Impact of correlation matrix estimation accuracy on the computation of stationarity intervals

The correlation matrix distance (CMD) is widely used to measure the stationarity of MIMO channels. The CMD measures the amount of change in the spatial structure of the MIMO channels, which can be estimated for the transmit (Tx) or receive (Rx) side but also for the full channel. Stationarity intervals can then be defined based on the CMD. In this paper, we analyze the influence of the correlation matrix estimation on the computation of stationarity intervals. More specifically, we emphasize on the quantity of data required to estimate accurately correlation matrices. Some examples are given, based on measured vehicle-to-vehicle (V2V) data.