Access Network Virtualization Research Articles

Real-Time Software Implementation of Filter-Separated MIMO Adaptive Equalizer for Fully Virtualized Access Networks

Real-Time Software Implementation of Filter-Separated MIMO Adaptive Equalizer for Fully Virtualized Access Networks

Real-time demonstration of industrial protocol applications by PHY softwarization for fully virtualized access networks

To realize optical access networks that can flexibly and quickly accommodate a variety of new services, we need virtualization of the access systems targeting physical-layer (PHY) softwarization as well as the softwarization of management and control functions. Our previous demonstrations showed 10-Gbps-class passive optical network (PON) PHY implementations on general-purpose servers using graphic processing units. However, only the PON PHY application was considered. As a promising use case, the application of industrial protocols to access networks is being considered. PROFINET isochronous real-time (IRT), which is one such industrial protocol, sets the network requirements of 1-ms latency and 1-µs jitter. This paper builds and tests a demonstration platform that runs industrial protocols combined with Ethernet PHY and PON PHY on a general-purpose server. Ethernet PHY is newly implemented through our technique of parallelizing frame synchronization. To achieve the deterministic communications required by industrial protocols, cyclic transfer and time slot allocation are implemented in the platform. The demonstration results show that the platform achieves real-time software implementation of 10G Ethernet PHY and that our industrial protocol implementations satisfy the PROFINET IRT requirements of latency and jitter. This confirms that the softwarization of optical access systems is possible for applications beyond PON.

Baseband-Function Placement With Multi-Task Traffic Prediction for 5G Radio Access Networks

The 5G Radio Access Network (RAN) virtualization aims to improve network quality and lower the operator’s costs. One of its main features is the functional split, i.e., dividing the instantiation of RAN baseband functions into different units over metro-network nodes. However, its optimal placement is non-trivial: it depends on the application requirements and on the expected traffic volume, whose daily variation highly impacts the total power consumption. Current optimization solutions fail to provide a placement solution capable of handling traffic fluctuations. In fact, the standard machine learning algorithms used in the literature for planning the network resources in advance result in an allocation that is inadequate to carry the actual traffic at all the time-slots. Hence, we must reserve an artificial buffer capacity in the nodes to ensure feasibility. Instead, our proposed method exploits a fine-grained two-step multi-task algorithm that predicts the mean and quantile traffic, making the artificial capacity no longer necessary. The subsequent placement uses mixed-integer linear programming and a heuristic. The former considers the expected traffic in the objective function (to estimate costs) and the quantile in the constraints (to enforce capacity limits). The heuristic combines the mean and quantile results to minimize the power and comply with the requirements. While using sufficiently large artificial buffers guarantees robustness with a mild power increase compared to the oracle, the fine-grained multi-task model improves the results, reducing the power consumption compared to the mean and meets all constraints. The heuristic enables significant computational time reduction.

Impact of Network Densification on Joint Slicing and Functional Splitting in 5G

The virtualization of the fifth generation radio access network (RAN), which distributes the next generation Node B functions between a central unit (CU) and a distributed unit (DU), along with the emergence of slicing as a corner-stone of mobile communications networks, pose new challenges. One of these challenges is the management of the joint slicing and allocation of appropriate distribution of functions between CU and DU, known as functional split. In this work, a discussion on the challenges and open issues of joint slicing and functional splitting is presented, taking into account the additional complexity resulting from instantiating multiple slices per OU, each one with a different functional split. It is shown that the additional complexity, which is translated into higher cost, is worthwhile depending on the RAN deployment density.

Wireless Network Virtualization Resource Sharing Based on Dynamic Resource Allocation Algorithm

In recent years, with the continuous development of IT industry, network virtualization technology has gradually developed into a hotspot of research and application in the field of communication because IT allows the construction of customized virtual network (VN) on shared physical infrastructure. In particular, the proposal of wireless network virtualization for the development of the enhanced Internet. Because these technologies require effective algorithms named virtual network mapping (VNE) to instantiate virtual networks on the SN, on the basis of satisfying the isolation between wireless network slices, a resource allocation mechanism based on bilateral auction is proposed. At present, many researchers have proposed that network virtualization technology is a feasible way to solve the rigid Internet architecture. In the field of existing research focuses on the cable network virtualization, mainly related to the data center network and backbone, then, the side of the wireless access network virtualization the key technology research is still in its infancy, in order to accelerate the process of wireless skill for innovation and meet the demand of the difference of the emerging business, wireless network virtualization becomes the research hotspot and focus of academia and industry. However, no one has applied the dynamic resource allocation algorithm to wireless network virtualization resource sharing field. Here, this paper analyzes and summarizes the existing virtual network resource allocation schemes at home and abroad. The wireless virtual network resource allocation model and cross-domain virtual network resource allocation model are established, and the network centrality theory in social network and complex network is introduced to analyze the topology properties of virtual network and physical network, and two efficient virtual network resource allocation methods are proposed.

