A spectral correction algorithm for I-Q channel imbalance problem

A novel spectral correction algorithm for an I-Q channel imbalance problem in a multiuser narrow band system is presented. If the receiver has imperfect I-Q channel oscillators, which are not exactly 90 degree phase shifted and/or are not of the same gains, interference in the image channel used by another user may substantially degrade system performance. By estimating an I-Q channel imbalance parameter that we denote as /spl gamma/, which is a function of the phase error and amplitude offset between the inphase and quadrature channels, the distorted spectrum of the narrow-band band-pass signal at the output of the quadrature channels can be corrected. We invented a recursive algorithm that does not depend on the received signal power to estimate /spl gamma/ in a real-time manner. The algorithm is a blind algorithm in the sense that we do not need to send a training sequence in advance. This paper starts with a brief complex envelope representation for narrow-band band-pass signals. We then give analytical derivations for the spectral correction algorithm along with some numerical examples.