A novel adaptive mismatch cancellation system for quadrature IF radio receivers

This paper investigates and resolves in-phase/quadrature phase (I/Q) imbalances between the input paths of quadrature IF receivers. These mismatches along the paths result in the image interference aliasing into the desired signal band, thus reducing the dynamic range and degrading the performance of the receivers. I/Q errors occur because of gain and phase imbalances between quadrature mixers. They are also caused by capacitor mismatches in analog-to-digital converters (A/Ds), which are designed to be identical for each input path. This paper presents a novel and feasible digital signal processing (DSP) solution for the I/Q mismatch problems. The system includes a novel complex least mean square algorithm and a modified adaptive noise canceler (signal separator) to separate the desired signal and the image noise caused by the mismatch. The noise canceler can also solve the signal leakage problem, which is that the noise reference includes signal components. This system was implemented in a Xilinx FPGA and an Analog Devices DSP chip. It was tested with a complex intermediate frequency receiver, which includes an analog front end and a complex sigma-delta modulator. Both simulation results and test results show a dramatic attenuation of the image noise. Extending applications of the system to N-path systems further indicates the robustness and feasibility of this novel adaptive mismatch cancellation system.