Hierarchical Content Caching in Fog Radio Access Networks:Ergodic Rate and Transmit Latency

In order to alleviate capacity constraints on the fronthaul and decrease the transmit latency, a hierarchical content caching paradigm is applied in the fog radio access networks(F-RANs). In particular, a specific cluster of remote radio heads is formed through a common centralized cloud at the baseband unit pool, while the local content is directly delivered at fog access points with edge cache and distributed radio signal processing capability. Focusing on a downlink F-RAN, the explicit expressions of ergodic rate for the hierarchical paradigm is derived. Meanwhile, both the waiting delay and latency ratio for users requiring a single content are exploited. According to the evaluation results of ergodic rate on waiting delay, the transmit latency can be effectively reduced through improving the capacity of both fronthaul and radio access links. Moreover, to fully explore the potential of hierarchical content caching, the transmit latency for users requiring multiple content objects is optimized as well in three content transmission cases with different radio access links. The simulation results verify the accuracy of the analysis, further show the latency decreases significantly due to the hierarchical paradigm.

The Interplay between Artificial Intelligence and Fog Radio Access Networks

The interplay between artificial intelligence(AI) and fog radio access networks(F-RANs) is investigated in this work from two perspectives: how F-RANs enable hierarchical AI to be deployed in wireless networks and how AI makes F-RANs smarter to better serve mobile devices. Due to the heterogeneity of processing capability, the cloud, fog, and device layers in F-RANs provide hierarchical intelligence via centralized, distributed, and federated learning. In addition, cross-layer learning is also introduced to further reduce the demand for the memory size of the mobile devices. On the other hand, AI provides F-RANs with technologies and methods to deal with massive data and make smarter decisions. Specifically, machine learning tools such as deep neural networks are introduced for data processing, while reinforcement learning(RL) algorithms are adopted for network optimization and decisions. Then, two examples of AI-based applications in F-RANs, i.e., health monitoring and intelligent transportation systems, are presented, followed by a case study of an RL-based caching application in the presence of spatio-temporal unknown content popularity to showcase the potential of applying AI to F-RANs.

Joint Resource Allocation and Admission Control in Sliced Fog Radio Access Networks

Network slicing based fog radio access network(F-RAN) has emerged as a promising architecture to support various novel applications in 5 G-and-beyond wireless networks. However, the co-existence of multiple network slices in F-RANs may lead to significant performance degradation due to the resource competitions among different network slices. In this paper, the downlink F-RANs with a hotspot slice and an Internet of Things(Io T) slice are considered, in which the user equipments(UEs) of different slices share the same spectrum. A novel joint resource allocation and admission control scheme is developed to maximize the number of UEs in the hotspot slice that can be supported with desired quality-of-service, while satisfying the interference constraint of the UEs in the Io T slice. Specifically, the admission control and beamforming vector optimization are performed in the hotspot slice to maximize the number of admitted UEs, while the joint sub-channel and power allocation is performed in the Io T slice to maximize the capability of the UEs in the Io T slice tolerating the interference from the hotspot slice. Numerical results show that our proposed scheme can effectively boost the number of UEs in the hotspot slice compared to the existing baselines.

Joint Design of Coalition Formation and Semi-Blind Channel Estimation in Fog Radio Access Networks

Coordinated signal processing can obtain a huge transmission gain for Fog Radio Access Networks(F-RANs).However,integrating into large scale,it will lead to high computation complexity in channel estimation and spectral efficiency loss in transmission performance.Thus,a joint cluster formation and channel estimation scheme is proposed in this paper.Considering research remote radio heads(RRHs)centred serving scheme,a coalition game is formulated in order to maximize the spectral efficiency of cooperative RRHs under the conditions of balancing the data rate and the cost of channel estimation.As the cost influences to the necessary consumption of training length and estimation error.Particularly,an iterative semi-blind channel estimation and symbol detection approach is designed by expectation maximization algorithm,where the channel estimation process is initialized by subspace method with lower pilot length.Finally,the simulation results show that a stable cluster formation is established by our proposed coalition game method and it outperforms compared with full coordinated schemes.

Deep Learning Based Channel Estimation in Fog Radio Access Networks

As a promising paradigm of the fifth generation networks,fog radio access network(F-RAN)has attracted lots of attention nowadays.To fully utilize the promising gain of F-RANs,the acquisition of accurate channel state information is significant.However,conventional channel estimation approaches are not suitable in F-RANs due to the large training and feedback overhead.In this paper,we consider the channel estimation in F-RANs with fog access point(F-AP)equipped with massive antennas.Thanks to the computing ability of F-AP and the sparsity of channel matrices in angular domain,Gated Recurrent Unit(GRU),a data-driven based channel estimation is proposed at F-AP to reduce the training and feedback overhead.The GRU-based method can capture the hidden sparsity structure automatically through the network training.Moreover,to further improve the channel estimation,a bidirectional GRU based method is proposed,whose target channel structure is decided by previous and subsequent structures.We compare the performance of our proposed channel estimation with traditional methods(Orthogonal Matching Pursuit(OMP)and Simultaneous OMP(SOMP)).Simulation results show that the proposed approaches have better performance compared with the traditional OMP and SOMP methods.

