We present a simple alternative method for the compensation of quadrature imbalance in optical quadrature phase-shift-keying (QPSK) coherent systems. The method is based on the determination and the compensation of the phase mismatch by the introduction of a relevant signal-to-noise ratio metric. The principle is validated numerically and the algorithm is validated experimentally through bit-error-rate (BER) and error vector magnitude (EVM) measurements. A 20 Gb/s optical QPSK experiment reveals a good agreement of the proposed method with the Gram-Schmidt orthogonalization procedure (GSOP). Moreover, the robustness of both methods was verified with up to 30° phase misalignment by comparing the signal after phase imbalance compensation to that without compensation. A 10% reduction of EVM is achieved with our method for a high phase misalignment of 30 . Keywords: in-phase/quadrature imbalance, digital signal processing, fiber optic communications, coherent communications.
INTRODUCTIONOver the last decade, there has been a renewed interest in coherent optical communication systems because of their improved receiver sensitivity and spectral efficiency, and ability to mitigate transmission impairments in the digital domain [1]. In particular, the quadrature phase-shift-keying (QPSK) format has been the object of intensive investigations and 100 Gb/s transmission systems using this format are now commercially available. Ideally, in a QPSK system, the in-phase and quadrature components of the optical field should be orthogonal to each other. However, hardware implementation imperfections such as incorrect bias-points settings in the modulator, imperfect splitting ratio of couplers, photodiodes responsivity mismatch and misadjustment of polarization splitters in the optical coherent receiver can create amplitude and phase imbalance, known as quadrature imbalance (or IQ imbalance), which destroys the orthogonality of the received signal [2]. Since digital signal processing (DSP) circuits are becoming increasingly faster, providing simple and efficient compensation of linear and, possibly, non-linear impairments, it is important to assess their potential for the compensation of this detrimental loss of orthogonality in the receiver. It has already been demonstrated that IQ imbalance in coherent QPSK systems can be corrected by applying different methods such as the Gram-Schmidt orthogonalization procedure (GSOP) [3], the ellipse correction method (EC) [4] or an IQ compensation based on the constant modulus algorithm [5].In this paper, we propose an alternative method for IQ imbalance compensation based on the definition and estimation of a suitable signal-to-noise ratio (SNR) metric for the detected signal. This approach, called maximum SNR estimation method (MSEM), provides an attractive alternative to existing algorithms thanks to its reduced complexity. Indeed, the proposed method requires simple mathematical functions such as exponentiation or division, whereas the GSOP method requires other more complex mathematical ...