Abstract With the growing demand for seamless connectivity and intelligent communication, the integration of artificial intelligence (AI) and sixth-generation (6G) communication networks has emerged as a transformative paradigm. By embedding AI capabilities across various network layers, this integration enables optimized resource allocation, improved efficiency, and enhanced system robust performance. This paper presents a comprehensive overview of AI and communication for 6G networks, with a focus on their foundational principles, inherent challenges, and future research opportunities. We first review the integration of AI and communications in the context of 6G, exploring the driving factors behind incorporating AI into wireless communications, as well as the vision for the convergence of AI and 6G. The discourse then transitions to a detailed exposition of the envisioned integration of AI within 6G networks, divided into three progressive stages. The first stage, AI for network, focuses on employing AI to augment network performance, optimize efficiency, and enhance user service experiences. The second stage, network for AI, highlights the role of the network in facilitating and buttressing AI operations and presents key enabling technologies. We compare wireless network large models with conventional large language models (LLMs), and identify key design principles and components for building wireless network architectures. In the final stage, AI as a service, it is anticipated that future 6G networks will innately provide AI functions as services, supporting application scenarios like immersive communication and intelligent industrial robots. Specifically, we define the quality of AI service, which refers to a framework for measuring AI services within the network. We further summarize the standardization process of AI for wireless networks, highlighting key milestones and ongoing efforts. In addition, we analyze the critical challenges faced by the integration of AI and communications in 6G. Finally, we outline promising future research opportunities that are expected to drive the development and refinement of AI and 6G communications.

The integration of artificial intelligence (AI) and mobile networks is regarded as one of the most important scenarios for 6G. In 6G, a major objective is to realize the efficient transmission of task-relevant data. Then a key problem arises, how to design collaborative AI models for the device side and the network side, so that the transmitted data between the device and the network is efficient enough, which means the transmission overhead is low but the AI task result is accurate. In this paper, we propose the multi-link information bottleneck (ML-IB) scheme for such collaborative models design. We formulate our problem based on a novel performance metric, which can evaluate both task accuracy and transmission overhead. Then we introduce a quantizer that is adjustable in the quantization bit depth, amplitudes, and breakpoints. Given the infeasibility of calculating our proposed metric on high-dimensional data, we establish a variational upper bound for this metric. However, due to the incorporation of quantization, the closed form of the variational upper bound remains uncomputable. Hence, we employ the Log-Sum Inequality to derive an approximation and provide a theoretical guarantee. Based on this, we devise the quantized multi-link information bottleneck (QML-IB) algorithm for collaborative AI models generation. Finally, numerical experiments demonstrate the superior performance of our QML-IB algorithm compared to the state-of-the-art algorithm.

Federated Learning (FL) is a promising distributed learning mechanism which still faces two major challenges, namely privacy breaches and system efficiency. In this work, we reconceptualize the FL system from the perspective of network information theory, and formulate an original FL communication framework, FedNC, which is inspired by Network Coding (NC). The main idea of FedNC is mixing the information of the local models by making random linear combinations of the original parameters, before uploading for further aggregation. Due to the benefits of the coding scheme, both theoretical and experimental analysis indicate that FedNC improves the performance of traditional FL in several important ways, including security, efficiency, and robustness. To the best of our knowledge, this is the first framework where NC is introduced in FL. As FL continues to evolve within practical network frameworks, more variants can be further designed based on FedNC.

Federated Learning (FL) is a promising distributed learning mechanism that revolutionizes our interaction with data in the IoT ecosystem. Due to the rapidly growing scale of smart devices and the limited transmission resources of networks, a simple, consistent and scalable FL framework aiming to address the communication bottleneck is urgently needed. In this work, we propose an efficient approach with Selective Aggregation of Models (SAM) to mitigate the communication overload in FL systems. The introduction of SAM enables each local client to upload its model with a certain probability, resulting in a significant reduction in costly communication expenses. We design the algorithm for SAM, analyze the convergence bound on non-convex objectives for heterogeneous data, which illustrates the impact of the selection probability as well as the set size of participating clients on the system performance, and assess the conservation for the network resource utilization by modeling queuing systems. We conduct various experiments to evaluate the performance of SAM, whose outcomes suggest that significant alleviation of the communication bottleneck can be accomplished with marginal cost of performance loss. It will also be shown that SAM is a communication-efficient method that can be freely applied to other frameworks.

