Adaptive Limited Feedback for Sum-Rate Maximizing Beamforming in Cooperative Multicell Systems


Ramya Bhagavatula and Robert W. Heath, Jr.


submitted to IEEE Transactions on Signal Processing.


Base station cooperation improves the sum-rates that can be achieved in cellular systems. Conventional cooperation techniques require sharing large amounts of information over finite-capacity backhaul links and assume that base stations have full channel state information (CSI) of all the active users in the system. In this paper, a new limited feedback strategy is proposed for multicell beamforming where cooperation is restricted to sharing only the CSI of active users among base stations. The system setup considered is a linear array of cells based on the soft handoff model, where each cell contains singleantenna users and multi-antenna base stations. Beamforming vectors are designed using a generalized eigenvector approach to maximize the sum-rate in a single-interferer scenario, at high signal to noise ratio. Users are assumed to feedback quantized CSI of the desired and interfering channels using a finite-bandwidth feedback link. An upper bound on the mean loss in sum rate due to random vector quantization is derived. A new feedback-bit allocation strategy, to partition the available bits between the desired and interfering channels, is developed to reduce the mean loss in sum-rate due to quantization for the soft handoff model. The proposed feedback-bit partitioning algorithm is shown, using simulations, to yield sum-rates close to the those obtained using full CSI at base station.

Preprint available at: [ArXiv].