Wireless cellular networks have many parameters that are normally tuned upon
deployment and re-tuned as the network changes. Many operational parameters
affect reference signal received power (RSRP), reference signal received
quality (RSRQ), signal-to-interference-plus-noise-ratio (SINR), and,
ultimately, throughput. In this paper, we develop and compare two approaches
for maximizing coverage and minimizing interference by jointly optimizing the
transmit power and downtilt (elevation tilt) settings across sectors. To
evaluate different parameter configurations offline, we construct a realistic
simulation model that captures geographic correlations. Using this model, we
evaluate two optimization methods: deep deterministic policy gradient (DDPG), a
reinforcement learning (RL) algorithm, and multi-objective Bayesian
optimization (BO). Our simulations show that both approaches significantly
outperform random search and converge to comparable Pareto frontiers, but that
BO converges with two orders of magnitude fewer evaluations than DDPG. Our
results suggest that data-driven techniques can effectively self-optimize
coverage and capacity in cellular networks.

本研究比较了深度确定性策略梯度（DDPG）和多目标贝叶斯优化（BO）两种方法，并利用实验表明 BO 比 DDPG 收敛更快，少用两个数量级的计算。这些结果表明数据驱动技术可以有效地进行无线蜂窝网络的覆盖和容量自优化。

利用机器学习优化蜂窝网络的覆盖范围和容量

Optimizing Coverage and Capacity in Cellular Networks using Machine  Learning

The model-driven power allocation (PA) algorithms in the wireless cellular
networks with interfering multiple-access channel (IMAC) have been investigated
for decades. Nowadays, the data-driven model-free machine learning-based
approaches are rapidly developed in this field, and among them the deep
reinforcement learning (DRL) is proved to be of great promising potential.
Different from supervised learning, the DRL takes advantages of exploration and
exploitation to maximize the objective function under certain constraints. In
our paper, we propose a two-step training framework. First, with the off-line
learning in simulated environment, a deep Q network (DQN) is trained with deep
Q learning (DQL) algorithm, which is well-designed to be in consistent with
this PA issue. Second, the DQN will be further fine-tuned with real data in
on-line training procedure. The simulation results show that the proposed DQN
achieves the highest averaged sum-rate, comparing to the ones with present DQL
training. With different user densities, our DQN outperforms benchmark
algorithms and thus a good generalization ability is verified.

本文研究了基于模型驱动的电力分配算法在具有干扰多接入信道（IMAC）的无线蜂窝网络中的应用。通过深度强化学习（DRL）的两步训练框架，采用深度 Q 网络（DQN）和深度 Q 学习算法（DQL）从离线学习中获得指定数据集，而对于在线学习过程中的真实数据来说，DQN 可进一步微调。这种方法比现有的 DQL 训练方法更好，具有很好的推广能力。

多用户蜂窝网络中的功率分配策略 —— 基于深度 Q 学习方法

Power Allocation in Multi-user Cellular Networks With Deep Q Learning  Approach

Wireless cellular networks feature two emerging technological trends. The
first is the direct Device-to-Device (D2D) communications, which enables direct
links between the wireless devices that reutilize the cellular spectrum and
radio interface. The second is that of Machine-Type Communications (MTC), where
the objective is to attach a large number of low-rate low-power devices, termed
Machine-Type Devices (MTDs) to the cellular network. MTDs pose new challenges
to the cellular network, one if which is that the low transmission power can
lead to outage problems for the cell-edge devices. Another issue imminent to
MTC is the \emph{massive access} that can lead to overload of the radio
interface. In this paper we explore the opportunity opened by D2D links for
supporting MTDs, since it can be desirable to carry the MTC traffic not through
direct links to a Base Station, but through a nearby relay. MTC is modeled as a
fixed-rate traffic with an outage requirement. We propose two network-assisted
D2D schemes that enable the cooperation between MTDs and standard cellular
devices, thereby meeting the MTC outage requirements while maximizing the rate
of the broadband services for the other devices. The proposed schemes apply the
principles Opportunistic Interference Cancellation and the Cognitive Radio's
underlaying. We show through analysis and numerical results the gains of the
proposed schemes.

本文论述了利用设备到设备通信解决无线蜂窝网络中大量低速、低功率设备接入和偏远设备断开的问题，并提出了两种网络协助的设备到设备方案，以实现低速、低功率设备的联网并提高宽带服务的速度。