Applied Sciences

We propose and numerically investigate a four-dimensional (4-D) multiplexed polarization division multiplexing twin single-sideband (PDM-Twin-SSB) PAM-4 transmission scheme enabled by a Stokes vector Kramers–Kronig (SV-KK) receiver, targeting high-capacity short-reach optical interconnects with low-cost implementation. At the transmitter, the PDM-Twin-SSB signal, composed of four independent optical SSB sub-channels, is efficiently generated using two low-cost dual-drive Mach–Zehnder modulators (DDMZMs) driven by a single laser, eliminating the need for multiple laser sources and complex optical filtering. At the receiver, a Stokes vector receiver (SVR) is employed to simultaneously detect the left and right sidebands from both polarizations. Polarization de-rotation, which is essential for recovering the multiplexed signals, is performed directly in the Stokes space. Furthermore, Kramers–Kronig (KK) processing is applied after polarization alignment to reconstruct the complex optical field and mitigate the signal-to-signal beating interference (SSBI) inherent in square-law direct detection. We demonstrate that the 448 Gb/s PDM-Twin-SSB PAM-4 signal can be successfully transmitted over 80 km standard single-mode fiber (SSMF) with a pre-FEC bit-error-rate (BER) below the 7% hard-decision forward error correction (HD-FEC) threshold of 3.8 × 10⁻³, achieving an OSNR requirement of 41 dB. The system performance is systematically evaluated through optimization of the carrier-to-signal power ratio (CSPR) and the guard band between the optical carrier and the sidebands. The results confirm that the proposed DDMZM-based PDM-Twin-SSB architecture, combined with the SV-KK receiver, provides a practical and cost-effective polarization-demultiplexing solution for single-wavelength high-speed optical interconnects, enabling robust transmission performance without requiring a local oscillator or active polarization control.

Under non-uniform temperature conditions, the reflected spectrum of a fiber Bragg grating (FBG) can broaden and distort, making a single Bragg wavelength insufficient to describe the axial temperature distribution. This study investigates a one-dimensional convolutional neural network (1D-CNN) for reconstructing non-uniform temperature profiles from the full reflected spectra of a single uniform FBG. A numerical dataset of 2000 paired spectra and temperature profiles was generated using the transfer matrix method (TMM), which relates forward- and backward-propagating optical field amplitudes across grating segments. Four representative profiles—linear, exponential, hyperbolic, and Gaussian—were considered to represent typical temperature distributions associated with heat conduction and localized heating. The CNN was trained to map the 1001-point normalized reflectivity spectrum to a 200-point axial temperature profile. Across five independent training runs, the CNN achieved a mean absolute percentage error (MAPE) of 11.02 ± 0.60%, a root mean square error (RMSE) of 3.06 ± 0.38 K, and a coefficient of determination (R²) of 0.971 ± 0.005 on the independent test set. Partial least squares (PLS) regression yielded a higher RMSE of 4.08 K. TMM-based iterative optimization reproduced the target profile with errors close to numerical precision but required 28.50 s per spectrum, compared with 6.85 ms for the CNN. Under simulated Gaussian intensity noise, the RMSE of a fixed CNN increased to 4.89 K at 30 dB and 10.73 K at 20 dB, whereas wavelength jitter with a standard deviation of 5 pm yielded an RMSE of 2.76 K. These results suggest that the proposed CNN can reduce online reconstruction time for the investigated temperature profiles, although experimental validation remains necessary.

Reinforced galvanic isolation and solid-state overcurrent protection (eFuse) are critical safety barriers in automotive high-voltage (HV) power electronics, yet they are rarely evaluated together. This paper reports destructive robustness tests of both barriers on hardware of the same automotive HV subsystem, at up to 850 VDC, as a complement to IEC 60747-17 certification. For the isolation barrier, 39 samples from two 5 kVrms-rated digital-isolator families (reinforced and basic insulation) were stressed with continuous DC/AC voltage. Except for two samples tested without energy limitation, which failed catastrophically, all samples passed a 1.5 kV/60 s HIPOT test and read above the 500 MΩ range of the multimeter, even when visibly punctured. For the overcurrent barrier, five eFuse power stages were tested in an 850 VDC capacitor-fed short circuit and at 10–360 VDC with a 120 A current limit. All five failed open; in the 850 VDC test the current was interrupted within 0.5 ms (limit 1.2 s), and no thermal event occurred outside the enclosure within the 12 s observation window. The results are exploratory, given the small samples, and support several recommendations for HV safety-validation testing. After every eFuse fault, the HV–LV isolation of the subsystem remained intact.