Large Language Models (LLMs), with advanced content creation and inference capabilities, can provide immersive intelligent services to users in mobile edge networks. However, the increasing demand for real-time artificial intelligence (AI) applications aggravates the limitations of cloud-based LLMs due to the long response time. Meanwhile, Small Language Models (SLMs), which are cost-effective and locally deployable for terminal devices, can serve as an efficient supplement to LLMs for performing latency-sensitive tasks with lower generalization capability. Due to the resource constraints of edge networks and the diverse requirements of user tasks, it is critical to design an inference framework that effectively coordinates the deployment and collaboration of LLMs and SLMs. In this paper, we propose an LLM-SLM collaborative inference (LSCI) scheme under a mobile edge computing (MEC) architecture, which jointly decides where to cache models and how to offload inference tasks to balance latency, accuracy, and resource costs. To optimize inference performance subject to resource constraints, we jointly solve the inference task offloading and model caching problem in LSCI scheme. Specifically, we employ deep reinforcement learning (DRL) to select highly popular SLMs to be cached on the edge server, and distributed belief propagation technique to solve the associated inference task offloading issue. Numerical results show that the proposed LSCI scheme can achieve significant performance gain in terms of inference performance when compared with a number of baseline solutions.

Hierarchical clustering learning (HCL) extends traditional parameter server-based distributed learning by clustering heterogeneous user equipments (UEs) via cluster nodes (CN s) located at the edge of the network. Currently, most vanilla model dissemination strategies in distributed learning rely on one-to-many transmissions, inevitably consuming excessive precious bandwidth resources. Consequently, communication-efficiency becomes crucial for HCL in resource-constrained edge networks. In this paper, we propose a multistage cooperative model dissemination strategy to sequentially determine the subsets of CNs that can concurrently transmit models during individual scheduling stages, thereby improving communication efficiency in HCL. First, we formulate the strategy design as an optimization problem to minimize the maximum completion time of the slowest straggler in a communication round of HCL. Then, we design an online learning algorithm, called sequential combinatorial multiarmed bandit (SCMAB) to make sequential and combinatorial decisions in individual stages. Numerical results demonstrate the superiority of our proposed strategy over some benchmarks in terms of communication efficiency.

High frequency directional communication is considered as a key technology to improve the performance of mobile Ad Hoc networks owing to its advantages in terms of communication distance and interference. Neighbor discovery plays a key role for efficient routing and topology control in mobile Ad Hoc networks. It is also a very challenging issue due to the use of directional antennas and movement of mobile nodes. Beam configuration is the key step of neighbor discovery in mobile Ad Hoc networks with directional antennas. Therefore, it is imperative to develop an efficient beam configuration algorithm to reduce the neighbor discovery latency. In this paper, we propose a novel beam configuration algorithm based on personalized federated learning. Considering the characteristics of mobile Ad Hoc networks (e.g., dynamic topology and directional communication), we use Deep Deterministic Policy Gradient (DDPG) as the local model of federated learning. Since the local data of ad hoc network nodes is heterogeneous, Model Agnostic Meta Learning (MAML) is applied to personalize the federated learning. Numerical results demonstrate that our proposed algorithm has better performance than some baseline algorithms.

Proactive reconfiguration of network slices according to uncertain traffic demands is essential to improve network resource utilization while ensuring service quality in 5G-and-beyond systems. Existing researches on network slice reconfiguration are either model-driven or data-driven methods. However, model-driven methods may cause resource over-provisioning due to a lack of prediction mechanism, while data-driven methods are unrealistic in inter-slice reconfiguration that involves costly and time-consuming operations such as VNF migration. To address these issues, in this paper, we propose a Hybrid Model-Data driven (HMD) framework that intelligently performs inter-slice reconfiguration by leveraging prediction interval and robust optimization. We design a Prediction Interval-oriented Predictor (PIP) to produce a prediction interval that can bracket the future traffic demand with a prespecified probability. Based on the prediction interval, we design an inter-slice reconfiguration scheme (named box optimizer) to perform fast inter-slice reconfigurations. To tackle the over-conservativeness of the box optimizer, we further design the ellipsoid optimizer with better optimality at a cost of increased complexity. Numerical results demonstrate that the proposed framework can provide high robustness with low power consumption. Meanwhile, the trade-off between the power consumption and the realized robustness can be flexibly adjusted according to the type of slice and the level of traffic demand fluctuations.

