Energy Efficient Multi-aerial Base Station Deployment Via DDPG-Mix Algorithm

Compared to ground base stations, aerial base stations (AeBSs) offer advantages in terms of overcoming limitations posed by complex terrain and transportation constraints, providing broader coverage and faster response speeds. In this paper, we investigate the problem of energy efficiency maximization in a multiple AeBSs network and adopt multi-agent deep reinforcement learning (MADRL) for the deployment of AeBSs. Specifically, considering the partial observation range of AeBSs, the multi-AeBS deployment problem is modeled as a decentralized partially observable markov decision process (Dec-POMDP) and a deep deterministic policy gradient mix (DDPG-MIX) algorithm is designed to solve the problem. The proposed algorithm uses the value decomposition framework to solve the lazy agent problem. Simulation results show that our proposed DDPG-MIX algorithm performs better than other baseline algorithms in the multi-AeBS deployment scenario.