The wide applications of artificial intelligence (AI) on massive Internet-of-things or smartphones raises significant concerns about privacy, heterogeneity, and resource efficiency. Correspondingly, federated learning emerges as an effective way to enable AI over massively distributed nodes without uploading the raw data. Conventional works mostly focus on learning a single unified model for one solitary task. Multi-task learning (MTL) outperforms single-task learning by training multiple models concurrently, leading to reduced model sizes and increased flexibility. However, existing federated learning efforts often face challenges in efficiently managing MTL scenarios, particularly with the presence of stragglers, without incurring prohibitive computation and communication costs. In this paper, inspired by the natural cloud-BS-terminal hierarchy of cellular networks, we provide a viable resource-aware hierarchical federated MTL (RHFedMTL) solution to meet the task heterogeneity corresponding to different non-IID (independent and identically distributed) training datasets. Specifically, a primal-dual method has been leveraged to effectively transform the coupled MTL into some local optimization sub-problems within BSs. Therefore, it enables solving different tasks within a BS and aggregating the multi-task result in the cloud without uploading the raw data. Furthermore, compared with existing methods that reduce resource costs by simply changing the aggregation frequency, we dive into the intricate relationship between resource consumption and learning accuracy, and develop a resource-aware learning strategy for adjusting the iteration number on local terminals and BSs to meet the resource budget. Extensive simulation results demonstrate the effectiveness and superiority of RHFedMTL in terms of improving the learning accuracy and boosting the convergence rate.

As a promising learning paradigm integrating computation and communication, federated learning (FL) proceeds the local training and the periodic sharing from distributed clients. Due to the non-i.i.d. data distribution on clients, FL model suffers from the gradient diversity, poor performance, bad convergence, etc. In this work, we aim to tackle this key issue by adopting importance sampling (IS) for local training. We propose importance sampling federated learning (ISFL), an explicit framework with theoretical guarantees. Firstly, we derive the convergence theorem of ISFL to involve the effects of local importance sampling. Then, we formulate the problem of selecting optimal IS weights and obtain the theoretical solutions. We also employ a water-filling method to calculate the IS weights and develop the ISFL algorithms. The experimental results on CIFAR-10 fit the proposed theorems well and verify that ISFL reaps better performance, convergence, sampling efficiency, as well as explainability on non-i.i.d. data. To the best of our knowledge, ISFL is the first non-i.i.d. FL solution from the local sampling aspect which exhibits theoretical compatibility with neural network models. Furthermore, as a local sampling approach, ISFL can be easily migrated into other emerging FL frameworks.

Channel capacity estimation plays a crucial role in beyond 5G intelligent communications. Despite its significance, this task is challenging for a majority of channels, especially for the complex channels not modeled as the well-known typical ones. Recently, neural networks have been used in mutual information estimation and optimization. They are particularly considered as efficient tools for learning channel capacity. In this paper, we propose a cooperative framework to simultaneously estimate channel capacity and design the optimal codebook. First, we will leverage MIM-based GAN, a novel form of generative adversarial network (GAN) using message importance measure (MIM) as the information distance, into mutual information estimation, and develop a novel method, named MIM-based mutual information estimator (MMIE). Then, we design a generalized cooperative framework for channel capacity learning, in which a generator is regarded as an encoder producing the channel input, while a discriminator is the mutual information estimator that assesses the performance of the generator. Through the adversarial training, the generator automatically learns the optimal codebook and the discriminator estimates the channel capacity. Numerical experiments will demonstrate that compared with several conventional estimators, the MMIE achieves state-of-the-art performance in terms of accuracy and stability.