Millimeter wave (mmWave) communication is one of the key technologies in 5G and beyond systems to address the tremendous growth in mobile data traffic owing to the abundant spectrum resources. Ultra-dense network deployment is a promising solution to combat the limited coverage, high propagation loss and attenuation of mmWave signals. This study investigates the beam management, with focus on beam configuration of mmWave base stations, in the ultra-dense mmWave network. To fulfill adaptive and intelligent beam management while protecting user privacy, we employ a double deep Q-network under a federated learning to tackle the beam management problem which is formulated to maximize the long-term system throughput. Simulation results demonstrate the performance gain of our proposed scheme.

Network Slicing (NS) has been widely identified as a key architectural technology for 5G-and-beyond systems by supporting divergent requirements sustainably. With the widespread of emerging smart devices, access control becomes an essential yet challenging issue in NS-based wireless networks due to the device-base station (BS)-NS three-layer association relationship. Meanwhile, stringent data security and device privacy concerns are increasing dramatically. In this paper, we propose an efficient access control scheme for radio access network (RAN) slicing by exploiting a federated deep reinforcement learning framework, called FDRL-AC, to improve network throughput and communication efficiency while enforcing the data security and device privacy. Specifically, we use deep reinforcement learning to train local model on devices, where horizontally federated learning (FL) is employed for parameter aggregation on BS, while vertically FL is employed for feature aggregation on the encrypted party. Numerical results show that the proposed FDRL-AC scheme can achieve significant performance gain in terms of network throughput and communication efficiency in comparison with some state-of-art solutions.

Radio Access Network (RAN) slicing is a promising architectural technology to address extremely diversified service demands for future mobile networks. As an essential requirement for RAN slicing, self-healing is to provide services with certain quality requirements by minimizing the impact of mobile network failings. In this paper, we propose a Multi-objective Pareto Optimization based Self-healing (MPOS) scheme to solve the SRANS problem. We model the SRANS problem as a multi-objective optimization problem with aim of maximizing the self-healing profits of individual RAN slices and demonstrate the NP-hardness. In proposed MPOS scheme, we employ self-conditioned GANs to replace the offspring reproduction module in the traditional Multi-Objective Evolutionary Algorithm (MOEA), where the insufficiency of diversity maintenance in MOEA is effectively overcome. Furthermore, we theoretically prove that MPOS framework is guaranteed to converge to the optimal Pareto solution set with probability 1. Numerical results demonstrate that our MPOS scheme is effective in reducing the inverted generational distance of optimal Pareto solutions and achieving high profit and isolation level of RAN slices.

The beyond fifth-generation (B5G) and future wireless networks face the challenges of spectral, energy and cost efficiency for large scale machine-type communications. Recently, emerging ambient backscatter communication (AmBC) technology provides a promising paradigm for the development of green Internet of Things (IoT) networks in beyond B5G era. In this paper, we consider a multi-nodes scenario where a backscatter network is symbiotic with primary network consisting of multiple ambient radio frequency (RF) sources, thereby allowing the system to use the appropriate RF to support high throughput and wide coverage for IoT devices. Unlike existing work on AmBC, which focuses on physical layer with relatively ideal model, i.e., classic three-nodes model composed of RF, backscatter device (BD) and IoT device, this paper studies the access control strategy, including coefficient design and device association of BDs and IoT devices, for multi-RF backscatter network from the perspective of maximizing device transmission rate. Under the guarantee of quality of service (QoS), we develop an access control strategy with aim of maximizing the weighted sum of primary and backscatter transmission rates, and design a distributed access control strategy called DCA-S, by using the difference of two convex functions approximation (DCA) and dual decomposition to transform the non-concave optimization problem into the solvable concave subproblems. Numerical results show that the proposed DCA-S can achieve significantly performance improvement of the system compared with benchmark schemes.