Cross-quadrature interference canceller-XQIC: A fresh view of the asymmetric equalizer

This paper develops an alternate configuration of the familiar complex-valued baseband linear adaptive equalizer which endows it with greater degrees of freedom, allowing it to equalize more general I/Q imbalances suffered by wireless signals in modern transceiver designs, in addition to equalizing the difficult propagation channel characteristics. As most spectrally-efficient wireless signals employ quadrature-amplitude modulation (QAM), it is natural to treat the data symbols as complex-valued quantities. However, some of the distortions, especially I/Q delay mismatches and frequency-dependent imbalances, do not fit the complex-valued model whereby the I and Q components are treated as the real and imaginary parts. Such complex-valued DSP blocks obviously cannot compensate for these I/Q mismatches. In this paper, the complex-valued adaptive equalizer is re-architected in a more general vector form [4] which allows it to compensate for such I/Q imbalances. Although not a new concept in itself, its development is outlined as a natural extension of the conventional complex-valued, blind (decision-directed) adaptive equalizer to a vector-valued equalizer without an increase in complexity. It is presented as a candidate for blind IQ signal separation [1] where the original I and Q components are dominant. The extension is further related to the adaptive cross-polarization interference canceller (XPIC), inspiring the name XQIC for this equalizer architecture. The developments are supported by simulation and laboratory test results on a microwave radio link employing a broadband wireless modem.

[1]  Mikko Valkama,et al.  Blind I/Q Signal Separation-Based Solutions for Receiver Signal Processing , 2005, EURASIP J. Adv. Signal Process..

[2]  Unto K. Laine,et al.  Splitting the unit delay [FIR/all pass filters design] , 1996, IEEE Signal Process. Mag..

[3]  Hikmet Sari,et al.  Asymmetric baseband equalization , 1988, IEEE Trans. Commun..

[4]  Inbar Fijalkow,et al.  Fractionally spaced equalizers , 1996, IEEE Signal Process. Mag..