Research Projects

With the advent of next-generation communication systems, new use cases such as virtual reality (VR), augmented reality (AR), autonomous robotics, smart healthcare, and industrial automation are becoming increasingly prominent. Most of these emerging services are location-aware, accurate, and real-time localization with centimeter-level precision, essential to ensure seamless and responsive user experiences [1-3]. Consequently, for applications like AR/VR, localization is becoming increasingly vital to the success of these technologies, enabling immersive, adaptive, and intelligent system behavior in complex environments [4]. In indoor environments, localization becomes critical, as techniques like GPS do not function effectively. Furthermore, existing indoor localization methods (e.g., Wi-Fi, Bluetooth) often fall short due to signal blockage, multipath distortion, and limited spatial resolution [5]. Localization using the Terahertz (THz) band is a promising solution to meet these stringent requirements. Due to its ultra-wide bandwidth and shorter wavelengths, THz-based localization offers several distinct advantages, making it a key enabling technology. THz signals support high-resolution localization, enabling sub-centimeter accuracy and precise orientation tracking. THz signals do not penetrate objects, leading to a more direct and predictable relationship between propagation paths and the environment. Together, these advantages position THz-based localization as a critical component for enabling precise, reliable, and accurate positioning in next-generation wireless systems [6-7]. Furthermore, machine learning (ML) techniques have demonstrated improved performance across a variety of tasks due to their ability to learn from data and adapt to changing environments [8]. Over the past decade, ML has proven to be valuable in numerous applications, including wireless communication and signal processing. Building on this, the project will explore and implement advanced ML algorithms to enhance the accuracy, adaptability, and robustness of the THz-based indoor localization system for six degrees of freedom tracking—capturing not only the three-dimensional spatial position but also the pitch, yaw, and roll for immersive and responsive interaction [4]. Since applications like AR/VR are highly sensitive to latency, any delay in the localization system can significantly hinder the user experience. It is essential to develop algorithms that operate efficiently with low computational complexity and minimal processing delay. This approach enables real-time, high-accuracy performance by learning from the environment and dynamically adjusting to changes in signal conditions and user movement. The primary objective of the project is to develop and validate a high-precision, real-time indoor localization framework that integrates THz localization with advanced ML techniques, specifically targeting 6DoF localization required in immersive AR/VR environments.