Numerical Study of Statistical Properties for Semiconductor Laser Chaos by Exploring the Injection Parameter Space

The statistical properties of semiconductor laser chaos are of great importance for practical applications. By exploring the injection parameter space of an optically injected semiconductor laser, the associated probability density functions (PDFs) are numerically investigated for generation of chaotic waveforms with rogue waves (RWs) or Gaussian distributions. First, generation of RWs are demonstrated by considering chaotic waveforms with long-tailed PDFs. The dynamical behaviors for RW generations are systematically investigated by exploring the entire injection parameter space, where a pattern is identified for parameter regions of nonzero RWs. The occurrence of RWs is also investigated for different injection parameters by evaluating the effects of laser bias current, spontaneous emission noise, and detection bandwidth. Alternatively, Gaussian PDFs are generated by bandwidth-limited detection of chaotic waveforms. The bandwidth-dependent statistical properties are investigated through tuning the injection parameters for different detection bandwidths. Generation of Gaussian PDFs are demonstrated for a large region of injection parameters and for a detection bandwidth as comparable to the laser relaxation resonance frequency. This numerical study reveals the importance of injection parameters for tuning the statistics associated with chaotic lasers, and confirms the feasibility of generating RWs and Gaussian PDFs on demand for potential applications in various fields.