Joint User Association and Resource Allocation for mmWave Communication:A Neural Network Approach

With the rapid growth of wireless data demand and the shortage of global bandwidth,the use of millimeter-wave(mmWave)frequency band for wireless cellular networks has become the core content of the fifth generation cellular network.Because mmWave communication has different characteristics from microwave communication,using traditional optimization techniques to manage the resource of mmWave communication networks is inappropriate.In this paper,we propose a neural network-based algorithm to solve the joint user association and resource allocation for mmWave communication system with multi-connectivity(MC)and integrated access backhaul(IAB).The resource allocation problem is formulated as a mixed-integer quadratically constrained quadratic programming(MIQCQP),which is very difficult to solve.First,we decompose the MIQCQP into two sub-problems,i.e.,binary associated matrix sub-problem and continuous IAB ratio sub-problem.Then we propose a neural network to solve the binary associated matrix inference problem and a resource allocation algorithm to find the sub-optimal IAB ratio.Simulation results show that the proposed algorithm can achieve good performance with a fast inference speed.

通感一体去蜂窝超大规模MIMO与高频段无线接入技术

随着全息类通信、数字孪生、虚实空间融合互动等新兴应用的快速涌现,未来移动通信系统需要能够支持超大容量无线传输和极高精度定位。基于去蜂窝超大规模MIMO系统的分布式天线阵列和以用户为中心的可伸缩服务架构,进行通信与感知一体化设计,实现高能/谱效传输与带内高精度感知定位,是一种有效的解决思路。首先介绍了未来移动通信系统的潜在技术和应用需求;接着从波束成形与系统性能分析、导频分配与信道估计、回程/前向链路受限下的性能优化等方面介绍了去蜂窝超大规模MIMO系统的研究现状;之后从基本信息理论分析、通信感知一体化接入方法、联合信道估计和位置更新方法、位置信息辅助的分布式波束成形、高能/谱效的分布式资源优化、高频段近场无线接入等方面分析了通感一体去蜂窝超大规模MIMO技术的未来研究方向;最后总结了通感一体去蜂窝超大规模MIMO系统的主要优势和未来仍需关注的问题。