Estimating the Cross-Correlation Properties of Large-Scale Parameters in Multilink Distributed Antenna Systems: Synchronous Measurements Versus Repeated Measurements

It is essential to capture the cross-correlation properties of large-scale parameters (LSPs) among different base station links in cooperative multilink systems in order to make realistic performance assessments. In this work, propagation measurements are used to study the cross-correlation properties of different LSPs, namely large-scale fading, delay spread, azimuth spread, and elevation spread of four links. The interlink cross-correlation coefficients of these LSPs are assessed based on two different measurement approaches: 1) synchronous measurements, where the values of the LSPs of the considered links are estimated from the same measurement run; and 2) repeated measurements, where the values of the LSPs of the considered links are estimated from different measurement runs. Repeated measurements are attractive because they are simpler and less expensive. In this paper, we address the following question: Can repeated measurements be used instead of synchronous measurements in order to estimate the LSPs' cross-correlation properties of different links? Based on analysis of wideband synchronous and repeated multilink measurements in a suburban microcell environment at 2.6 GHz, we found that: 1) the mean values of the cross-correlation coefficients are preserved with repeated measurements, and 2) the estimates of the cross-correlation coefficients from repeated measurements are less spread around the mean value than those from synchronous measurements. These findings are explained based on detailed investigation of specific measured cases and further supported by results obtained from Monte Carlo simulations.