Access Point Selection Strategies for Indoor 5G Millimeter-Wave Distributed Antenna Systems

In this paper, we study the use of three candidate Access Point (AP) selection mechanisms for use with indoor millimetre-wave (mmWave) Distributed Antenna Systems (DASs). These are per-sample random AP selection, one-shot AP selection and per-sample optimal AP selection. To facilitate our analysis, we used a customized measurement system operating at 60 GHz, to record the signal power time series simultaneously received at 9 ceiling mounted AP locations while a mobile user imitating a voice call application. Using the time series data, the localized cross correlation coefficient (CCC) was subsequently estimated from the raw received signal strength (RSS) using the Spearman's rank-order correlation. It was found that the resultant time series of the localized CCCs was well-described by the Gaussian distribution across all of the considered measurement scenarios. Moreover, it was observed that the line-of-sight (LOS) and quasi-LOS (QLOS) paths typically led to higher CCC values with broader spreads than the non-LOS (NLOS) scenarios. To study the potential for signal enhancement by using diversity combining in mmWave DASs, we applied selection combining, equal gain combining and maximal ratio combining before investigating the AP selection mechanisms. Finally, we provide some useful insights into the influence of differing AP numbers on the diversity gain when considering the aforementioned AP selection methods.