A Dynamic Distributed Spectrum Allocation Mechanism Based on Game Model in Fog Radio Access Networks

With the explosive growth of highspeed wireless data demand and the number of mobile devices, fog radio access networks(F-RAN) with multi-layer network structure becomes a hot topic in recent research. Meanwhile, due to the rapid growth of mobile communication traffic, high cost and the scarcity of wireless resources, it is especially important to develop an efficient radio resource management mechanism. In this paper, we focus on the shortcomings of resource waste, and we consider the actual situation of base station dynamic coverage and user requirements. We propose a spectrum pricing and allocation scheme based on Stackelberg game model under F-RAN framework, realizing the allocation of resource on demand. This scheme studies the double game between the users and the operators, as well as between the traditional operators and the virtual operators, maximizing the profits of the operators. At the same time, spectrum reuse technology is adopted to improve the utilization of network resource. By analyzing the simulation results, it is verified that our proposed scheme can not only avoid resource waste, but also effectively improve the operator's revenue efficiency and overall network resource utilization.

An Efficient Scheduling Scheme for Fronthaul Load Reduction in Fog Radio Access Networks

Fog radio access network(F-RAN) is one of the key technology that brings cloud computing benefit to the future of wireless communications for handling massive access and high volume of data traffic. The high fronthaul burden of a typical cellular system can be partially diminished by utilizing the storage and signal processing capabilities of the F-RANs, which is still not desirable as user throughput requirement is in the increasing trend with the increment of the internet of things(IoT) devices. This paper proposes an efficient scheduling scheme that minimizes the fronthaul load of F-RAN system optimally to improve user experience, and minimize latency. The scheduling scheme is modeled in a way that the scheduler which provides the lower fronthaul load while fulfilling the minimum user throughput requirement is selected for the data transmission process. Simulation results in terms of user selection fairness, outage probability, and fronthaul load for a different portion of user equipments(UEs) contents in fog access point(F-AP) are shown and compared with the most common scheduling scheme such as round robin(RR) scheme to validate the proposed method.

Deep reinforcement learning based computation offloading and resource allocation for low-latency fog radio access networks

Fog Radio Access Networks(F-RANs)have been considered a groundbreaking technique to support the services of Internet of Things by leveraging edge caching and edge computing.However,the current contributions in computation offloading and resource allocation are inefficient;moreover,they merely consider the static communication mode,and the increasing demand for low latency services and high throughput poses tremendous challenges in F-RANs.A joint problem of mode selection,resource allocation,and power allocation is formulated to minimize latency under various constraints.We propose a Deep Reinforcement Learning(DRL)based joint computation offloading and resource allocation scheme that achieves a suboptimal solution in F-RANs.The core idea of the proposal is that the DRL controller intelligently decides whether to process the generated computation task locally at the device level or offload the task to a fog access point or cloud server and allocates an optimal amount of computation and power resources on the basis of the serving tier.Simulation results show that the proposed approach significantly minimizes latency and increases throughput in the system.

Joint uplink and downlink resource allocation for low-latency mobile virtual reality delivery in fog radio access networks

Fog radio access networks(F-RANs),in which the fog access points are equipped with communication,caching,and computing functionalities,have been anticipated as a promising architecture for enabling virtual reality(VR)applications in wireless networks.Although extensive research efforts have been devoted to designing efficient resource allocation strategies for realizing successful mobile VR delivery in downlink,the equally important resource allocation problem of mobile VR delivery in uplink has so far drawn little attention.In this work,we investigate a mobile VR F-RAN delivery framework,where both the uplink and downlink transmissions are considered.We first characterize the round-trip latency of the system,which reveals its dependence on the communication,caching,and computation resource allocations.Based on this information,we propose a simple yet efficient algorithm to minimize the round-trip latency,while satisfying the practical constraints on caching,computation capability,and transmission capacity in the uplink and downlink.Numerical results show that our proposed algorithm can effectively reduce the round-trip latency compared with various baselines,and the impacts of communication,caching,and computing resources on latency performance are illustrated.

