Objective  In lidar remote ranging, the lidar receiver needs to extract weak signals from noisy received signals. The detection ability of the lidar receiver for weak signals is an important part of ensuring the detection distance and accuracy of the lidar system. Improving the signal-to-noise ratio (SNR) of the demodulated signal can effectively improve the detection distance and accuracy of lidar. Under the premise of a certain laser transmission power, it is necessary to improve the SNR during the reception process. Traditional heterodyne lidar receivers typically down-convert the lidar echo signal to the baseband through two down-conversions. The first down-conversion directly mixes the echo light signal with the local oscillator light, which introduces additional mirror frequency noise. Due to the mirror frequency noise being the same as the useful signal frequency after demodulation, the noise cannot be eliminated through a filter after demodulation, resulting in a deterioration of the SNR of the demodulated signal. In contrast, using orthogonal demodulation can suppress the elimination of mirror frequency noise through the phase relationship between the two branches during the demodulation process, thereby improving the SNR of the demodulated signal. In practical applications, affected by the non-ideal state of the device, the two branches in the quadrature demodulation structure may not be able to achieve a fully balanced state. Therefore, the algorithm compensation for the orthogonality of the two branch signals is also worth in-depth study. The use of appropriate mirror frequency suppression down-conversion structures and the use of relevant algorithms for imbalance compensation is expected to improve the detection performance of lidar systems.

  Methods  Based on the principle of receiving and demodulating echo signals in a lidar system, and referring to structures such as Hartley and Weaver, an orthogonal down-conversion structure for optical signals in the receiver was constructed. At the same time, based on the statistical characteristics of the orthogonal signal, a compensation algorithm is used to compensate for the error of the demodulated quasi orthogonal signal to ensure that the mirror frequency noise is completely eliminated. Subsequently, the demodulated intermediate frequency signal is subjected to secondary down-conversion to obtain the baseband signal. Finally, by computing the SNR of two schemes can effectively determine the performance of suppression of the mirror frequency.

  Results and Discussions   Simulation analysis and experimental verification show that the SNR using the biorthogonal demodulation scheme is superior to traditional heterodyne demodulation schemes. In addition, when there is amplitude or phase imbalance in the orthogonal structure, the use of correlation compensation algorithms can effectively eliminate the additional noise interference generated by this imbalance (Fig.9), further improving the SNR of the modulated signal. The results show that compared to traditional heterodyne lidar receivers, the SNR of the demodulated signal processed by this scheme is improved by about 3 dB (Fig.12).

  Conclusions  This study is based on the background of coherent lidar remote ranging, and improves the down-conversion structure of the lidar receiver. By adopting an optoelectronic dual orthogonal down-conversion method, it effectively suppresses mirror frequency noise, improves the SNR of the demodulated signal, and ultimately improves the lidar detection distance and detection accuracy. Due to the fact that the I/Q imbalance compensation algorithm used in this scheme is based on the statistical characteristics of the signal, which relies on the length of the sampled signal and the integrity of its period, its compensation performance is not stable for different sampling signals. Therefore, researching more advanced compensation algorithms can achieve better mirror frequency suppression effect. Finally, this scheme only suppresses mirror frequency interference in frequency mixing, and in reality, there is also frequency related noise. The suppression of frequency related noise can further improve the effectiveness of signal noise control.