基 于 科 斯 塔 斯 环 的 零 差 相 干 光 接 收 技 术（特 邀）

Objective Space laser communication has emerged as a transformative technology addressing the increasing demands of next-generation space exploration and science missions. Conventional microwave communication systems face fundamental constraints: limited capacity, intense competition for spectrum resources, and exponential increases in size, weight, and power consumption with increasing transmission distance and data rates. These limitations significantly impede deep^-space exploration, Earth observation constellations, and satellite networks that require high^-data^-rate links across thousands to tens of thousands of kilometers without repeaters. Coherent optical communication systems, particularly those utilizing homodyne detection, present a compelling solution in this context. Leveraging the ultra^-short wavelength and high frequency of laser light, these systems deliver substantially higher bandwidth potential and superior receiver sensitivity compared to direct detection or conventional microwave^-based systems. This sensitivity is essential for overcoming the substantial link losses inherent in space^-based free^-space optical links. Additionally, coherent detection provides exceptional background light rejection, a crucial capability in the space environment saturated with solar radiation. However, achieving the necessary phase and frequency synchronization between the signal light and local oscillator light for homodyne detection presents significant technical challenges, particularly under dynamic space conditions (e.g., platform vibrations, large Doppler shifts). The development of robust, high-sensitivity, and power^-efficient homodyne receivers therefore represents a cornerstone for realizing the full potential of high-capacity space laser communications. Methods This research successfully developed and validated a high-performance 10 Gbit/s binary phase^-shift keying (BPSK) optical homodyne receiver prototype based on the Costas loop architecture. The prototype receiver includes several key technical achievements. 1) Delayed XOR phase-frequency detector (PFD). A PFD architecture (Fig. 2) is implemented using a time^-delayed XOR gate preceded by limiting amplifiers. Theoretical analysis [Eq. (6)] and numerical simulations (Fig. 3 and Fig. 4) optimized the delay, balancing a -6.25 to 6.25 GHz frequency capture range with high phase discrimination sensitivity while eliminating BPSK modulation artifacts. 2) Auxiliary^-controlled composite optical phase^-locked loop (OPLL). A dual^-loop architecture (Fig. 5) including a fast loop and a slow loop has been proposed and realized. The fast loop with MHz bandwidth uses an acousto-optic frequency shifter (AOFS) -50 to 50 MHz range controlled by an analog active loop filter (LF) (Fig. 6) with lead^-lag compensation for loop stability. The slow loop with kHz bandwidth uses the piezoelectric transducer (PZT) -30 to 30 GHz range of a narrow^-linewidth fiber laser controlled via digital LF implemented with digital signal processor (DSP) and digital to analog converter (DAC). In addition, an auxiliary controller monitors LF output and filtered PFD error. These enabled an autonomous frequency scanning, lock detection, and proactive Doppler compensation by coordinating the fast/slow loops. 3) Prototype implementation & testing. A 10 Gbit/s BPSK receiver (Fig. 7) is built using a 90° optical hybrid (within the error range of -5° to 5), balanced photodetectors (10 GHz bandwidth), home^-made PFD/LF boards, AOFS, and fiber laser based local oscillator laser. The performance verification is completed through the following four tests. 1) Autonomous locking tests (Fig. 8). Measuring acquisition time under ±12 GHz initial frequency offset conditions of -12 to 12 GHz. 2) Doppler tolerance tests (Fig. 10). Apply a sine frequency scan of -8 to 8 GHz at a rate of 600 MHz/s. 3) Sensitivity tests (Fig. 12). Measuring BER vs. received power using a PRBS 2^23-1 pattern. 4) Phase noise analysis. Integrating residual phase error from PFD output spectra (Fig. 13). Results and Discussions The prototype demonstrated exceptional performance exceeding typical space communication requirements. 1) Autonomous acquisition. Lock acquisition completed in ≤7.3 s under extreme initial frequency offset conditions of -12 to 12 GHz (Fig. 9), demonstrating robust link initialization essential for operational systems. The auxiliary controller’s scanning and detection capability proved instrumental in this achievement. 2) Doppler tolerance. The receiver compensated simulated Doppler shifts of -8 to 8 GHz magnitude at 600 MHz/s slew rates [Fig. 11(b)], substantially exceeding typical maximum requirements. Notably, disabling the auxiliary control resulted in immediate loss of lock [Fig. 11(c)], confirming its essential role in extending the effective tracking range beyond the AOFS’s inherent ±50 MHz limit. 3) Phase^-locking fidelity. Measured loop bandwidth reached ~500 kHz (Fig. 13), enabling effective suppression of broadband phase noise. Integrated residual phase error is 3.6°, substantially below the ~10° threshold for low^-penalty BPSK demodulation, confirming loop stability under closed^-loop operation. 4) Sensitivity with discrete implementation. Achieved a sensitivity of -41.3 dBm at BER is 1×10^-7 (Fig. 14), exceeding the -40 dBm benchmark for 10 Gbit/s space transmission links. The sensitivity at BER (1×10^-9) is -39.5 dBm, with clear eye opening observed (Fig. 14, inset). This approaches theoretical homodyne limits and surpasses reported integrated alternatives at similar speeds. 5) System robustness. The composite loop design effectively integrated the wide tuning range of the PZT^-based slow loop (±30 GHz) with the rapid response of the AOFS^-based loop (10 MHz bandwidth), resolving the inherent tradeoff in single-loop designs. The auxiliary control’s sophisticated management of frequency drift prevention ensured sustained lock under dynamic conditions. Conclusions This research successfully designed, implemented, and rigorously validated a high-performance 10 Gbit/s BPSK optical homodyne receiver prototype based on the Costas loop architecture. The developed prototype represents a significant milestone in space coherent receiver technology utilizing discrete components. The results demonstrate the technical maturity and space^-readiness of the Costas loop homodyne architecture, fulfilling the critical requirements for high-sensitivity, high^-data^-rate space laser communications. The prototype’s performance confirms its advantages in laser phase and frequency noise tolerance, acquisition robustness, and Doppler resilience compared to alternative coherent receiver configurations. Future improvements can focus on size reduction, power consumption optimization, support for higher^-order modulation formats, and system integration. These technical achievements establish a fundamental platform for developing next-generation integrated coherent terminals essential for high-capacity inter^-satellite links and provide significant reference value for the engineering implementation of coherent receivers in space optical communications.