Least mean square as error correction method of multi-port reflectometer for microwave imaging of human head

This article present an error correction of a wideband multi-port reflectometer for microwave imaging application of the human head using Least Means Square (LMS) technique. The configuration of the wideband reflectometer is formed by passive circuits, which are four couplers and two power dividers. The analysis of the proposed error correction is done through the mathematical equation calculation based on the LMS technique, which accomplished via the use of the Agilent´s Advance Design Systems (ADS) software. Through the application of this error correction procedure, any imperfect characteristics of the wideband reflectometer have been removed. A number of loads have been used as the Device Under Tests (DUTs). The corrected reflection coefficients are comparable with the theoretical values. The good performance of the corrected reflection coefficients has been achieved when the LMS is applied.