Research starts with curiosity; Engineering continues with discipline.

Performance Characterization of OpenFOAM UAV Aerodynamics Simulations Across Mesh Resolutions and Flow Regimes Using Linux perf

This study will investigate the computational performance of OpenFOAM simulations for flow around rotary- and fixed-wing UAV configurations under different mesh resolutions and flow regimes, including laminar and turbulent cases. Simulations will be executed in both serial and MPI-based parallel modes to evaluate core scalability, memory usage, and solution time. Linux perf will be used to examine processor-level hardware performance counters and identify major computational bottlenecks. The study will also assess how mesh size, solver settings, and parallel decomposition affect hardware utilization and scaling efficiency. The resulting performance data will be used to characterize the most computationally intensive stages of the CFD workflow and to provide a basis for selecting candidate operations for future hardware acceleration.

Eligibility: This project is open exclusively to undergraduate students of the Istanbul Technical University's Faculty of Aeronautics and Astronautics.

Aerodynamic System Identification Using Multivariate Orthogonal Functions for In-Flight Applications

This study will implement a multivariate orthogonal functions (MOF)-based aerodynamic system identification method intended for real-time in-flight use. Focusing on longitudinal dynamics, force and moment model terms will be selected automatically using a statistical criterion, and their coefficients estimated by sequential orthogonalization. The method will be evaluated with synthetic maneuver data from two openly documented nonlinear aircraft models: a closed-form polynomial F-4 model, enabling term-by-term comparison, and a table-based F-16 reference model. The latter will be assessed on an independent validation grid using error norms and local stability/control derivatives. Orthogonal multisine excitations will be designed to reduce regressor correlation and compared with conventional inputs. Initial testing will be offline and batch-based, with a modular codebase supporting future lateral-directional dynamics, realistic measurement noise, and recursive real-time updating.

Eligibility: This project is open exclusively to undergraduate students of the Istanbul Technical University's Faculty of Aeronautics and Astronautics.

Improving Multirotor Velocity Estimation During GNSS Outages Using Aerodynamic Modeling

In this study, the aim is to improve velocity estimation performance for multirotor aerial vehicles during temporary loss of GNSS measurements by exploiting the aerodynamic characteristics of the vehicle. The relationships among vehicle motion, wind, and aerodynamic forces acting on the airframe will be investigated to develop a suitable model. This model will be combined with available inertial sensor measurements to provide additional constraints on vehicle velocity during GNSS outages. Different aerodynamic model alternatives will be compared in terms of estimation accuracy, practical applicability, computational cost, and sensitivity to model uncertainty. The developed approach will be evaluated by examining velocity error growth and drift during GNSS-denied intervals, and will be validated using simulations and suitable experimental flight data.

Eligibility: This project is open exclusively to undergraduate students of the Istanbul Technical University's Faculty of Aeronautics and Astronautics.

Development of an Automated Allan Variance Characterization Framework and Test Setup for MEMS Inertial Sensors

Noise modeling of MEMS IMU sensors used in autopilot and navigation systems is critical to the performance of state estimation algorithms such as Kalman filters. This final-year project will develop a desktop application for automated characterization of accelerometer and gyroscope noise from long-duration static measurements. The software will compute Allan deviation curves, automatically identify characteristic slope regions, and estimate parameters including Angle/Velocity Random Walk, Bias Instability, and Rate/Acceleration Random Walk. A physical test setup will also be developed to provide controlled and repeatable data acquisition from different IMUs. The resulting framework will support visualization, comparison, and reporting of sensor characteristics, enabling repeatable evaluation of candidate inertial sensors and providing experimentally derived noise parameters for subsequent navigation and state-estimation studies.

Eligibility: This project is open exclusively to undergraduate students of the Istanbul Technical University's Faculty of Aeronautics and Astronautics.

Design and Implementation of a Modular SDR-Based GNSS Signal Simulator

In this project, a modular GNSS signal simulator capable of generating synthetic RF signals will be developed using a Software-Defined Radio (SDR) platform. The initial implementation will focus on GPS, modeling satellite ephemeris, propagation delays, and Doppler shifts to synthesize baseband I/Q signals using PRN codes and BPSK modulation. The generated digital signal will be converted to RF using SDR hardware and validated with a commercial GNSS receiver. Validation will include receiver acquisition and positioning performance, and, where accessible, comparison of generated and observed navigation measurements. The software architecture will be modular, allowing additional constellations such as Galileo, GLONASS, and BeiDou, as well as additional frequency bands, to be incorporated depending on the project schedule.

Eligibility: This project is open exclusively to undergraduate students of the Istanbul Technical University's Faculty of Aeronautics and Astronautics.

Feasibility Analysis of Signals of Opportunity (SoOP) for GNSS-Denied Environments and SDR-Based Extraction of Navigation Observables

In this project, the feasibility of Signals of Opportunity (SoOP) for alternative navigation in scenarios where GNSS signals are unavailable or jammed will be investigated. RF signals from terrestrial broadcasts, cellular networks, and LEO satellite constellations will be evaluated in terms of bandwidth, coverage, geometry, synchronization characteristics, and potential navigation accuracy. In the implementation phase, selected signals will be recorded in the Istanbul/campus environment using Software-Defined Radio (SDR) hardware. Digital signal processing (DSP) algorithms will then be developed to extract applicable navigation observables, such as Doppler shift, Time Difference of Arrival (TDoA), and carrier phase. The noise characteristics and temporal stability of the extracted observables will be analyzed, and their suitability for integration into alternative navigation systems will be assessed.

Eligibility: This project is open exclusively to undergraduate students of the Istanbul Technical University's Faculty of Aeronautics and Astronautics.