Pricing-based edge caching resource allocaƟon in fog radio access networks

The edge caching resource allocation problem in Fog Radio Access Networks(F-RANs)is investigated.An incentive mechanism is introduced to motivate Content Providers(CPs)to participate in the resource allocation procedure.We formulate the interaction between the cloud server and the CPs as a Stackelberg game,where the cloud server sets nonuniform prices for the Fog Access Points(F-APs)while the CPs lease the F-APs for caching their most popular contents.Then,by exploiting the multiplier penalty function method,we transform the constrained optimization problem of the cloud server into an equivalent non-constrained one,which is further solved by using the simplex search method.Moreover,the existence and uniqueness of the Nash Equilibrium(NE)of the Stackelberg game are analyzed theoretically.Furthermore,we propose a uniform pricing based resource allocation strategy by eliminating the competition among the CPs,and we also theoretically analyze the factors that affect the uniform pricing strategy of the cloud server.We also propose a global optimization-based resource allocation strategy by further eliminating the competition between the cloud server and the CPs.Simulation results are provided for quantifying the proposed strategies by showing their efficiency in pricing and resource allocation.

A Broad Learning-Driven Network Traffic Analysis System Based on Fog Computing Paradigm

The development of communication technologies which support traffic-intensive applications presents new challenges in designing a real-time traffic analysis architecture and an accurate method that suitable for a wide variety of traffic types.Current traffic analysis methods are executed on the cloud,which needs to upload the traffic data.Fog computing is a more promising way to save bandwidth resources by offloading these tasks to the fog nodes.However,traffic analysis models based on traditional machine learning need to retrain all traffic data when updating the trained model,which are not suitable for fog computing due to the poor computing power.In this study,we design a novel fog computing based traffic analysis system using broad learning.For one thing,fog computing can provide a distributed architecture for saving the bandwidth resources.For another,we use the broad learning to incrementally train the traffic data,which is more suitable for fog computing because it can support incremental updates of models without retraining all data.We implement our system on the Raspberry Pi,and experimental results show that we have a 98%probability to accurately identify these traffic data.Moreover,our method has a faster training speed compared with Convolutional Neural Network(CNN).

基于RA-IDNC的D2D辅助F-RANs协作重传方案

针对D2D辅助F-RANs场景中网络资源利用不充分、重传平均完成时延较高的问题,提出了一种基于速率感知IDNC(Rate perception IDNC,RA-IDNC)的D2D辅助F-RANs重传策略,通过利用不同频段的蜂窝链路和带外D2D链路进行网络编码传输。在重传阶段,首先推导了D2D辅助F-RANs中使用网络编码重传的最小完成时间公式;然后利用图论的方法构造终端和增强远程无线电头(enhanced Remote Radio Head,eRRH)联合的RA-IDNC图,将最小化重传完成时间问题转换为联合图的最大权重独立集搜索问题,并综合考虑链路丢包率、终端丢失数据包个数、设备接口传输速率、接收终端个数等因素设计权重,为了降低运算复杂度,采用贪婪算法对联合RA-IDNC图进行搜索选取最优的编码传输策略;最后,利用接口传输速率不一致造成的发送持续时间不同,搜索满足再次发送条件的设备,并构建空闲时间下设备的联合RA-IDNC图,从提前处于空闲状态的终端和eRRH中搜索可行的编码传输方案,在不增加传输时间的基础上尽可能多的恢复单次重传过程中终端丢失数据包的个数。仿真结果表明,与现有的RA-IDNC编码方案相比,本文所提方案能够有效提高F-RANs中的重传效率,降低重传平均完成时间。

雾无线接入网中面向时延的协作缓存策略

为了改善雾无线接入网(Fog-Radio Access Networks,F-RANs)中多个边缘节点之间的协作缓存问题,提出基于异步优势演员评论家(Asynchronous Advantage Actor-Critic,A3C)算法的协作缓存策略。该策略根据用户的历史请求信息学习用户偏好模型,并利用区域用户的偏好模型预测每个雾接入节点(Fog-Access Point,F-AP)服务区域内的局部内容流行度。为了提高边缘节点存储空间的利用率,考虑F-AP以及用户设备(User Equipment,UE)间的协作缓存,以最小化用户获取请求内容的平均下载时延为目标,根据获得的内容流行度分布,优化热门内容的缓存位置。将所提策略与参考策略、贪婪缓存策略和随机缓存策略等3种策略相比,仿真结果表明,所提策略能够实现更低的平均内容下载时延。

