Xinping Huang

Gain/phase imbalance and DC offset compensation in quadrature modulators

This paper describes a technique to estimate gain and phase imbalances and DC offsets in a quadrature modulator and to pre-compensate the in-phase and quadrature signals, so that image suppression is maximized and LO leakage is minimized at the modulator output. To validate the technique, an experiment with two 2.5 GHz direct modulators was conducted, and the results show that both the image suppression and LO leakage can be improved by at least 20 dB.

I/Q-channel regeneration in 5-port junction based direct receivers

This paper presents a new I/Q regeneration technique for direct receivers which employ quadrature modulation. The technique is based on the property that the I/Q components of the received signal are uncorrelated, and is applicable to any quadrature modulation. A K-band 5-port junction based direct receiver prototype has been developed which supports a data rate of 89.472 Mbps with QPSK modulation. The bit error rate performance has been simulated and measured, and the results are presented.

Efficient Transmitter Self-Calibration and Amplifier Linearization Techniques

This paper shows how RF circuit imperfections such as imbalances in the quadrature vector modulator and nonlinearities in the power amplifier can be effectively mitigated digitally at baseband. Two patented techniques, one for vector modulator calibration, and the other for power amplifier linearization, are described. Simulation and experimental results are included to demonstrate the performance improvement.

Blind Self-Calibration Technique for I/Q Imbalances and DC-Offsets

The gain/phase imbalances and DC-offsets in the in-phase and quadrature branches in direct conversion transmitters (DCTs) are main sources of errors, especially at microwave frequencies and above. These impairments distort the spectrum of the transmitted signal and degrade the communications performance. In this paper, a blind technique based on higher order statistics is proposed to estimate the calibration parameters in order to compensate for these impairments. By minimizing a cost function composed of second and fourth moments of the modulator output signal, the DC-offsets, gain imbalance, and phase imbalance can be successively calibrated. The convergence of the proposed technique is also analyzed. This technique is blind in the sense that it does not require any prior knowledge about the DCT input signal, and is applicable to many communications systems including QAM, OFDM, and SC-FDMA. Computer simulations and experiments at 20 GHz demonstrate that the proposed technique can significantly improve the DCT performance in terms of local oscillator leakage and image suppressions.

Performance of a type-based digital predistorter for solid-state power amplifier linearization

In a wireless communication system, a predistorter is often employed to alleviate nonlinear distortions due to operating a power amplifier near saturation, thereby improving the system performance and reducing the interference to adjacent channels. In this paper, we investigate the performance of a new digital predistorter. A 16-QAM communication system is used as an example to show that the predistorter indeed eliminates the nonlinear distortion, and can greatly improve the performance of the communication system.

Ka-Band Multi-port Power Amplifier Calibration Experiment and Results

A multi-port power amplifier (MPA) is a multi-input multi-output system that is capable of amplifying multiple input signals simultaneously by a set of shared power amplifiers without mutual interference. In a practical MPA, the component imperfections reduce the port isolation, which introduces leakage or cross-port interference. A patent-pending type-based calibration technique is studied in this paper. It exploits the uniqueness of the statistics for a given communication signal to estimate a calibration matrix that minimizes the effects of the MPA impairments and suppresses the cross-port interference. Experimentation with a 4-port 20 GHz prototype MPA subsystem is conducted to evaluate the type-based calibration technique. Results show that it can accurately estimate the calibration matrix to improve the MPA performance significantly.

Analytical Performance of M-PSK Communications Systems in the Presence of Gain/Phase Imbalances and DC-Offsets

This letter addresses the performance of M-PSK communications systems in the presence of gain/phase imbalances and DC-offsets in in-phase and quadrature channels. A model is established to represent the gain/phase imbalances and DC-offsets in both the transmitter and the receiver, and two closed-form formulas are derived to calculate the symbol error rate (SER) of the M-PSK communications systems in the presence of these impairments. Computer simulations are conducted to validate the derivation, and the results show that the SER calculated from the formulas closely matches the simulation results. The formulas provide an accurate guideline for communications system architects/engineers to evaluate the impacts of various imbalances and offsets on the system performance and to determine the required component specifications.

Type-Based Group Delay Equalization Technique

This paper presents a patented type-based group- delay equalization technique to compensate for in-band group- delay distortion typically existing in analog/RF filter circuits. The underlining principle is that for a given modulation scheme and pulse-shaping function, the modulated signal has a unique statistical distribution, and that when the modulated signal passes through a circuit, any in-band group-delay distortion in the circuit distorts its output statistical distribution. The technique employs an equalization filter to minimize the in-band group-delay distortion, with the filter coefficients derived from a measure of the distortion in the output statistical distribution using a weighted nonlinear least square algorithm. It requires simple analog and digital hardware and firmware to implement, and its implementation is inherently adaptive, capable of tracking and compensating for any variation in the group-delay distortion characteristic due to component aging and temperature variation. Computer simulations have been performed to show that an accurate equalization filter can be obtained to effectively compensate for the group-delay distortion, achieving significant performance improvements.

A novel type-based vector modulator self-calibration technique

In this paper, we present a patent-pending type-based technique to compensate for gain/phase imbalances and DC offsets in vector modulators. Its underlining principle is that, given a modulation scheme and pulse-shaping and/or channel filter, the modulated signal has a unique statistical distribution, and any impairment in the vector modulator distorts its output statistical distribution. The difference between the ideal and the output statistics contains the impairment information, and is used to derive calibration coefficients. Computer simulations have shown that the type-based technique can effectively compensate for the gain/phase imbalances and DC offsets in vector modulators, and significantly improves the performance.

Experimental results of a type-based predistorter for SSPA linearization

In a wireless communication system, operation of a power amplifier near saturation is often desirable to fully utilize its capacity. Such an operation results in nonlinear distortions, thereby degrading performance of the communication system and introducing interference to adjacent channels. In this paper, we present experimental results using a novel, statistical technique to linearize the power amplifier response. The technique does not assume any parametric modeling of the power amplifier, and it is simple and easy to implement digitally at baseband. The results show that the technique reduces the effects of the nonlinearity significantly.