Kim Seung-Yeon, Ko Haneul
Department of Computer Convergence Software, Korea University, Sejong 30019, Republic of Korea.
Department of Electronic Engineering, Kyung Hee University, Yongin-si 17104, Republic of Korea.
Sensors (Basel). 2024 Dec 4;24(23):7750. doi: 10.3390/s24237750.
In beyond fifth-generation networks, millimeter wave (mmWave) is considered a promising technology that can offer high data rates. However, due to inter-cell interference at cell boundaries, it is difficult to achieve a high signal-to-interference-plus-noise ratio (SINR) among users in an ultra-dense mmWave network environment (UDmN). In this paper, we solve this problem with the cooperative transmission technique to provide high SINR to users. Using coordinated multi-point transmission (CoMP) with the joint transmission (JT) strategy as a cooperation diversity technique can provide users with higher data rates through multiple desired signals. Nonetheless, cooperative transmissions between multiple base stations (BSs) lead to increased energy consumption. Therefore, we propose a multi-agent Q-learning-based power control scheme in UDmN. To satisfy the quality of service (QoS) requirements of users and decrease the energy consumption of networks, we define a reward function while considering the outage and energy efficiency of each BS. The results show that our scheme can achieve optimal transmission power and significantly improved network energy efficiency compared with conventional algorithms such as no transmit power control and random control. Additionally, we validate that leveraging channel state information to determine the participation of each BS in power control contributes to enhanced overall performance.
在超五代网络中，毫米波（mmWave）被认为是一种能够提供高数据速率的有前景的技术。然而，由于小区边界处的小区间干扰，在超密集毫米波网络环境（UDmN）中，用户之间很难实现高信干噪比（SINR）。在本文中，我们通过协作传输技术解决这个问题，为用户提供高SINR。将具有联合传输（JT）策略的协作多点传输（CoMP）用作协作分集技术，可以通过多个期望信号为用户提供更高的数据速率。尽管如此，多个基站（BS）之间的协作传输会导致能耗增加。因此，我们提出了一种基于多智能体Q学习的UDmN功率控制方案。为了满足用户的服务质量（QoS）要求并降低网络能耗，我们在考虑每个基站的中断和能量效率的同时定义了一个奖励函数。结果表明，与无发射功率控制和随机控制等传统算法相比，我们的方案能够实现最优传输功率，并显著提高网络能量效率。此外，我们验证了利用信道状态信息来确定每个基站参与功率控制有助于提高整体性能。