Zero-Touch Multi-Service Provisioning With Pluggable Module-Type OLT in Access Network Virtualization Testbed

This paper investigates a multi-service activation procedure for telecommunications carriers’ zero-touch provisioning workflows using access network virtualization technologies in an optical access network. We examine a procedure in which one network function or a set of functions is initiated depending on the services to be activated after a line is automatically opened in response to a customer’s action. The feasibility of this procedure is experimentally verified in a joint testbed environment between three Asia-Pacific telecommunications carriers. The results showed that even when 32 customers simultaneously request service activation, it takes less than 30 seconds for all of the customers’ requests to be completed. In addition, we discuss potential implementations for activating various multi-service provisioning in response to customer requirements and network isolation in terms of virtualized resources and infrastructure. We also propose an implementation that utilizes network resources shared by various services more efficiently than that presented in our previous conference report, and we experimentally verify its feasibility. In addition, the experimental configurations briefly described in the previous report are described in detail in this paper.

Standardization and Technology Trends in Optical, Wireless and Virtualized Access Systems

This paper reviews access system standardization activities and related technologies from the viewpoints of optical-based PON access, mobile access systems including LPWAN, and access network virtualization. Future study issues for the next access systems are also presented.

Scalable RAN Virtualization in Multitenant LTE-A Heterogeneous Networks

Cellular communications are evolving to facilitate the current and expected increasing needs of quality of service, high data rates, and diversity of offered services. Toward this direction, radio access network (RAN) virtualization aims at providing solutions of mapping virtual network elements onto radio resources of the existing physical network. This paper proposes the Resources nEgotiation for NEtwork Virtualization (RENEV) algorithm, which is suitable for application in heterogeneous networks in Long-Term Evolution Advanced (LTE-A) environments, consisting of a macro evolved Node B overlaid with small cells (SCs). By exploiting radio resource management principles, RENEV achieves slicing and on-demand delivery of resources. Leveraging the multitenancy approach, radio resources are transferred in terms of physical radio resource blocks among multiple heterogeneous base stations, which are interconnected via the X2 interface. The main target is to deal with traffic variations in geographical dimension. All signaling design considerations under the current Third-Generation Partnership Project LTE-A architecture are also investigated. Analytical studies and simulation experiments are conducted to evaluate RENEV in terms of the network's throughput and additional signaling overhead. Moreover, we show that RENEV can be applied independently on top of already proposed schemes for RAN virtualization to improve their performance. The results indicate that significant advantages are achieved both from the network's and users' perspective and that it is a scalable solution for different numbers of SCs.

Fronthaul network design for radio access network virtualization from a CAPEX/OPEX perspective

With the increase of mobile traffic demand and the need to reduce expenses to handle this demand, a novel solution, known as Cloud Radio Access Network (C-RAN), has been proposed for future radio access network. This solution involves virtualizing base stations and centralizing processing resources into a baseband processing unit (BBU) pool. C-RAN also helps fully deploying cooperative schemes used in LTE and LTE-Advanced. In this paper, we analyze C-RAN cost structure. Then, unlike previous works, we mathematically formulate cell-BBU pool assignment, taking into account fronthaul network expenditure. Two optimization models are proposed for two different architectures. We then use these formulations to develop solutions to our problem, which optimize the C-RAN costs subject to demand constraints. Through extensive experiments, cost efficiency of C-RAN architecture is discussed and the effect of different parameters is analyzed. We also derive conditions where utilizing C-RAN architecture can help cost savings.

Evaluating Dynamic Bandwidth Allocation of Virtualized Passive Optical Networks Over Mobile Traffic Traces

Deploying passive optical networks (PONs) requires substantial investments, despite their savings compared to point-to-point solutions. Consequently, network-sharing business models where a network operator leases its system to other service providers, are being investigated and deployed in some countries. These types of schemes require that operators are either physically isolated or have reserved capacity so that they can operate independently. We present the benefits that our group-assured dynamic bandwidth assignment (DBA) algorithm brings to access network virtualization. This algorithm enables operators to take advantage of the benefits of statistical multiplexing, while maintaining isolation from other operators in the PON. To test our algorithm, we considered the particular use case of a mobile system backhauling multiple base stations through a PON. Our tests, operated over real mobile traffic traces, show increased quality of service with a reduction of backhaul capacity of >30% compared to a legacy DBA algorithm.

Wireless World 2020: Radio Interface Challenges and Technology Enablers

Future pervasive communication system requirements for two to three orders of magnitude capacity improvement, flexible, fast deployment, and cost/energy efficiency are expected to revolutionize the way we design and use wireless networks. From a network infrastructure perspective, the emphasis is placed on achieving ubiquitous, real-time high data rate communications ?anytime-anywhere,? including at cell-edge, through Small Cell Network architectures and Heterogeneous Cellular Networks (HetNets). From a pervasive systems? perspective, the vision of the Internet of Things suggests the integration between ubiquitous computing and wireless communications targeting a reliable connectivity of things, i.e., computers, sensors, and everyday objects equipped with transceivers. From a backhaul bandwidth, network resource sharing, and optimization perspective, cloud-based processing and radio access network virtualization provide a revolutionary approach toward balancing the degree of centralization of physical and virtual resources management. In this article we analyze these three trends, present key technology enablers, and assess suitable performance merits in an effort to set the scene for the wireless evolution in the era beyond 2020.

