This letter proposes a joint cell activation and user association scheme to address the power consumption and backhaul load balancing issues in green dense heterogeneous networks (HetNets). In this scheme, an optimization problem is formulated for jointly minimizing power consumption and optimizing backhaul load balance of dense HetNets, subject to diverse quality of service (QoS) requirements and heterogeneous backhaul capacities. To solve this problem, a Q-learning-based algorithm is developed. Results show that the proposed scheme achieves substantial performance improvements over several baseline schemes in terms of fairness, QoS and energy efficiency.

中文翻译：

---

用于绿色 HetNet 中回程负载平衡的联合小区激活和用户关联

这封信提出了联合小区激活和用户关联方案，以解决绿色密集异构网络（HetNets）中的功耗和回程负载平衡问题。在该方案中，根据不同的服务质量 (QoS) 要求和异构回程容量，制定了一个优化问题，用于联合最小化功耗和优化密集 HetNet 的回程负载平衡。为了解决这个问题，开发了一种基于 Q-learning 的算法。结果表明，所提出的方案在公平性、QoS 和能源效率方面比几个基线方案实现了实质性的性能改进。