The explosive growth of data results in more scarce spectrum resources. It is important to optimize the system performance under limited resources. In this paper, we investigate how to achieve weighted throughput (WTP) maximization for cell-free (CF) multiuser MIMO (MU-MIMO) multicarrier (MC) systems through resource allocation (RA), in the cases of finite blocklength (FBL) and infinite blocklength (INFBL) regimes. To ensure the quality of service (QoS) of each user, particularly for the block error rate (BLER) and latency in the FBL regime, the WTP gets maximized under the constraints of total power consumption and required QoS metrics. Since the channels vary in different subcarriers (SCs) and inter-user interference strengths, the WTP can be maximized by scheduling the best users in each time-frequency (TF) resource and advanced beamforming design, while the resources can be fully utilized. With this motivation, we propose a joint user scheduling (US) and beamforming design algorithm based on the successive convex approximation (SCA) and gene-aided (GA) algorithms, to address a mixed integer nonlinear programming (MINLP) problem. Numerical results demonstrate that the proposed RA outperforms the comparison schemes. And the CF system in our scenario is capable of achieving higher spectral efficiency than the centralized antenna systems (CAS).

翻译：数据爆炸式增长导致频谱资源日益稀缺，在有限资源下优化系统性能至关重要。本文研究了在有限块长（FBL）和无限块长（INFBL）场景下，如何通过资源分配（RA）实现无小区（CF）多用户MIMO（MU-MIMO）多载波（MC）系统的加权吞吐量（WTP）最大化。为保障每个用户的服务质量（QoS），特别是FBL场景下的块错误率（BLER）与延迟要求，本文在总功耗与所需QoS指标的约束下最大化WTP。由于不同子载波（SC）的信道特性及用户间干扰强度存在差异，通过在每个时频（TF）资源上调度最优用户并设计先进的波束成形方案，可最大化WTP并充分资源利用。基于此动机，我们提出一种联合用户调度（US）与波束成形设计的算法，该算法融合逐次凸近似（SCA）与基因辅助（GA）方法，以解决混合整数非线性规划（MINLP）问题。数值结果表明，所提RA方案性能优于对比方案；且本文场景下的CF系统相比集中式天线系统（CAS）可实现更高的频谱效率。