Femtocell双层网络中基于Q-learning的子信道分配方案

在Femtocell家庭基站(Femtocell Base Station, FBS)组成的异构网络中，为提升网络的频谱效率，FBS与Macrocell宏基站(Macrocell Base Station, MBS)一般要求是同频部署，然而同频部署会产生同信道干扰。为了实现FBS的大规模部署，降低网络同信道干扰影响变得尤为重要。该文提出一种基于Q-learning的子信道分配方案，既保证大量部署的FBS不会对MBS带来过高的跨层干扰，同时也降低了FBS之间的同层干扰。同时针对FBS稀疏部署和密集部署的场景，分别进行了算法的仿真验证，其仿真结果表明该算法降低了干扰，验证了理论的正确性。

Subchannel Allocation Scheme for Two-tire Femtocell Networks Based on Q-learning

In order to improve the spectrum efficiency in the Femtocell home Base Station (FBS) heterogeneous network, FBS and Macrocell Base Station (MBS) are usually deployed with the same frequency. However, the same frequency deployment will inevitably lead to larger co-channel interference. In order to achieve the large-scale deployment of FBS, reducing the interference of the network with the channel is particularly important. In this paper, a sub channel allocation scheme is propsed based on Q-learning. It can ensure that FBS will not bring high cross-layer interference of MBS, while it reduces the same layer interference between two FBS. The simulation of the algorithm of FBS sparse deployment and dense deployment situation are performed, respectively. Simulation results show that this algorithm reduces the same layer interference and verifies the correctness of the theory.