超密集蜂窝网络智能干扰协调算法

 

秦爽¹, 董星辰², 冯钢¹ 摘要：提出了一种面向超密集蜂窝网智能干扰协调的动态功率控制算法。从网络动态决策的角度出发，将动态干扰环境下，超密集蜂窝网络中的小基站发送功率的动态调整问题建模为一个马尔科夫决策过程。在此基础上，设计了一种基于Actor-Critic（AC）方法的小基站功率智能控制算法，并通过仿真实验验证了所提算法的性能。仿真结果显示，与传统的干扰协调算法相比，所提出智能功率控制算法能有效降低网络中基站间的干扰，提升网络传输性能。
关键词：超密集蜂窝网络；干扰协调；功率控制；强化学习；AC算法


Abstract： An intelligent power control algorithm for interference coordination in ultra dense cellular networks (UDN) is proposed. By considering the impact of dynamical interference between different base stations, the problem of dynamical power control for small base stations in UDN is modeled as a Markov decision process (MDP). Then, an intelligent power control algorithm based on AC method is designed, and the performance of the proposed algorithm by using simulation experiments is validated. The numerical results show that, comparing to the traditional interference coordination algorithms, the proposed intelligent power control algorithm can reduce the inter-cell interference and improve transmission performance in UDN.
Keywords: UDN; interference coordination; power control; reinforcement learning; AC algorithm