Multi-Agent Reinforcement Learning (MARL) has become a classic paradigm to solve diverse, intelligent control tasks like autonomous driving in Internet of Vehicles (IoV). However, the widely assumed existence of a central node to implement centralized federated learning-assisted MARL might be impractical in highly dynamic scenarios, and the excessive communication overheads possibly overwhelm the IoV system. Therefore, in this paper, we design a communication efficient cooperative MARL algorithm, named RSM-MAPPO, to reduce the communication overheads in a fully distributed architecture. In particular, RSM-MAPPO enhances the multi-agent Proximal Policy Optimization (PPO) by incorporating the idea of segment mixture and augmenting multiple model replicas from received neighboring policy segments. Afterwards, RSM-MAPPO adopts a theory-guided metric to regulate the selection of contributive replicas to guarantee the policy improvement. Finally, extensive simulations in a mixed-autonomy traffic control scenario verify the effectiveness of the RSM-MAPPO algorithm.

Federated learning (FL) has prevailed as an efficient and privacy-preserved scheme for distributed learning. In this work, we mainly focus on the optimization of computation and communication in FL from a view of pruning. By adopting layer-wise pruning in local training and federated updating, we formulate an explicit FL pruning framework, FedLP (Federated Layer-wise Pruning), which is model-agnostic and universal for different types of deep learning models. Two specific schemes of FedLP are designed for scenarios with homogeneous local models and heterogeneous ones. Both theoretical and experimental evaluations are developed to verify that FedLP relieves the system bottlenecks of communication and computation with marginal performance decay. To the best of our knowledge, FedLP is the first framework that formally introduces the layer-wise pruning into FL. Within the scope of federated learning, more variants and combinations can be further designed based on FedLP.

Multi-Agent Reinforcement Learning (MARL) has reached astonishing achievements in various fields such as the traffic control of vehicles in a wireless connected environment. In MARL, how to effectively decompose a global feedback into the relative contributions of individual agents belongs to one of the most fundamental problems. However, the volatility of the environment (e.g., the vehicle movement and wireless disturbance) could significantly shape the time-varying topological relationships among agents, thus making the Value Decomposition (VD) challenging. Therefore, in order to cope with this annoying volatility, it becomes imperative to design a dynamic VD framework. Hence, in this paper, we propose a novel Stochastic VMIX (SVMIX) methodology by embedding the dynamic topological features into the VD and incorporating the corresponding components into a multi-agent actor-critic architecture. In particular, the Stochastic Graph Neural Network (SGNN) is leveraged to effectively extract underlying dynamics embedded in topological features and improve the flexibility of VD against the environment volatility. Finally, the superiority of SVMIX is verified through extensive simulations.

6G mobile communication system is expected to be intelligence inclusive, that is, its system architecture would be designed to provide AI services to anyone, anywhere, including the network system itself. This mandates a clean-slate approach to its design. In this paper, this broad question is addressed from the perspective of providing distributed learning services, more specifically, using the Federated Learning (FL) paradigm. The relationship between two network metrics and the FL performance in a hierarchical federated learning system is explored, in order to determine the new features that the network architecture needs to support. An estimation of the dependency of bandwidth requirements on different ML models is also provided.

Federated learning (FL) has recently emerged as an innovative paradigm to train models among distributed agents. Conventional FL considers the center as an aggregator and trains from distributed data, while the collected global information at the center is not effectively utilized. Thus, the restricted information from local observations may limit the model accuracy. If FL can introduce data sets from the network server, the distributed models may be largely improved by the extra global information. Since network agents may not be completely trusted, the center cannot directly broadcast its raw data for security concern. Then, how to combine the central sets with FL? In this article, we propose to add a learning model at the center, which obtains the central sets as input. The outputs can be transmitted to network agents and integrated into local models instead of the raw data. The central and local models could be trained to form an integration for intelligent inference. Then, what is the integrated performance gain comparing with the original horizontal FL (HFL) and how to implement it? To figure out these two problems, we propose the vertical-HFL (VHFL) scheme, where models of the center and agents are trained collaboratively. We further analyze its convergence and the related communication channel, proposing the theoretical bounds to guide the network implementation of VHFL. Some simulation results will demonstrate the effectiveness of our proposed VHFL scheme. It is expected that VHFL will be an important block for the next generation of smart services.

