Federated Learning (FL) enables distributed model training across edge devices in a privacy-friendly manner. However, its efficiency heavily depends on effective device selection and high-dimensional resource allocation in dynamic and heterogeneous wireless environments. Conventional methods demand a confluence of domain-specific expertise, extensive hyperparameter tuning, and/or heavy interaction cost. This paper proposes a Tool-aided Evolutionary Large Language Model (T-ELLM) framework to generate a qualified policy for device selection in a wireless FL environment. Unlike conventional optimization methods, T-ELLM leverages natural language-based scenario prompts to enhance generalization across varying network conditions. The framework decouples the joint optimization problem mathematically, enabling tractable learning of device selection policies while delegating resource allocation to convex optimization tools. To facilitate the evolutionary process, T-ELLM interacts with a sample-efficient, model-based virtual learning environment that captures the relationship between device selection and learning performance. This developed virtual environment reduces reliance on real-world interactions, thus minimizing communication overhead while refining the LLM-based decision-making policy through group relative policy optimization. Theoretical analysis proves that the discrepancy between virtual and real environments is bounded, ensuring the advantage function learned in the virtual environment maintains a provably small deviation from real-world conditions. Experimental results demonstrate that T-ELLM outperforms benchmark methods in energy efficiency and exhibits robust adaptability to environmental changes.

In distributed edge networks employing frequency division duplex, the feedback of channel state information (CSI) from the edge devices to the edge server always consumes a lot of spectrum resources, resulting in a serious communication burden. In this paper, we first propose an end-to-end unfolding neural network framework inspired by the soft threshold iterative algorithm (U-ISTANet). The proposed U-ISTANet integrates the advantages of compression awareness and neural networks. Especially, the compression matrix and sparse transformation of channel matrix can be learned for accurate CSI compression and recovery. And a lightweight version of U-ISTANet, called U-ISTANet-L, is proposed to reduce the training parameters. To reduce the data transmission overhead in the centralized learning framework, we extend the proposed U-ISTANet-L to a federated U-ISTANet-L (FU-ISTANet-L), which can train a more generalizable model by increasing the number of edge devices to enlarge the data set in a distributed learning manner. The proposed FU-ISTANet-L reduces the transmission overhead and increases the training speed while achieving a performance close to that of centralized learning. Furthermore, we propose a personalized FU-ISTANet-L (P-FU-ISTANet-L) to solve the heterogeneous data training problem in different communication environments. Specifically, we first obtain a pre-trained model by federation unfolding learning, and then each edge device fine-tunes the model using only a small amount of train data to obtain a personalized model for local channel environment. Extensive experimental results are provided to show that the proposed networks achieve a significant performance over the benchmarking schemes in terms of the normalized mean square error.

Spatial non-stationarity channel estimation for uplink massive MIMO systems can be formulated as a non-uniform block sparse signal recovery problem, in which the hierarchical sparsity of the channel matrix is classified as row sparsity and in-row sparsity. This paper proposes an efficient threshold-enhanced hierarchical estimation (TEHE) algorithm without prior information. More precisely, the non-zero rows of spatial non-stationarity channel matrix are estimated according to the in-row correlation in the first layer; while the non-zero elements of the estimated non-zero rows are further refined in the second layer. Different from the existing two-layer iteration algorithms, an adaptive threshold is designed to estimate the non-zero elements replacing the iterative algorithm in the second layer. In the proposed TEHE algorithm, row-wise sparse adaptive matching pursuit (SAMP) is used to find the non-zero rows in the first layer, which has high precision and lower complexity, compared to the conventional SAMP. To further improve the efficiency of the row estimation for larger antenna array, an adaptive threshold-enhanced hierarchical estimation (A-TEHE) algorithm is proposed. In addition, a sufficient condition and a halting condition for theoretical guarantee to obtain accurate row estimation are developed. Finally, the computation complexity is analyzed and compared. The simulation results demonstrate that the proposed threshold-enhanced hierarchical spatial non-stationary channel estimation algorithms achieve better performance compared to various state-of-the-art baselines in terms of support set estimation, channel coefficient estimation, and computational efficiency. Specifically, the proposed algorithms are robust to the in-row sparsity.

This paper proposes a hierarchical sparse estimation of spatial non-stationarity channel for uplink massive multiple-input multiple-output (MIMO) systems without prior information. Especially, the non-zero rows of non-stationarity channel matrix are estimated according to the in-row correlation in the first layer; while the non-zero elements of the estimated non-zero rows are further refined in the second layer. A row-wise sparse adaptive matching pursuit (SAMP) is used to find the non-zero rows in the first layer of the proposed algorithms, and multiple non-zero rows can be estimated in one iteration, which has higher precision and lower complexity, compared to the conventional SAMP. Different from the existing two-layer iteration algorithms, a threshold is designed to estimate the non-zero elements replacing the iterative algorithm in the second layer. Further, the computation complexity is analyzed and compared. The simulation results demonstrate that the proposed threshold-enhanced hierarchical spatial non-stationary channel estimation algorithms achieve better performance compared to various state-of-the-art baselines in terms of channel coefficient estimation, and computational efficiency.

MOTIVATION: Subspecies identification is one of the most critical issues in microbiome studies, as it is directly related to their functions in response to the environmental stress and their feedbacks. However, identification of subspecies remains a challenge largely due to the small variation between different strains within the same species. Accurate identification of subspecies primarily relies on variant identification and categorization through microbiome data. However, current SNP calling and subspecies identification for microbiome data remain underdeveloped. RESULTS: In this work, we have proposed Strain-GeMS for subspecies identification from microbiome data, based on solid statistical model for SNP calling, as well as optimized procedure for subspecies identification. Results on simulated, ab initio and in vivo datasets have shown that Strain-GeMS could always generate more accurate results compared with other subspecies identification methods. AVAILABILITY AND IMPLEMENTATION: Strain-GeMS is available at: https://github.com/HUST-NingKang-Lab/straingems. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.