This paper proposes a joint communication and computation scheme (JCCS) for cell-free multiple-input multiple-output (CF-MIMO) network to support federated learning (FL). The JCCS allows users to choose between a model update process or a data offloading process, where the offloaded data and the uploaded model gradient are respectively sent to the distributed processing units (DPUs) deployed on the access points (APs) for further model updates. Moreover, we define a performance metric called iteration error gap as the difference between model errors of adjacent iterations and decouple the total training time minimization problem into the error gap maximization problem within fixed time limit. Base on common machine learning (ML) assumptions, we derive a lower bound of iteration error gap, which is determined by the minimum batch size among all DPUs. An optimization problem aiming to maximize the minimum batch size is then formulated to jointly optimize the time division, power control, and user selection. By employing the block coordinate descent approach, we develop a new algorithm to solve the formulated non-convex mixed integer programming problem. Our simulation results verify the convergence of proposed algorithm and show that it reduces the relative error of a single FL iteration by more than 1.5 dB compared with other baseline schemes. Furthermore, the CF-MIMO network integrated with JCCS achieves a relative error reduction exceeding 1 dB per iteration when compared to collocated MIMO network. In addition, an FL example of handwritten digits classification shows that the JCCS indeed accelerates the convergence of FL model.

Previous studies on federated learning (FL) often encounter performance degradation due to data heterogeneity among different clients. In light of the recent advances in multimodal large language models (MLLMs), such as GPT-4v and LLaVA, which demonstrate their exceptional proficiency in multimodal tasks, such as image captioning and multimodal question answering. We introduce a novel federated learning framework, named Multimodal Large Language Model Assisted Federated Learning (MLLM-LLaVA-FL), which employs powerful MLLMs at the server end to address the heterogeneous and long-tailed challenges. Owing to the advanced cross-modality representation capabilities and the extensive open-vocabulary prior knowledge of MLLMs, our framework is adept at harnessing the extensive, yet previously underexploited, open-source data accessible from websites and powerful server-side computational resources. Hence, the MLLM-LLaVA-FL not only enhances the performance but also avoids increasing the risk of privacy leakage and the computational burden on local devices, distinguishing it from prior methodologies. Our framework has three key stages. Initially, we conduct global visual-text pretraining of the model. This pretraining is facilitated by utilizing the extensive open-source data available online, with the assistance of MLLMs. Subsequently, the pretrained model is distributed among various clients for local training. Finally, once the locally trained models are transmitted back to the server, a global alignment is carried out under the supervision of MLLMs to further enhance the performance. Experimental evaluations on established benchmarks, show that our framework delivers promising performance in the typical scenarios with data heterogeneity and long-tail distribution across different clients in FL.

Federated learning (FL) has attracted increasing attention as a promising technique to drive a vast number of edge devices with artificial intelligence. However, it is very challenging to guarantee the efficiency of a FL system in practice due to the heterogeneous computation resources on different devices. To improve the efficiency of FL systems in the real world, asynchronous FL (AFL) and semi-asynchronous FL (SAFL) methods are proposed such that the server does not need to wait for stragglers. However, existing AFL and SAFL systems suffer from poor accuracy and low efficiency in realistic settings where the data is non-IID distributed across devices and the on-device resources are extremely heterogeneous. In this work, we propose FedSEA - a semi-asynchronous FL framework for extremely heterogeneous devices. We theoretically disclose that the unbalanced aggregation frequency is a root cause of accuracy drop in SAFL. Based on this analysis, we design a training configuration scheduler to balance the aggregation frequency of devices such that the accuracy can be improved. To improve the efficiency of the system in realistic settings where the devices have dynamic on-device resource availability, we design a scheduler that can efficiently predict the arriving time of local updates from devices and adjust the synchronization time point according to the devices' predicted arriving time. We also consider the extremely heterogeneous settings where there exist extremely lagging devices that take hundreds of times as long as the training time of the other devices. In the real world, there might be even some extreme stragglers which are not capable of training the global model. To enable these devices to join in training without impairing the systematic efficiency, Fed-SEA enables these extreme stragglers to conduct local training on much smaller models. Our experiments show that compared with status quo approaches, FedSEA improves the inference accuracy by 44.34% and reduces the systematic time cost and local training time cost by 87.02× and 792.9×. FedSEA also reduces the energy consumption of the devices with extremely limited resources by 752.9×.