雾网络中基于统计分布的内容缓存与交付方案

作为5G中的一种重要模型,雾无线接入网络(Fog Radio Access Network,F-RAN)通过设备到设备通信和无线中继等技术获得了显著的性能增益,而边缘设备中合适的缓存则可以让内容缓存用户(Caching Users,CUs)向内容请求用户(Requesting Users,RUs)直接发送缓存内容,有效减小前传链路的负担和下载延迟。考虑一个F-RAN模型下用户发出请求并获得交付的场景,将每个CU的内容请求队列建模为独立的M/D/1模型,分析导出CUs缓存命中率和平均下载延迟关于内容缓存与交付方案的表达式,证明CUs缓存命中率与内容统计分布之间的联系有助于实现前者的近似最优解。针对在一段时间内的期望视角下建立的优化问题,提出了基于统计分布的算法并注意了执行时的交付控制。仿真结果表明,相较于现有缓存策略,优化内容整体统计分布的方案能够最大化CUs缓存命中率,同时减小平均下载延迟。

6G智慧雾无线接入网：架构与关键技术

融合雾无线接入网、非正交多址技术以及人工智能技术优势的智慧雾无线接入网,被认为是满足6G无线网络巨容量、极低时延及超连接性能需求的有效解决方案。将概述6G智慧雾无线接入网的基本原理、关键技术、国内外研究和产业进展、智慧雾无线接入网标准化进展等,探讨智慧雾无线接入网面临的技术挑战。

雾无线接入网中的多层协作缓存方法

针对边缘缓存和雾无线接入技术中降低后传链路负载的问题,提出了一种雾无线接入网中的多层协作缓存方法。考虑网络架构、文件流行度估计、链路容量等因素,将相关的优化问题分解为每个层中缓存布置的背包问题,并分别使用贪心算法求解。仿真结果表明,所提出的多层协作缓存方法能够有效降低后传链路负载,同时具有较高的缓存命中率。

5G室内密集立体覆盖的计算通信:架构、方法与增益

提出了一种新型的室内密集立体覆盖的计算通信一体化架构,通过挖掘信道计算、容量计算以及网络资源优化计算之间的内在联系,并利用基于云计算和雾计算的密集分布式接入网络的优势,该架构完成了计算电磁学、计算信息论与大规模优化理论到计算通信理论的深度融合。介绍了该架构的实现方法,即以密集异构分布式无线接入网络作为通信接入网络基础架构,利用分布式的计算资源结合计算电磁学理论实现并行化的信道计算,据此进一步依据计算信息论实现容量计算,并基于大规模优化理论完成多用户的网络资源优化计算,最终实现由传播环境到信道容量与资源分配的计算通信。

雾无线接入网:架构、原理和挑战

为了降低前传(fronthaul)链路开销、避免传统云无线接入网基带单元池中无线信号处理大规模/高实时要求、充分利用边缘网络设备的计算和存储能力,提出了雾无线接入网络(fog computing based radio access network,F-RAN),作为异构云无线接入网络的演进。F-RAN的核心是利用用户和边缘网络设备的计算和存储功能,进行本地业务分发、分布式信号处理和分布式资源管理等。详细介绍了F-RAN的系统架构、关键技术及未来需研究的问题。

基于深度增强学习的无人机赋能雾无线电接入网络的能效优化

雾无线电接入网络适合用于广域范围内的诸如管线管网监测等国家重要行业的物联网应用场景。然而基于地面雾接入节点的网络将受到环境、地形等影响,无法及时有效地提供雾接入服务。利用低空无人机作为雾接入点实现空地的边缘通信和雾计算方面引起了普遍的关注。本文探讨怎样利用深度增强学习来提高无人机雾接入点的能效,延长无人机的任务时间。深度增强学习可以保障无人机雾接入点及时地调整空地通信和计算的配置策略,包括资源优化、动态任务卸载以及缓存,也可以优化无人机在三维空间中的飞行航迹,提高无人机赋能的雾无线电接入网络的总体性能。研究的创新性在于综合论述了深度增强学习用于无人机赋能的雾无线电接入网络要解决的主要优化问题,并且总结了解决相关优化问题的技术细节,最后对深度增强学习应用于无人机赋能的雾无线电接入网络的技术挑战和未来研究方向展开讨论。

雾无线接入网中基于神经网络的资源分配方案

考虑雾无线接入网(Fog Radio Access Network,F-RAN)中的性能优化问题,提出一种基于深度神经网络(Deep Neural Network,DNN)的资源分配方案。该方案旨在通过资源分配策略来最大化经济频谱效率(Economical Spectral Efficiency,ESE)。为解决传统资源分配方案需要大量计算的问题,该方案借助神经网络模型,将ESE作为损失函数,使用更少的计算量来确定用户的波束赋形,从而实现实时处理。仿真结果表明,相比于基于传统凸优化功率分配方案或者是基于监督学习的CNN方法,所提出的方案的光谱效率(Spectral Efficiency,SE)和ESE的最大增益分别可以达到5%和20%。此外,该方案在执行时间上与CNN方案接近,明显优于传统算法。