Rebuttal to “Photonic Microwave Generation With Frequency Octupling Based on a DP-QPSK Modulator”

Abstract

In the title paper, Gao et al. have experimentally demonstrated a scheme for filter-less photonic millimetre-wave (mm-wave) generation using a dual polarization quadrature phase shift keying modulator (DP-QPSK). For frequency octo-tupling generation, the X-QPSK modulator is biased such that all the harmonics are suppressed except those that are order of ${4l}$ , where ${l}$ is an integer and combined with the output of the non-modulated Y-QPSK modulator using a polarization beam combiner. Precise setting of the polarizer angle suppresses the carrier, leaving the ±4th sidebands as the strongest harmonics, which is the innovative part of their design. However, a demerit of the design is that, following any change in modulation index, a precise readjustment of the polarizer angle is required in order to suppress the unwanted harmonics (e.g. carrier) without undue suppression of the wanted harmonics. The goal of this comment is to show that, for an appropriate modulation of the Y-QPSK modulator with the design presented by Gao for a fixed polarizer angle, all the harmonics can be suppressed except those that are order of $4{l}$ , where ${l}$ is an odd integer. Hence, a filter-less frequency octo-tupling can be generated without restricting its operation by the careful adjustment of the RF drive signal; rather the circuit can function for a broad range of modulation index ( $m\sim 2.5 \to 7.25)$ . Moreover, the unwanted harmonics suppressed by design mandate 6.66 dB optical and 11 dB electrical advantages over the design presented by Gao and the precise setting of the polarizer angle can be avoided.