Moynul Rifat

IEEE CONFERENCE PAPER  ·  2ND AUTHOR

Performance and Efficiency Analysis of Modern Filesystems on Consumer-Grade NVMe SSDs Under Real-World Mixed Workloads

A comprehensive benchmarking study evaluating the throughput, latency, and I/O efficiency of modern filesystems [NTFS, BTRFS, ZFS, XFS & EXT4] on Consumer-Grade NVMe SSDs. This research provides data-driven insights into filesystem behavior and resource overhead under real-world mixed workloads.

JOURNAL PAPER  IN PROGRESS

A Closed-Loop Ensemble Learning Framework for High-Fidelity Forward Prediction and Inverse Design of Ring Core-Photonic Crystal Fibers via Machine Learning

Data-driven investigation and automated inverse design of ring-core photonic crystal fibers (RC-PCFs) to enable robust and stable orbital angular momentum (OAM) mode transmission tailored for high-capacity optical networks via interpretable ensemble machine learning framework.

OFFICIAL HARDWARE BASELINE  ·  RLSOK v1.5.3

ROS2 Modular DiffBot Integrated into RLSOK Configuration Management Tool

My custom-built ROS2 Sim-to-Real architecture was selected and integrated as the official physical hardware baseline for RLSOK, a global open-source configuration management tool. This third-party integration validates the robustness of my custom TCP/WiFi kinematics bridge and URDF modeling, allowing operators to track offline hardware changes without active ROS dependencies.