UAV service enhancement mechanism deep deterministic policy gradient for scalable video coding transmission

Abstract

Unmanned aerial vehicles (UAV) are an important component of future air-space-ground integrated network due to their onboard storage, communication, and computing capabilities.However, existing research on UAV-assisted edge networks often focuses more on the network perspective and lacks consideration of user perspectives and their requirements.Therefore, a UAV service enhancement mechanism based on deep deterministic policy gradient (DDPG) for scalable video coding (SVC) transmission was proposed from the user’s perspective.Firstly, an elastic video transmission method based on SVC was proposed in conjunction with UAV to improve differentiated user experiences.Secondly, a UAV trajectory planning design based on the DDPG algorithm was proposed to maximize the number of users receiving enhanced layer videos and ensure effective coverage enhancement by UAV in hotspot areas.Simulation results show that compared with both the deep Q network (DQN) algorithm and shortest path (SP) algorithm under different user distributions, the proposed algorithm can increase the average number of enhanced layers received by 47.9% and 76.4%, respectively.This study successfully achieves improved coverage in hotspot areas while also providing differentiated user experiences through its proposed methods.

Keywords

• DDPG;UAV;SVC;trajectory planning