space division multiple access (SDMA). The exact solution for the sum capacity of such channels does not exist in closed form, but requires an expensive iterative algorithm. By imposing certain constraints on the capacity equation, a sub-optimal closed-form solution can be obtained. The paper presents two such solutions. The first, referred to as "block-diagonalization" arises from forcing all inter-user interference to zero. The second - "successive optimization" - is an alternative method of minimizing transmit power for an arbitrary information rate per user, and allows solutions which, under certain circumstances, are superior to the block-diagonalization approach. Both algorithms have sub-optimal performance, but they lead to simpler transmitter and receiver structures, and allow a tradeoff between performance and complexity.