Description
There are many undesired attributes caused by imperfect analog components in quadrature receivers. These include distortion due to gain and phase mismatch in the analog quadrature down converter mixers, DC offset due to self-mixing in the down converter, DC injection due to truncation arithmetic, harmonics from the digital clocks into the down converted band, and spectral cross talk due to phase and gain mismatch of the analog filters in the signal paths. The undesired artifacts can be eliminated using digital signal processing techniques that perform DC offset cancelling and single or multiple channel I-Q imbalance correction. All modern radio receivers contain DSP based functions that perform the DCcancelling and I-Q balancing tasks. Many modern communication receivers use analog quadrature down conversion of their radio frequency (RF) signals from an input center frequency or from an intermediate center frequency to baseband. Gain and phase imbalances of the analog mixers or of the analog low pass filters in the down conversion process lead to an undesirable signal interference from the image frequency of the down converted center frequency. High performance receivers must reduce the level of the insufficient rejection of the image frequency band. In this thesis, imbalances between I and Q branches due to the finite tolerance of analog components are modeled separately to account for artifact contributions from the quadrature mixers and for the imbalanced low pass filters. We present and discuss detailed analysis of adaptive interference cancellation based imbalance compensation based on baseband digital signal processing. The Least Mean Square (LMS) and Recursive Least Square (RLS) algorithms are used to implement the adaptive cancelling techniques. These exhibit significantly different convergence profiles and levels of computational complexity.
