Joint optimized multi-user access and UAV deployments based on heterogeneous revenue in IoT network

With the explosion demands of computation in the internet-of-things (IoT) sector, as an emerging technique with considerable computational capability, unmanned aerial vehicle (UAV) assisted mobile edge computing (MEC) has been proposed to reduce the latency and improve the quality-of-experience (QoE) of the user equipments (UEs). However, being existed massive heterogeneous UEs with different time sensitivity, how to offload the tasks of UEs, and where to deploy the positions of UAVs are two critical factors for the MEC system. In this work, we investigate a multi-UAV-assisted MEC system to maximize the sum of UEs' revenue in the event that natural disasters block communication conditions. A two-hierarchy Stackelberg game framework model is constructed, with the upper-layer UAVs as leaders performing location deployments, while the lower-layer UEs as followers conducting offloading selections. Different from the existing revenue functions, for followers, we adopt three revenue functions in terms of different UEs having distinct responses to time consuming. Furthermore, both the leader-layer and follower-layer subgame are proved to be exact potential games (EPG) with at least one Nash equilibrium (NE), then demonstrating the existence of Stackelberg equilibrium (SE). Additionally, we evaluate the price of anarchy (PoA) of the NE and illustrate that the value of PoA is relatively small, which corroborates that our solution is close to the global optimum. Finally, the simulation results demonstrate that the proposed solution can improve the total revenue of UEs under guaranteeing the convergence.