Cooperative communication aided downlink DS-CDMA using preprocessing based on vector quantized channel impulse responses

Abstract

In this contribution, we investigate the performance of downlink (DL) direct sequence code division multiple access (DS-CDMA) system with multi-user transmitter preprocessing (MUTP) based on noisy feedback of vector quantized (VQ) channel impulse responses (CIRs). Specifically, in this contribution, the channel impulse responses required to conceive the preprocessing matrix are estimated at each of the mobile stations (MSs) and conveyed to the base station (BS) with the aid of feedback channels that conflict noise. In particular, the CIRs are estimated with a known estimation technique, and the estimated CIRs are VQ and the magnitudes and phases are fed back through noisy feedback channels to the BS. At the BS, this noise infected VQ-CIRs are recovered using a linear detector and the recovered CIRs are then exploited to formulate the preprocessing matrix to mitigate the DL multi-user interference (MUI). Our simulation study shows that MUTP realized with perfect CIRs completely eliminates the DL MUI whereas, noise tainted VQ-CIRs based MUTP results in imperfect removal of MUI as the feedback channel induced errors play a domineering role in determining the performance of the system in the context of interference elimination. Further, the attainable bit error rate (BER) degrades when noise contaminated VQ-CIRs are invoked to realize the preprocessing matrix compared to its perfect CIRs counterpart. Nevertheless, from our simulation results it can be inferred that vector quantization can serve as an efficient approach to quantize the CIRs at the MSs and feedback them via ideal feedback channels to the BS to formulate the preprocessing matrix particularly in frequency division duplexing (FDD) type wireless systems.