Resource allocation based on differentiated user services for indoor hybrid multi-color VLC/WiFi networks
Resource allocation based on differentiated user services for indoor hybrid multi-color VLC/WiFi networks
复制标题
DOI：
10.1016/j.optcom.2022.129072
复制
发表时间：
2022-10
影响因子：
2.4
通讯作者：
Huanlin Liu;Shuai Yang;Yong Chen;X. Yuan;Jinlin Chen;Ke Chen;Haonan Chen
中科院分区：
文献类型：
--
作者：
Huanlin Liu;Shuai Yang;Yong Chen;X. Yuan;Jinlin Chen;Ke Chen;Haonan Chen
多色可见光通信（VLC）和无线保真（WiFi）网络为室内无线通信提供了一种有前途的解决方案，但混合网络面临接入点（AP）负载和资源分配不平衡的挑战。针对多色VLC/WiFi混合网络，提出了一种基于用户业务差异化（RAMUSD）的资源分配方法。在我们的设计中，用户的付费级别是根据用户的信道条件和请求的数据速率定义的。异构网络根据用户的付费等级为用户选择业务AP，付费等级高的用户由VLC接入点（access point， AP）中的一个接入点提供服务，付费等级低的用户由WiFi AP提供服务，然后以平均满意度作为负载指标，调整WiFi AP或VLC AP提供服务的用户数，实现AP间的负载均衡。RAMUSD利用潜在博弈理论优化用户迁移和ap间资源分配，提高负载均衡指标。在潜在博弈中，用户可以通过改变接入AP来提高其返回值，从而使得追求更多资源的用户放弃拥塞的AP。仿真结果表明，该算法比基于进化博弈理论的增强比例公平算法（EPF-EGT）具有更好的系统吞吐量和用户满意度。
Multi-color visible light communication (VLC) and wireless fidelity (WiFi) networks provide a promising solution for the indoor wireless communication, while the hybrid networks face the challenge of imbalanced access point (AP) load and resource allocation. A resource allocation method based on user service differentiation (RAMUSD) is proposed for multi-color VLC/WiFi hybrid networks. In our design, the user’s premium level is defined according to the user’s channel conditions and requested data rate. The heterogeneous network selects the service AP for the users according to its premium level, user with high premium level is served by one of VLC access points (APs), while user with low premium level is served by WiFi AP. Then, the average satisfaction is used as the load index to adjust the number of users served by the WiFi AP or the VLC AP to achieve load balance among APs. Furthermore, the RAMUSD uses the potential game theory to optimize user’s migration and resource allocation among APs for improving the load balancing index. In the potential game, user can improve its return value by changing access AP. So, it makes user who pursue more resources give up the congested APs. Simulation results show that the proposed RAMUSD can achieve better system throughput and user satisfaction than an enhanced proportional fairness algorithm based on evolutionary game theory (EPF-EGT).