A Time Series Based Study of Correlation, Channel Power Imbalance and Diversity Gain in Indoor Distributed Antenna Systems at 60 GHz

In this paper, we investigate the potential enhancements in signal reliability which can be achieved by using a millimeter-wave distributed antenna system (DAS) within an indoor environment. To achieve this, we measured the signal power simultaneously received at 9 ceiling mounted access point (AP) locations likely to be used in future indoor DAS deployments while a mobile user imitated making a voice call on a hypothetical user equipment. Key metrics, associated with the performance of multiple antenna systems, such as the cross correlation coefficient (CCC) and channel power imbalance (CPI) are determined. It was found that line-of-sight (LOS) and quasi-LOS (QLOS) links with the APs typically led to higher CCC values than the non-LOS (NLOS) cases. Similarly, LOS and QLOS links typically produced higher CPI values between APs than the NLOS case. To enable the reproduction of our results, we have successfully employed autoregressive moving average and autoregressive integrated moving average modeling to the CCC and CPI time series. The performance improvement that can be achieved using a DAS instead of a single AP was evaluated using three commonly deployed diversity combining schemes, namely, selection combining, equal gain combining and maximal ratio combining along with three AP selection mechanisms, namely, per-sample random AP selection, one-shot AP selection and per-sample optimal AP selection. Finally, we have provided some useful insights into the influence of differing AP numbers on the diversity gain when considering the aforementioned AP selection methods.