Quantum dot laser noise is an important design consideration wherever optical sources must deliver stable intensity, including high-speed communications and photonic systems used in radar. A laser’s intensity and phase can both fluctuate, but the mechanisms that shape these forms of noise are not identical. In quantum dot (QD) lasers, carrier fluctuations in the ground and excited states can significantly increase relative intensity noise (RIN), while fluctuations in the carrier reservoir have little effect in the model examined here.
This distinction matters to engineers evaluating laser behavior or developing models for low-noise optical sources. Accounting only for spontaneous emission, or including carrier noise from just one energy state, may give an incomplete picture of intensity stability.
Why relative intensity noise matters
RIN describes fluctuations in optical output power relative to the average power, expressed as a noise spectrum. In communication links, excessive intensity noise can impair the transmission of broadband data. For radar-related applications, limiting intensity fluctuations is also important across the operating bandwidth. The relevant noise therefore depends not just on the laser’s average output, but on how fluctuations vary with frequency.
Quantum dot lasers have discrete, atom-like energy levels and can offer properties such as low-threshold operation and narrow spectral linewidth. Their carrier dynamics involve several interacting populations: a carrier reservoir (RS), an excited state (ES), and a ground state (GS). Carriers move from the reservoir into the excited state and then relax into the ground state; thermal processes can also move carriers back to higher levels. Spontaneous recombination and stimulated emission further affect the populations and the optical field.
A rate-equation model can represent these processes and add Langevin noise sources to describe fluctuations. In the study considered here, stimulated emission is modeled from the ground-state transition. Including carrier noise from the different populations reveals effects that a model omitting those sources—or considering only ground-state carrier noise—may not capture.
Quantum dot laser noise: the role of each carrier state
The simulations show that carrier noise from both the GS and ES contributes substantially to RIN. By comparison, carrier noise from the reservoir is negligible in the modeled conditions. This means that the carrier population feeding the quantum dots is not necessarily the most important noise source once fluctuations within the dot energy states are considered.
The two influential states affect the spectrum differently. Ground-state carrier noise raises RIN across the frequency range examined. Excited-state carrier noise mainly increases RIN below the laser’s resonance frequency. The RIN spectrum is relatively flat at low frequencies, develops a peak near resonance, and decreases at higher frequencies. Increasing bias current reduces the overall RIN in the reported simulations, with the resonance peak also becoming smaller and shifting to a higher frequency.
At 1 MHz and 1.5 times the threshold current, including carrier noise raised the calculated RIN from −144 dB/Hz to −134 dB/Hz. The result illustrates why carrier noise cannot automatically be treated as a minor correction in QD laser intensity-noise analysis. It is a result from the modeled laser and operating conditions, not a universal performance value for all quantum dot devices.
Energy-level spacing and relaxation affect RIN
The separation between the ground and excited states influences how strongly excited-state carrier fluctuations affect the ground-state laser output. In simulations at fixed photon numbers, a larger GS–ES energy separation reduced RIN. For a photon number of 2 × 105, increasing the separation from 0.05 eV to 0.16 eV reduced the calculated RIN by about 6 dB.
The practical design implication is that weaker coupling between the GS and ES can favor lower intensity noise by reducing the excited state’s contribution. The study also identifies longer ES-to-GS relaxation time as a factor that reduces RIN, particularly for relaxation times below 10 ps in the modeled case. Beyond that range, the calculated RIN was comparatively insensitive to further increases. These relationships should be interpreted within the assumptions of the model; device design involves multiple interacting parameters.
Photon lifetime is another relevant consideration. The simulations indicate that a longer photon lifetime is expected to reduce RIN, as also seen in quantum well lasers. Possible ways to increase photon lifetime include increasing cavity length or facet reflectivity, or reducing internal loss. However, assessing such changes accurately requires a model that also accounts for effects such as spontaneous-emission-factor changes and spatial hole burning.
Frequency noise is a different design question
Carrier noise has a much smaller effect on frequency noise (FN) than on RIN in the reported analysis. It slightly enhances the FN peak near resonance, but contributes little elsewhere. The study therefore indicates that distributed carrier noise in the GS, ES, and reservoir can generally be neglected when investigating frequency noise and spectral linewidth under its modeling assumptions.
This difference is useful when choosing the level of detail in a laser-noise model. An analysis focused on intensity stability should include carrier noise from both the ground and excited states. An analysis focused on spectral linewidth may be less sensitive to those distributed carrier-noise terms, although other phase-noise mechanisms remain relevant.
Implications for photonic system design
For engineers, system integrators, and technical decision-makers assessing QD lasers for communications, optical frequency combs, or radar-related systems, the main lesson is to match the noise model to the performance question. When RIN is critical:
- Include ground- and excited-state carrier noise rather than relying only on spontaneous-emission noise.
- Evaluate reservoir noise separately: it was negligible in the modeled case, while GS and ES contributions were significant.
- Consider energy-level separation and relaxation dynamics as factors affecting the excited-state contribution to intensity noise.
- Distinguish RIN from frequency noise: carrier fluctuations that strongly affect intensity may have little influence on spectral linewidth.
These findings apply to a model of a QD laser emitting on the ground-state transition, with stated simplifications including the exclusion of dot-size dispersion and excited-state stimulated emission. They provide a useful framework for understanding noise mechanisms, but should not be treated as a substitute for device-specific characterization.
Conclusion
Ground- and excited-state carrier fluctuations can materially increase relative intensity noise in quantum dot lasers, whereas reservoir carrier noise has little influence in the modeled system. Larger ground-to-excited-state energy separation can reduce the excited-state contribution, and carrier noise has comparatively little impact on frequency noise. For practical design and evaluation, engineers should use a model that reflects the noise metric of interest and the carrier dynamics relevant to the specific laser and application.