Adopting Ordered Weighted Averaging Approach in Multitaper Temporal Channel Estimates to Enhance Channel Capacity

This letter proposes to collect the most of temporal channel information from the frequency domain by exploiting the diversity offered from the fusion of multiple windows, commonly known as tapers. For this, frequency-domain measurements are filtered out by using the selected windows, weighted, and then combined prior to the application of the inverse Fourier transform, thereby adopting a dual temporal concept of multitaper spectrum estimate technique. We have opted for multitaper principle due to the previous findings with the spectrum estimate technique that, with a unique window, more or less a significant amount of the original data is discarded, leading to the increase of the estimates variance, and causing a leakage of energy across frequencies. In this letter, the windows weights are selected by adopting ordered weighted averaging method, which involves the solution of a constrained nonlinear optimization problem, such as this one, with a given orness degree as the constraint, and the maximum entropy as the objective function. This concept has been successfully applied to our multiple-input-multiple-output underground channel measurements carried out in a mine in Northern Canada using channel sounding technique in the frequency domain.