Using autoregressive and ADALINE neural network modeling to improve downlink performance of smart antennas

In time-division-duplex (TDD) mode wireless communications, downlink beamforming performance of a smart antenna system at the base station can be degraded due to variation of spatial signature vectors of an antenna array especially in fast fading scenarios. To mitigate this, downlink beams must be controlled by properly adjusting their weight vectors in response to changing propagation dynamics. This can be achieved by modeling the spatial signature vectors in the uplink period and then predicting them for the new mobile position in the downlink transmission period. In this paper, we show that linear prediction of spatial signature vectors using autoregressive (AR) and ADALINE (adaptive linear neuron) network modeling provide certain level of performance improvement compared to conventional beamforming method under varying channel (mobile speed) and filter (delay) order conditions. ADALINE structure outperforms AR modeling in terms of downlink SNR improvement and relative error improvement.