Research On Reconfigurable RF Measurement Systems Based On Parallel Modulation And Demodulation

With the continuous development of modern electronic information technology,various radio frequency(RF)measurement instruments are moving towards modularization,generalization,and miniaturization in the direction of updating iteration.Traditional RF measurement instruments mainly include the vector network analyzer,spectrum analyzer,and signal source for applications such as radar,communications,and integrated circuit testing.Despite the excellent performance of RF measurement instruments,the shortcomings of solidified functionality,specialization,and poor expandability make them unable to meet the functional requirements of modern multitask joint testing.Additionally,due to the increasingly complex transient characteristics of electronic signals during testing,higher demands are placed on the instantaneous bandwidth(IBW)metrics of signal sources,spectrum analyzers,and other instruments.However,it is difficult to expand the IBW metrics of measurement instruments under the traditional architecture restricted by the corresponding metrics of existing functional chips,such as the sampling rate of analog-to-digital converter(ADC)and the bandwidth of quadrature demodulation chips.Therefore,reconfigurable technology has been gradually introduced into the test field to fulfill various test tasks and design corresponding test systems.Based on the inherent characteristics of RF measurement instruments with transceiver channels as the hardware core,the dissertation proposes a novel reconfigurable RF measurement system around the goals of functional reconfiguration and IBW expandability.Also,the error analysis and modeling of this measurement system are carried out to solve the problems,including the compensation of singlechannel quadrature demodulation in-phase quadrature(IQ)mismatch error,the joint estimation of single-channel quadrature modulation IQ mismatch and DC-offsets errors,as well as the estimation of multi-channel quadrature modulation and demodulation IQ mismatch errors.The specific research contents are as follows:(1)A novel universal RF measurement system is constructed based on reconfigurable technology and modularization ideas.This system takes the multichannel quadrature modulation and demodulation module as the hardware core;on the one hand,through the flexible regulation of the matrix switch module,the specific function module is selected to realize the reconfiguration of the system function.On the other hand,by configuring the frequency value of each channel’s digital complex mixing and analog quadrature mixing oscillator,the IBW index of the system can be expanded to break through the limitation of the relevant index of the key chip.Meanwhile,the errors existing in this system are analyzed and decomposed into singlechannel quadrature demodulation IQ mismatch error,single-channel quadrature modulation IQ mismatch error,DC-offsets,and multi-channel quadrature modulation and demodulation IQ mismatch errors,which provide the preconditions and theoretical support for the subsequent estimation of the errors and compensation.(2)The compensation method of IQ mismatch error for single-channel quadrature demodulation is investigated.By theoretically analyzing the specific presentation of IQ mismatch error during quadrature demodulation,a compensation method relying on quadrature signal reconstruction is proposed.This method utilizes the Weaver architecture to implement the orthogonal transformation of the demodulated I-branch signals,thus avoiding introducing the Hilbert Transform(HT)filter to ensure that the method is still effective in broadband applications.Compared with existing studies,this method does not require previous information,such as second-order statistical properties and additional hardware-assisted circuits,and the method’s complexity is significantly reduced because it involves no iterative operations.The experimental results show that this method has favorable compensation ability in application scenarios where frequency-independent(FI)and frequency-dependent(FD)mismatch coexist.(3)Research on single-channel quadrature modulation IQ mismatch error and DCoffsets estimation method is carried out,which proposes an estimation method based on the self-closed loop of modulation and demodulation channels.By setting the local oscillator signals to have a fixed frequency difference,the joint estimation is realized by using specific guide sequences.On the one hand,this estimation method avoids the introduction of extra measurement instruments for assistance.On the other hand,the estimation accuracy is proved to be independent of the interference caused by the selfclosed loop link and demodulation channel through strict theoretical derivation.Numerical results validate the effectiveness of the joint estimation method and demonstrate excellent suppression of the local oscillation leakage introduced by the DCoffsets.Moreover,the estimation accuracy of the IQ mismatch error is up to the thousandth order.(4)Based on the above research,a multi-channel error estimation method is proposed.Under the self-closed loop condition of single-channel quadrature modulation demodulation,the expressions for the amplitude-phase information at specific frequency points are derived by further analyzing the frequency domain characteristics of the demodulated complex baseband signal.By utilizing the fixed frequency difference between the local oscillation signals and multiple pilot sequences,the amplitude-phase effects introduced by the matrix switch module are eliminated so that the expression is converted to an equation relationship between the pilot sequences,the quadrature modulation IQ mismatch error,the quadrature demodulation IQ mismatch error,and the amplitude-phase information of specific frequency points.Based on the equation relationship,the error parameter matrix is obtained after traversing all the modulation and demodulation channels using the matrix switch module,thereby estimating the IQ mismatch error in the proposed reconfigurable measurement system.Besides,the proposed method can be extended to the joint estimation at arbitrary channels for the mismatch error.Simulation results and physical tests jointly show that the proposed method can accurately estimate multiple IQ mismatch error values without the aid of external instruments,and the estimation accuracy can still be up to one-thousandth level under the influence of the feedback loop.