A Novel Robust Technique to compensate I/Q imperfections In Quadrature Modulator using LMS Algorithm

In the modern era, the increasing demand for high data rate transmission enquires for wireless communication devices with large spectral bandwidth. Large spectral bandwidth increases the level of fault in the transmitter block. Quadrature modulator is one of the sensitive blocks in the transmitter. Having large spectral bandwidth makes the transmitter's quadrature modulators more prone to high level of physical and electrical faults. Physical faults in quadrature modulators, known as I/Q imperfections, will cause incorrect demodulation on the receiving end resulting in high bit error rate. In this paper, we propose a new 2-feedbacks compensation scheme based on Least Mean Square algorithm, its related channel modeling, a novel 1-bit online least mean square algorithm and finally a theoretical modulation/compensation method based on 1-bit modulation. To evaluate our proposed scheme and algorithm, we calculated normalized mean square error, image rejection ratio and convergence speed of our proposed algorithm for 128 iterations, gain and phase imbalances. While the normalized mean square error of adaptive algorithms used in other studies was -68dB, normalized mean square error for our proposed 2-feedbacks method is only -79.8dB. Moreover, image rejection ratio, signal errors and speed of the convergence shows improvements in our method as well.