In this work, we study the performance of reconfigurable intelligent surface (RIS)-aided communication systems over the sum of cascaded Rician fading channels. To facilitate the performance analysis of the practical scenario, we consider the case that imperfect channel state information (CSI) is available at the RIS. We derive the closed-form expressions of several performance metrics in terms of the exact outage probability, ergodic capacity and average bit error rate (BER). Through analytical and numerical results, we examine the effect of the number of reflecting elements at the RIS and different system parameters on the overall system performance.

Wireless federated learning (FL) holds the potential of preserving data privacy and reducing network traffic congestion, thereby attracting much recent attention. Due to the fading nature of wireless channels, wireless FL suffers from the random delay in each uplink and downlink transmission. As a result, how to analyze the overall random delay of a FL task over wireless fading channels remains open. To solve this challenging problem, we present a saddle point approximation based approach to obtain the distribution of the delay caused by communication in wireless FL systems. In particular, we obtain the uplink delay distribution and the downlink delay distribution by Lugannani-Rice formula. The overall delay distribution is then obtained through the convolution of those two distributions and the generating function. Simulation results demonstrate that the theoretical results provide accurate characterizations for the empirical results, which corroborates the validity of the analysis in this paper.

Federated learning (FL) is a collaborative machine learning paradigm, which enables deep learning model training over a large volume of decentralized data residing in mobile devices without accessing clients’ private data. Driven by the ever increasing demand for model training of mobile applications or devices, a vast majority of FL tasks are implemented over wireless fading channels. Due to the time-varying nature of wireless channels, however, random delay occurs in both the uplink and downlink transmissions of FL. How to analyze the overall time consumption of a wireless FL task, or more specifically, a FL’s delay distribution, becomes a challenging but important open problem, especially for delay-sensitive model training. In this paper, we present a unified framework to calculate the approximate delay distributions of FL over arbitrary fading channels. Specifically, saddle point approximation, extreme value theory (EVT), and large deviation theory (LDT) are jointly exploited to find the approximate delay distribution along with its tail distribution, which characterizes the quality-of-service of a wireless FL system. Simulation results will demonstrate that our approximation method achieves a small approximation error, which vanishes with the increase of training accuracy.

Predicting the popularity of online contents especially videos has drawn a lot of attention recently, since successful prediction of popularity can benefit many practical applications such as recommender systems and proactive caching, and help optimize the advertisement strategies or balance the throughput in the network. In this paper, we formulate a popularity prediction problem and present an Attention-based Long Short Term Memory (LSTM) with Feature Embedding method (ALFE) to tackle the popularity prediction problem. Several features including publish time, follower count, and type of the video are considered and compared. Experiments on a real world dataset show that our method outperforms other competitive baselines from existing works in terms of prediction accuracy. The attention mechanism and feature embedding contribute to the improvement of accuracy. Among all the features, timestamp of popularity and video duration are shown to be the most informative ones, due to the regularity and periodicity of human daily activities.

The coronavirus disease 2019 (COVID-19) has recently attracted extensive attention due to its serious impact on public health worldwide. In this paper, we study and verify that the popularity of virus-related content has a negative correlation with the epidemic spread by means of statistical analysis. Inspired by this result, a practical solution of recommender system is proposed for pushing virus-related content, aiming to gain insight about the newly discovered virus for people and thus reduce the epidemic spread to the utmost extent. First, we formulate the optimization of recommendation policy subject to quality of experience (QoE) loss constraints as a finite-horizon Constrained Markov Decision Problem (CMDP). To solve this problem, then, we present both enumeration and heuristic methods, from perspectives of achieving optimal recommendation policy and reducing computational complexity, respectively. Finally, our simulations validate the benefit of our solution by showing that to recommend virus-related content following our strategy does help slow down the spread of the epidemic.

