Gui-Rong Zhou

Computation of full-vector modes for bending waveguide using cylindrical perfectly matched layers

A new full-vector approach to calculate leaky modes on three-dimensional bending waveguides is developed and demonstrated with the help of the cylindrical perfectly matched layer (CPML) numerical boundary conditions. By utilizing the complex coordinate stretching technique in the cylindrical system, a new set of full-vector wave equations for the bending waveguide structures are derived for the perfectly matched layer regions. Numerical solutions by the finite-difference schemes for the new wave equations are shown to yield highly accurate complex propagation constants (e.g., the bending-induced phase shifts and leakage losses) and modal field patterns, due primarily to the effective CPML.

A novel planar waveguide wavelength demultiplexer design for integrated optical triplexer transceiver

A novel wavelength demultiplexer design is proposed for the development of integrated optical triplexer transceiver in fiber-to-the-home (FTTH) applications. Still taking the slab waveguide as the beam confining element along the vertical direction but following a different approach from the arrayed waveguide grating, we use planar lenses to convert the phase front and planar diffraction grating to split the collimated beams in different wavelengths. The planar lenses and grating are realized through the effective refractive index difference arising from the difference on the cladding layer thickness of the slab waveguide. In applications such as optical triplexer transceivers where relatively large wavelength spacing is required among signal channels over a broad wavelength range, this design seems more appealing on simulated performance and fabrication cost in comparison with various existing integrated structures. Preliminary simulation results have shown that this component, once integrated with laser diode and photodetectors, has the potential to reach the required specifications as the optical triplexer transceiver in optical network unit for FTTH applications.

Space mapping technique for design optimization of antireflection coatings in photonic devices

Space-mapping (SM) technique is applied for design optimization of antireflection (AR) coatings for photonic devices such as the semiconductor optical amplifiers (SOA). The approximate and efficient transfer matrix method (TMM) serves as the coarse model for design optimization, whereas the time-intensive and accurate finite-difference time-domain (FDTD) method is used as the fine model for model calibration. A mapping is established between the parameter spaces of the coarse and the fine models so that the fine model design becomes the inverse mapping of the optimized coarse model design. Remarkable performance of the SM technique in terms of efficiency and accuracy in the design optimization is demonstrated by way of examples. It is shown that, in the context of multilayer coating design, the desired broadband ultralow reflectivities can be obtained within three fine model (FDTD) calculations.

Design of deeply etched antireflective waveguide terminators

An alternative solution to achieve an antireflective waveguide terminator is proposed by adopting a deeply etched waveguide structure to replace the conventional facet interference coatings. The performance is evaluated by different numerical approaches and optimum designs can be achieved based on the combination of the finite-difference time-domain method and the transfer matrix method. Perfectly matched layer absorbing boundary conditions are employed and pre-optimized in order to eliminate any nonphysical reflections due to the computation window introduced artificially. Results show that a power reflectivity of less than 5.0/spl times/10/sup -3/ over almost the entire C-band with a minimum value as low as 1/spl times/10/sup -5/ can be achieved. The effects on etching with a tilted angle and etching with finite depth are also studied.

Mode calculation by beam propagation method combined with digital signal processing technique

A robust computational scheme for calculation of multiple modes of optical waveguides is developed and presented. The new method uses the beam propagation methods to generate the modal fields and a digital signal processing technique for mode parameter extraction. It can be applied to both guided and leaky modes with different formulations (i.e., scalar, semi-vector or full-vector) and discretizations (e.g., finite difference or finite element). Salient features of the new method are discussed and advantages over other computational methods based on boundary-value eigen solvers and initial-value propagation solvers are demonstrated by way of examples.

An efficient split-step time-domain beam-propagation method for modeling of optical waveguide devices

This paper presents a novel split-step time-domain beam-propagation method (TD-BPM) for simulation of time-domain pulse propagating in optical waveguide devices. Different from the traditional methods, where the time-marching step is strictly limited by the relation /spl Delta/t=/spl Delta/z/v/sub g/, the new split-step algorithm allows much more relaxed temporal and spatial steps, and therefore is much more efficient.

Wave equation-based semivectorial compact 2-D-FDTD method for optical waveguide modal analysis

A wave equation-based semivectorial compact 2-D finite-difference time-domain (2-D-FDTD) method is developed and validated for optical waveguide modal analysis. This approach is a combination of the Maxwells equation-based compact 2-D-FDTD and the wave equation-based semivectorial FDTD methods. Perfectly matched layer (PML) absorbing boundary condition (ABC) is also extended to this approach. Excellent accuracy is achieved for the entire spectrum even in the region near the cutoff. Through extensive study on the excitation conditions, it indicates that this method, when used as an explicit optical mode solver, is extremely robust.

A scalar finite-difference time-domain method with cylindrical perfectly matched layers: application to guided and leaky modes of optical waveguides

A new scheme for the scalar finite-difference time-domain (FDTD) method with perfectly matched layer boundary conditions in cylindrical coordinate system is developed and presented. The FDTD method is applied successfully for the calculation of both guided and leaky modes of optical waveguides with circular symmetry.

A hybrid time-domain technique for simulation of high-density integrated optical circuits

This paper presents a new simulation method for modeling and simulation of high-density integrated optical circuits based on high index contrast (HIC) waveguides with complex topology. The method combines the time-domain reflective beam propagation method (TD-RBPM) and the slow-wave finite-difference time-domain method and is hence referred to as the time-domain hybrid BPM (TD-HBPM). It is capable of handling arbitrary optical integrated circuits with perpendicularly located input and output ports. The application to the two HIC optical circuits shows the accuracy and efficiency of this method.

A digital filter approach for complex frequency-shifted perfectly matched layer in semivectorial FDTD

A new formulation of the perfectly matched layer (PML) for the semivectorial finite-difference time-domain (FDTD) method in optical waveguide simulation is presented by incorporating the infinite-impulse response (IIR) digital filter technique. The complex frequency-shifted PML is implemented through Z transformation, where the second-order derivatives in semivectorial FDTD are realized by two cascaded first-order recursive IIR digital filters. The numerical examples indicate that the new scheme has better performance compared with the normal PML.