arXiv Science⌕ Search

arXiv subjects

Mohammad Asgari

Publications and source records attributed to Mohammad Asgari.

2 recordsLinked to original sources

BIC-Engineered Silicon Nitride Microring Resonators Enabled by an Etch-Free Laser-Writing Approach

We report the initial experimental demonstration of a complementary, etch-free approach with fewer fabrication steps, in useful SU-8 ridges written by direct ultraviolet laser lithography on an unpatterned silicon nitride film define waveguides and microring resonators. Because the nitride film is never patterned, scattering at patterned nitride sidewalls does not arise, and the remaining radiative loss is engineered by design. The guided transverse-magnetic mode lies within the continuum of transverse-electric slab modes, and we use bound states in the continuum (BICs) as a design tool to control this radiation. Coupled-mode analysis and full-vectorial simulations show that radiation from the two ridge sidewalls cancels by destructive interference at discrete widths, with a first BIC at 2.25 micron and radiation below 1 dB/cm over a 0.3 micron width window and about 200 nm of bandwidth, and that curvature confines this cancellation to discrete bands of bending radius and width. Laser-written waveguides follow the predicted width dependence of the loss, reaching approximately 1.5 dB/cm near the BIC, where non-radiative loss dominates. A microring designed on a radius-selected BIC band exhibits loaded quality factors up to 1.6 \times 10^4 and, under the over-coupled assignment supported by a coupler model, an intrinsic quality factor of 3.3 \times 10^5. Because radiation is set by a patterned low-index layer on an unetched film, the same approach could be extended to other nanostructures. Optimization of the writing conditions and materials is ongoing.

physics.optics↗

Audio Compression Using Graph-based Transform

Graph-based Transform is one of the recent transform coding methods which has been used successfully in the state-of-art data decorrelation applications. In this paper, we propose a Graph-based Transform (GT) for audio compression. Hence, we introduce a proper graph structure for audio. Then the audio frames are projected onto an orthogonal matrix consisting of eigenvectors of the introduced graph matrix, leading to the sparse coefficients. The results show that the proposed method outperforms the conventional transform methods like Discrete Cosine Transform (DCT) and Walsh-Hadamard Transform (WHT) in decorrelation of the audio signals.

eess.AS↗