Joint Antenna and User Scheduling in the Massive MIMO System Over Time-Varying Fading Channels

Abstract

Massive multiple-input multiple-output (MIMO) technology, mainly equipped with dozens or even hundreds of antennas at transmitter and/or receiver, is one of the most important technologies in 5G/6G era due to its capability of achieving high transmission rate. However, since each transmit antenna typically needs a complete radio frequency (RF) chain, a large number of RF chains need to be installed accordingly, resulting in high economic cost, high hardware complexity, and high power consumption. To resolve these problems, architectures where a small number of RF chains are installed have been proposed in recent years, and an antenna selection technique that activates antennas only as many as the number of RF chains has been envisioned as one of solutions. In this paper, we study the joint antenna and user scheduling problem for the downlink massive MIMO system over time-varying fading channels to maximize the weighted average sum rate while ensuring users' minimum average data rate requirements. To solve the problem, we first develop an opportunistic joint antenna and user scheduling algorithm (OJAUS) using the dual and stochastic subgradient methods, which makes it possible to schedule antennas and users without any underlying distributions of the fading channels. However, it requires solving a joint antenna and user selection (JAUS) problem to maximize the instantaneous weighted sum rate in every time slot. Thus, we additionally develop a simple heuristic JAUS algorithm with low computational complexity, called JAUS-LCC, which is executed in every time slot within OJAUS. Finally, through simulation results, we first show that our JAUS-LCC provides near-optimal performance despite requiring very low computational complexity, and then show that our OJAUS with JAUS-LCC well guarantees given minimum average data rate requirements.INDEX TERMS Antenna selection, multiple-input multiple-output (MIMO), quality-of-service (QoS), timevarying fading channels, user scheduling, weighted sum rate maximization, zero-forcing (ZF) precoding.

A DRL-Based Automated Algorithm Selection Framework for Cross-Layer QoS-Aware Scheduling and Antenna Allocation in Massive MIMO Systems

“…We consider a single-cell massive MIMO system consisting of an M -antenna BS, K single-antenna UEs, and cross-layer controls [8]. Thus we have m ∈ M = {1, 2, .…”

“…Then, the AS precoder, which matches the greedy-based UBLAA algorithm, is applied. • LWDF-JAUS: is a combination of LWDF-PF [61] QoS scheduling and JAUS-LCC [8] massive MIMO antenna allocation algorithms. LWDF-PF is a landmark proportional fair algorithm that adopts weighted delay fairness for UE scheduling, while JAUS-LCC determines the number of antennas used for receiving UEs.…”

“…The work considered only scalable video streaming. Zhu et al [8] propose a joint antenna and user scheduling solution with antenna/user deletion sub-problem solved by a low complexity rule-based algorithm named JAUS-LCC for massive MIMO scenarios. Though the sub-problem can be solved effectively, the overall QoS-aware algorithm is not as scalable.…”

“…In summary, current cross-layer resource allocation works appear not fully utilize machine learning for decision-making, while massive MIMO-focused works consider less on the application scenarios and QoS. To advance the state-of-the-art, i.e., [8], [12], we proposed to deal with the open problem of coordinating the interaction of cross-layer resource allocation algorithms in a MIMO system with QoS in mind.…”