Research And Implementation Of Wideband Zero-if Receiver’s IQ Imbalance Calibration Technology

Compared with traditional superheterodyne receiver, zero-IF receiver has a simple circuit structure, higher integration, lower power consumption and cost. It’s more suitable for consumer electronic products and multimode applications, and gains great attention and research in recent years. However, the performance of zero-IF receiver is constrained by the drawbacks stemming from its hardware strucure, especially the IQ imbalance caused by the unmatched devices and circuit PCB layout errors.Currently, zero-IF receiver IQ imbalance calibration studies concentrated in the narrowband signal calibration, which considers the imbalance is unchanged within the receiving bandwidth. While, in wideband applications, the IQ imbalance of receiver is frequency dependent, that is, it varys with frequency within the receiving bandwidth, which makes the imbalance calibration more difficult. With the increasing bandwidth of the receiver, we need to solve the wideband zero-IF receiver’s IQ imbalance calibration.This paper designs and implements a wideband zero-IF receiver’s IQ imbalance calibration scheme. Details are as follows:Firstly, this paper analyses the imbalance characteristic of the handhold software radio platform’s wideband zero-IF receiver by testing. Using the test method, this paper analyses the characteristic of IQ imbalance within the 30~88MHz bandwidth of the receiver, and verifys the IQ imbalance of our wideband receiver is frequency dependent.Secondly, this paper develops the wideband receiver IQ imbalance calibration solutions in the digital domain. Based on the principle of the low order statistic method of blind IQ imbalance compensation, this paper designs a ‘semi-blind’ digital domain imbalance parameter estimation and calibration program for narrowband signal. The final digital domain imbalance calibration program for wideband signal is based on the former narrowband signal’s calibration scheme and FIR filters designed by using interpolation algorithm. The interpolation algorithm and interpolation order is determined by simulation and realizability. At the same time, this paper analyses the performance of wideband calibration program and optimizes the parameters by using simulation method.Thirdly, according to the characteristics of handhold software radio platform’s receiver, this paper completes the implementation of wideband calibration scheme. This paper divides the 30~88MHz receiving bandwidth into several sub-bands, and designs a process for getting the characteristic of each sub-band. Then, this paper designs the calibration FIR filters with variable coefficients, and completes FPGA implementation by using these FIR filters. Our implementation has the characteristics of low complexity and low resource consumption.Finally, the effect of wideband zero-IF receiver’s IQ imbalance calibration scheme is tested and analysed. This paper compares the image rejection ratio(IRR) of before and after calibration within 30 MHz ~ 88 MHz receiving bandwidth. According to the final test results, this calibration scheme can improve the image rejection ratio to more than 43 d B within the bandwidth of the receiver. At the same time, according to the image rejection ratio’s trends before and after calibration, the factors affecting the performance of the calibration scheme are analysed.This paper designs and implements a wideband zero-IF receiver’s IQ imbalance calibration scheme, by testing on the handheld software radio platform, verifys the feasibility of scheme. These enrich the solutions of frequence dependent IQ imbalance.