In TDD systems wireless channel reciprocity is not preserved due to the different front-end components at the transmitter and the receiver. Anyhow, the application of channel estimation procedures and adaptive algorithms is possible by using relative calibration methods, which are based on so-called total least squares (TLS) algorithms. These algorithms estimate the influence of the front-ends, which are assumed to be time-independent filter matrices, and allow to estimate the overall channel. This project deals with the applicability of TLS methods in this field of interest.

This project starts with a literature research, especially concerning total least square methods and its application in calibration and estimation of channels. Afterwards, an existing program has to be modified according to the TLS methods in a MATLAB environment. A performance comparison between the uncalibrated and calibrated system in terms of bit and frame error rates (BER/FER) concludes this project.