Ultra-dense network (UDN) is recognized as a key technology for the fifth generation (5G) network to deliver high capacity. Due to cell densification, UDN suffers from heavy intercell interference (ICI). To mitigate ICI, an accurate interference modeling is essential. Compared to distributed mechanism, centralized interference management is more adept at addressing server ICI problem, especially in dense deployments. Hence, we adapt centralized radio access network (C-RAN) architecture to obtain global channel state data. In this paper, a novel regression-based uplink interference identification and signal-to-interference ratio (SINR) prediction algorithm is proposed for C-RAN based 5G UDN. By leveraging big data technology with in-depth analysis of the cause of ICI, this algorithm can precisely model the interference relationship and thus achieve very accurate SINR prediction. In addition, the proposed algorithm is highly efficient in computing, as demonstrated by complexity analysis.

It is expected that Device-to-Device (D2D) communication is allowed to underlay future cellular networks such as IMT-Advanced for spectrum efficiency. However, by reusing the uplink spectrums with the cellular system, the interference to D2D users has to be addressed to maximize the overall system performance. In this paper, a novel method to deal with the resource allocation and interference avoidance issues by utilizing the network peculiarity of a hybrid network to share the uplink resource is proposed and the implementation details are described in a real cellular system. Simulation results prove that satisfying performance can be achieved by using the proposed mechanism.

A hybrid system consisting of a cellular network and a device-to-device (D2D) network is considered in this paper where the D2D users operate in an underlay mode and reuse the spectrums with the cellular users. Most researches focus on reusing the uplink spectrums but how to share the downlink frequency bands is seldom addressed. To share the downlink spectrums and avoid the interference to the primary cellular devices, a labeled time slots based mechanism is proposed and the implementation details are described in a real cellular system. Simulation results prove that satisfying performance can be achieved by using the proposed mechanism.

A hybrid system of cellular mode and the peer-to-peer (P2P) mode is considered in this paper, where the cellular uplink resource is reused by the P2P transmission. In the objective of overall system throughput maximization, we addresses the problem of distributed power control of P2P transmission in the hybrid system model. Some preliminary results of the problem are presented first to investigate the optimal solution of the problem. Based on these results, by applying the game theory approach, that is, potential game, an asynchronously distributed power control scheme is devised. The associated properties of the proposed scheme are analyzed, including the global convergence, the conditional optimality and the Lyapunov stability. In the end, simulation is conducted to study the performance of the proposed scheme, which shows satisfying results.

A hybrid system of cellular mode and device-to-device (D2D) mode is considered in this paper, where the cellular uplink resource is reused by the D2D transmission. In order to maximize the overall system performance, the mutual interference between cellular and D2D sub-systems has to be addressed. Here, two mechanisms are proposed to solve the problem: One is mitigating the interference from cellular transmission to D2D communication by an interference tracing approach. The other one is aiming to reduce the interference from D2D transmission to cellular communication by a tolerable interference broadcasting approach. Both mechanisms can work independently or jointly to synergy the transmission in the hybrid system for the efficient resource utilization. In the end, simulation is conducted to study the performance of the proposed schemes, which shows satisfying results.

In cognitive radio networks, the limitation of control channel bandwidth is a challenge of cooperative spectrum sensing when the number of cognitive users becomes very large. Cluster- based architecture is applied for cooperative sensing to avoid the congestion on control channel and reduce the sensing delay. In this paper, we propose a cluster-based cooperative spectrum sensing scheme to improve the efficiency of the network. The number of clusters effects both the system efficiency and detection performance significantly. By balancing the tradeoff between the communication overhead and sensing reliability, we can obtain the optimal number of clusters, which can minimize the cooperation overhead without any performance loss of reliability. Moreover, a clustering strategy is proposed based on a given number of clusters and simulation results show the superiority of the proposed strategy.