Abstract
The implementation of advanced multi-level modulation schemes such as quadrature phase-shift keying (QPSK) is crucial to further boost terahertz (THz) communications speed beyond conventional on–off keying. However, carrier phase noise reduction in the THz range remains a critical challenge. This paper presents a low phase-noise photonic source based on a stimulated Brillouin scattering (SBS) optical fiber cavity applied to a 300-GHz-band QPSK wireless communication link. Utilizing this source, a data rate of 15 Gbaud was achieved at the forward-error-correction (FEC) limit with a transmit power of approximately −36 dBm. The SBS-based source demonstrates superior performance over conventional optical frequency comb generator (OFCG)-based sources in terms of phase noise, optical carrier-to-noise ratio, and bit error rate.
Introduction: The Phase Noise Challenge in THz Communications
To achieve wireless data rates exceeding 100 Gbit/s, researchers are targeting the Terahertz (THz) band, specifically the 300 GHz range, which offers vast available bandwidth. While simple on-off keying (OOK) has been the standard, reaching higher spectral efficiency requires multi-level modulation schemes like Quadrature Phase Shift Keying (QPSK).
The primary obstacle for QPSK at these frequencies is phase noise. In both electronic and photonic systems, phase instability scales poorly as frequencies are multiplied to reach the THz regime. High phase noise "smears" the signal constellation, making it difficult for receivers to distinguish between symbols, as illustrated in the theoretical models of BPSK and QPSK constellations.
Methodology: SBS-Based Photonic Source
The proposed solution utilizes a photonic source based on Stimulated Brillouin Scattering (SBS) within an optical fiber cavity. This process leverages the "noise-cleaning" properties of the SBS effect, where a physical interaction between light and acoustic waves creates a narrow-band Stokes wave. This wave acts as a natural filter, significantly narrowing the linewidth of the lasers used.
The system employs two independent continuous-wave (CW) lasers locked to different modes of the SBS cavity. This locking mechanism ensures that the relative frequency difference between the two lasers—which determines the 300 GHz carrier frequency—remains extremely stable. These "cleaned" optical signals are then mixed in a Uni-Traveling-Carrier Photodiode (UTC-PD) to generate the THz electrical signal.
Key Results and Performance Comparison
The SBS-based source was compared against a standard Optical Frequency Comb Generator (OFCG). Key findings include:
- Phase Noise Reduction: The source achieved a phase noise of −90 dBc/Hz at a 10 kHz offset at 300 GHz. Unlike traditional frequency multiplication, the SBS approach does not suffer from quadratic noise scaling with frequency.
- Improved Power Efficiency: The SBS source provided a Carrier-to-Noise (C/N) ratio approximately 8 dB higher than the OFCG source. This efficiency stems from the SBS system concentrating pump energy into the two required optical tones, whereas OFCGs waste energy on unused comb lines.
- High Data Rates: The experimental link successfully transmitted 15 Gbaud (30 Gbit/s) at the FEC-limited condition with a very low transmit power of −36 dBm.
- Sensitivity: At 5 Gbaud, the SBS source reached the FEC limit at −45 dBm transmit power, outperforming the OFCG source which required −42 dBm.
Conclusion and Future Outlook
The application of an SBS-based photonic source represents a significant step toward high-speed, low-noise THz wireless links. While the current transmitter demonstrates exceptional performance, future work will focus on upgrading the receiver with similar low-noise technology to eliminate the current electronic local oscillator bottleneck. Additionally, packaging the fiber cavity to mitigate environmental sensitivities (vibration and temperature) will be essential for practical deployment. The success with QPSK suggests that even higher-order modulations, such as 32 QAM, may be achievable with this technology.
Summary updated