Abstract:Contrary to conventional massive MIMO cellular configurations plagued by inter-cell interference, cell-free massive MIMO systems distribute network resources across the coverage area, enabling users to connect with multiple access points (APs) and boosting both system capacity and fairness across user. In such systems, one critical functionality is the association between APs and users: determining the optimal association is indeed a combinatorial problem of prohibitive complexity. In this paper, a solution based on deep learning is thus proposed to solve the user clustering problem aimed at maximizing the sum spectral efficiency while controlling the number of active connections. The proposed solution can scale effectively with the number of users, leveraging long short-term memory cells to operate without the need for retraining. Numerical results show the effectiveness of the proposed solution, even in the presence of imperfect channel state information due to pilot contamination.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in the Cell - Free Massive MIMO (CF - mMIMO) system, how to optimize the association between access points (APs) and user equipment (UEs) to maximize the sum spectral efficiency of the system while controlling the number of active connections. Specifically, the paper proposes a deep - learning - based method to solve the user clustering problem, aiming to overcome the high computational complexity and poor scalability of traditional combinatorial optimization methods. ### Problem Background In traditional cellular networks, due to inter - cell interference, especially at the cell edge, the performance is often poor. The CF - mMIMO system can serve a smaller number of user equipment (UEs) under the same time - frequency resources by removing cell boundaries and using a large number of distributed access points (APs) to work together, thus improving system capacity and fairness. However, in this system, determining which access points each user should connect to is a combinatorial optimization problem, and its complexity increases rapidly with the increase of network scale and becomes intractable. ### The Solution Proposed in the Paper To meet this challenge, the paper proposes a deep - learning - based method, especially using the Long - Short - Term Memory network (LSTM) to solve the user clustering problem. The goal of this method is to control the number of active connections while maximizing the total spectral efficiency. The advantage of the LSTM network is that it can be effectively extended to different numbers of users without retraining, thus improving the scalability and flexibility of the system. ### Main Contributions 1. **Optimization Objective**: By introducing a weighting factor \(\lambda\), the spectral efficiency and the number of active connections are balanced, enabling the system to find the best balance between performance and resource utilization. 2. **LSTM Architecture**: Using the recursive characteristics of the LSTM network, it can handle any number of users without retraining the model. 3. **Numerical Results**: The experimental results show that this method can still work effectively even in the presence of imperfect channel state information (such as pilot pollution), and in some cases is superior to traditional methods. ### Summary of Mathematical Formulas - **Spectral Efficiency**: \[ SE_k=\frac{\tau_d}{\tau_c}\log_2(1 + SINR_k)\quad\text{bit/s/Hz} \] where \(\tau_d\) and \(\tau_c\) represent the number of time symbols used for downlink data transmission and the entire coherent block respectively, and \(SINR_k\) is the signal - to - interference - plus - noise ratio of the \(k\)-th user. - **Optimization Problem**: \[ \max_{L_1,\ldots,L_K}\sum_{k\in K}(SE_k-\lambda|L_k|) \] Subject to the constraints: \[ L_k\neq\emptyset\quad\forall k\in K \] \[ \sum_{i\in K_\ell}\rho_{i\ell}\leq\rho_{\max}\quad\forall\ell\in L \] Through these formulas and methods, the paper shows how to achieve efficient user clustering and resource allocation in the complex CF - mMIMO system.

A Deep Learning Approach for User-Centric Clustering in Cell-Free Massive MIMO Systems

Energy-efficient access point clustering and power allocation in cell-free massive MIMO networks: a hierarchical deep reinforcement learning approach

Deep reinforcement learning based joint cooperation clustering and downlink power control for cell-free massive MIMO

Foundations of User-Centric Cell-Free Massive MIMO

Early Acceptance Matching Game for User-Centric Clustering in Scalable Cell-free MIMO Networks

Unsupervised Deep Learning for Power Control of Cell-Free Massive MIMO Systems

Joint User Association and Power Control for Cell-Free Massive MIMO

Joint Cooperative Clustering and Power Control for Energy-Efficient Cell-Free XL-MIMO with Multi-Agent Reinforcement Learning

Deep Learning-based Power Control for Cell-Free Massive MIMO Networks

Deep Learning-Based Approach for User Activity Detection with Grant-Free Random Access in Cell-Free Massive MIMO

A survey on user-centric cell-free massive MIMO systems

Cluster-Based Cell-Free Massive MIMO Systems: A Novel Framework to Enhance Spectral Efficiency with Low Complexity

User-centric Cell-free Massive MIMO Networks: A Survey of Opportunities, Challenges and Solutions

User Association in User-Centric Hybrid VLC/RF Cell-Free Massive MIMO Systems

Performance of Cell-Free Massive MIMO With Joint User Clustering and Access Point Selection

Distributed Combinatorial Optimization of Downlink User Assignment in mmWave Cell-free Massive MIMO Using Graph Neural Networks

Distributed Learning for Uplink Cell-Free Massive MIMO Networks

Downlink Resource Allocation in Multiuser Cell-free MIMO Networks with User-centric Clustering

Cell-Free Massive MIMO versus Small Cells

Energy-Efficient Power Control in Cell-Free and User-Centric Massive MIMO at Millimeter Wave