Joint Game Theory and Greedy Optimization Scheme of Computation Offloading for UAV-Aided Network

The UAV-aided network has great potential in emergency response scenarios, which is a necessary complement to the conventional terrestrial cellular network. With the increase of data to be processed, users' expectations for low latency and low energy consumption are also improved. Traditional networks are unable to meet the rapidly growing actual needs gradually. To solve this problem, as an important role of mobile edge computing (MEC), computing offloading technology makes edge nodes have processing ability and reduce time delay, which further improves the practicability of UAV-aided network. However, the computing resources deployed on the edge nodes are limited. In order to solve the competition of non-cooperative game scenario and ensure the algorithm complexity is as low as possible, we adopt a heuristic algorithm based on game theory, consider the weighted sum of delay and energy consumption as the objective function and decision basis, and propose to introduce the greedy strategy to optimize the MEC server selection policy in the algorithm. The simulation results show that the joint algorithm can significantly reduce the overall user loss compared with the two traditional schemes. Compared with the game theory algorithm itself, the greedy optimization strategy can also bring about 12% gain when the parameters meet certain conditions. Meanwhile, the selection strategy with O(n) time complexity introduced does not affect the highest term, and the total complexity is still O(mn) at each time slot.