Federated learning (FL) has recently emerged as a transformative paradigm that jointly train a model with distributed devices while avoiding the need for central data collection. Due to the limited observation range, the devices only contain local information, which limits the quality of trained models. In this case, combining the global information into FL may be helpful. However, in horizontal FL, the central agency only acts as a model aggregator without utilizing its global observation. Meanwhile, the global data may not be directly transmitted to agents for data security. Then how to utilize the global observation residing in the central agency while protecting its safety thus rises up as an important problem in FL. In this paper, we develop a vertical-horizontal federated learning (VHFL) scheme, where the global feature is shared with the agents in a procedure similar to that of vertical FL. It is shown by experiments that the proposed VHFL could enhance the accuracy compared with horizontal FL while protecting the central data from being announced.

Federated learning is an effective way to enable artificial intelligence over massive distributed nodes with security and communication efficiency. Some previous works primarily focus on learning a single global model for a unique task across the network, which is less competent to handle multi-task scenarios with stragglers and fault, after adopting the general gradient update methods in a federated environment. Others aim to learn a distinct model for each node, which is expensive in terms of the computation and communication cost. Using hierarchical network to reduce communication cost is becoming a new candidate. Thus, we propose a primal-and-dual method-based hierarchical federated multi-task learning system, supported with HFedMTL algorithm that allows massive nodes from distributed areas to join in the federated multi-task learning process. Empirical experiments verify the analysis and demonstrate the benefits of improving the learning performance and convergence rate.

Recently, federated learning (FL) has been treated as an important and promising learning scheme in IoT, enabling devices to jointly learn a model without sharing their data sets. Different from centralized training on some collected data sets, FL training suffers a lot of constraints from limited resources in the network. Therein, the bandwidth and package loss restrict interactions in training. Meanwhile, the highly distributed data sets and limited computation could also affect its convergence. To figure out the specific impact, we analyze the convergence rate of FL training considering both communication and training. Further taking in training costs in terms of time and power, the closed-form optimal settings for communication networks are proposed with principles to assist the parameter selection. The results build a bridge between AI and communication, giving us an intuitive knowledge of how the background system could influence the distributed training process.

Federated learning (FL) has recently emerged as an important and promising learning scheme in IoT, enabling devices to jointly learn a model without sharing their raw data sets. As FL does not collect and store the data centrally, it requires frequent model exchange through the wireless network. However, since the aggregation in FL can be partially participated with synchronized frequency, its communication pattern is different from the conventional network. Therein, limited bandwidth and package loss restrict interactions in training. Thus, the network scheduling could largely affect the FL convergence. To figure out the specific effects, we analyze the convergence rate of FL regarding the joint impact of communication and training. Combining it with the network model, we formulate the optimal scheduling problem for FL implementation. The theoretical results could guide the hyper-parameter design in the network and explain the principle of how the wireless communication could influence the FL training process.

This paper discusses a resource management abstraction framework which involves the clustering of densely deployed access points and the selection of access nodes to act as cluster schedulers based on the RAN characteristics, the resource situation and the network slice requirements. To this end, the adaptive placement of Radio Resource Management (RRM) functionalities to the RAN nodes and the interactions among functionalities are determined on per slice basis. By taking into account the slice requirements, the backhaul#x002F;access channel conditions and the traffic load, a central management entity assigns RRM functionalities to the controllers with different levels of centralization in order to meet the per slice SLAs (in terms of throughput, reliability, latency). The controller, cluster and RRM split configuration problems are formulated and interpreted as three dependent graph-based sub- problems, where heuristic approaches with low complexity and signalling cost are proposed within this framework.

This paper describes a new protocol SMP, which supports multimedia transfer for single-user, multiple-device scenarios. Through its novel naming and control/data plane designs, SMP is able to retain the current client and server protocol operations while placing new functions at the proxy. Our initial evaluation has confirmed its viability.

Resolution from identifiers to locators serves as a key component of mapping-based mobility solutions. In this paper we address the weakness of current resolution methods in supporting diverse mobility scenarios and propose a Global Resolution Service offered by Resolution Service Providers. We present a preliminary design and simulation, and results show that our approach is able to provide better resolution service compared to existing solutions in terms of performance.