Robust and Secure UAV-Assisted MEC Design: Enhancing User Experience Amidst Uncertainties

This paper proposes a novel design for UAV-assisted mobile edge computing (MEC) that emphasizes user experience, addressing often overlooked elements like security and system robustness. The system works through the collaboration of multiple ground users and a UAV, completing computational tasks despite potential threats and uncertainties like eavesdropping or unpredictable flight paths due to severe weather or UAV instability. We formulate these challenges as probabilistic constraints in our model and cleverly merge the safety and robustness problems into one, greatly reducing the complexity of the problem. The model utilizes variations in energy consumption and delay to calculate a user's body sensation index, which we aim to minimize within these constraints. Given the non-convex nature of the problem, we employ Bernstein-type inequalities and the posterior method to transform the constraints into a deterministic form and the objective function into a convex one. Non-convex constraints are approximated as convex ones through first-order Taylor expansion and handled using successive convex approximation methods. An iterative algorithm is then employed for efficient solutions. Numerical simulations validate the effectiveness of our design in enhancing user experiences.