Cellular capacity maximization via robust downlink beamforming

Abstract

In this paper we propose an optimized multi-cell downlink beamforming solution in which our objective is to maximize capacity of a cellular network. We first formulate an optimization problem, maximizing the received signal power of every active user in a cell, subjected to limiting the overall interference observed by other users below a specified level. In addition, we also put constraint on maximum transmit power of the serving base station. Next, in order to compute robust beamforming vector we accommodate channel estimation error in our formulation. We consider channel imperfection as error between true and estimated channel coefficients and we assume error is bounded with in an ellipsoidal set. The resulting formulation is a non-convex optimization problem and to get a tractable solution we exploit linear matrix inequality based S-procedure. The final reformulation is solved by using semi definite relaxation. The efficacy of proposed solution in improving cellular capacity and efficient power transmission is shown by simulations.