I am a researcher working at the intersection of Reinforcement Learning, Edge Intelligence, and Wireless Networking, focusing on adaptive resource allocation and distributed decision-making in dynamic Mobile Edge Computing (MEC) systems.

I developed QECO, a Deep Reinforcement Learning (DRL)-based computation offloading algorithm published in IEEE Transactions on Network Science and Engineering (2025). Built on Dueling Double DQN (D3QN), QECO integrates LSTM-based workload modeling to capture temporal workload dynamics and optimize long-term QoE under task deadline and energy constraints. Simulation results demonstrated up to 37% higher QoE and faster convergence than baseline approaches. Paper

My current research investigates Multi-Agent Reinforcement Learning (MARL) for collaborative computation offloading in MEC, focusing on partial observability, agent heterogeneity, and non-stationary system dynamics. I formulate interdependent offloading decisions as Decentralized Partially Observable Markov Decision Processes (Dec-POMDPs) and explore Centralized Training with Decentralized Execution (CTDE) to develop coordinated policies for distributed resource management.

• Research Methods: Reinforcement Learning, Stochastic Systems Modeling, Sequential Decision-Making, MDPs, Queueing Theory, Simulation, Performance Evaluation, Resource Allocation and Optimization, Experimental Design and Benchmarking
• Programming & Machine Learning: Python, R, C++, SQL, Bash, PyTorch, TensorFlow, Keras, Scikit-learn
• Systems & Tools: Linux, Git, NS-3, Mininet, NumPy, Pandas, Matplotlib, Jupyter, Flask, LaTeX