Radio access network virtualization for future mobile carrier networks

This article presents a survey of cellular network sharing, which is a key building block for virtualizing future mobile carrier networks in order to address the explosive capacity demand of mobile traffic, and reduce the CAPEX and OPEX burden faced by operators to handle this demand. We start by reviewing the 3GPP network sharing standardized functionality followed by a discussion on emerging business models calling for additional features. Then an overview of the RAN sharing enhancements currently being considered by the 3GPP RSE Study Item is presented. Based on the developing network sharing needs, a summary of the state of the art of mobile carrier network virtualization is provided, encompassing RAN sharing as well as a higher level of base station programmability and customization for the sharing entities. As an example of RAN virtualization techniques feasibility, a solution based on spectrum sharing is presented: the network virtualization substrate (NVS), which can be natively implemented in base stations. NVS performance is evaluated in an LTE network by means of simulation, showing that it can meet the needs of future virtualized mobile carrier networks in terms of isolation, utilization, and customization.

DSP-based optical access approaches for enhancing NG-PON2 systems

Motivated by recent progress in next-generation PON2, or NG-PON2, standardization, this article reviews digital signal processing technologies to further enhance NG-PON2 systems by realizing flexible and cost-effective optical access network deployments. First, flexible speed upgrades by DSP-enabled advanced modulation and multiplexing approaches are described, followed by a consideration of cost-performance trade-offs. Next, reach extension using DSP-based digital coherent reception and impairment compensation is overviewed. Finally, as a future goal, access network virtualization with reconfigurable optical line terminals and optical network units is discussed.

분산 안테나 시스템에 기반한 무선 액세스망 가상화

본 논문에서는 다수 무선 접속 기술 (Radio Access Technology; RAT)이 통합되어 운용되는 셀룰러 네트워크에서 차세대 이동통신 네트워크 진화 방향인 디지털 유닛 (DU)-라디오 유닛 (RU) 분리 구조를 기본으로 무선 액세스 네트워크 가상화 알고리즘을 제안한다. 제안 무선 액세스 네트워크 가상화 알고리즘은 크게 기본 액세스 네트워크 가상화, RAT 가상화, 액세스 경로 이주 알고리즘으로 구성되고, 순차적인 수행을 통하여 최종 무선 액세스 네트워크 가상화가 이루어진다. 제안 무선 액세스 네트워크 가상화 알고리즘은 3GPP LTE와 WiMAX의 두 무선 접속 기술을 예제로 한 시스템 레벨 시뮬레이션을 통하여 시스템 수율 측면에서 성능이 평가된다. 이 때, 네트워크 트래픽 로드의 비대칭성에 따른 두 가지 시나리오를 고려하고, 제안 무선 액세스 네트워크 가상화 알고리즘은 비대칭적 네트워크 트래픽 로드와 사용자 분포 상황에서도 상당한 성능 이득을 성취한다. In this paper, we propose a wireless access network virtualization algorithm based on a digital unit (DU)-radio unit (RU) separated network structure in a cellular network with multiple radio access technologies (RATs). The proposed wireless access network virtualization algorithm consists of a baseline access network virtualization, RAT virtualization, and access path migration algorithms. Final wireless access network virtualization is performed by sequentially performing these procedures. Through system-level simulations which assume 3GPP LTE and WiMAX systems, the performance of the proposed wireless access network virtualization is evaluated in terms of system throughput for two scenarios according to asymmetry of network traffic load. Numerical results show that our proposed wireless access network virtualization algorithm achieves significant system throughput gain even in asymmetric traffic load and user distribution situations.

차세대 이동통신 네트워크를 위한 무선 액세스 망 구조 및 가상화 시나리오

차세대 모바일 인터넷의 진화에 따라 2G, 3G, 4G, B4G 이동통신 무선접속네트워크가 공존하게 되며, 기존의 통신 서비스 제공자들은 통합 서비스 제공자로 융합되고, 다수의 가상 서비스 제공자들이 발생하게 될 것으로 예상된다. 이러한 복잡한 미래의 융합 서비스를 제공하기 위해서 미래 인터넷에서는 무선 액세스 네트워크가 융합되고 다양한 서비스 제공자들에 의해 공유되는 무선 네트워크 가상화가 필수적이다. 따라서 본 논문에서는 미래 무선 액세스 네트워크 상에서 여러 서비스 제공자들의 다양한 무선 액세스 기술들이 유연하게 통합 운영될 수 있는 무선 액세스 네트워크 가상화를 위한 망 구조와 가상화 시나리오에 대해서 고찰한다. 본 논문에서 도출된 차세대 이동통신 네트워크를 위한 무선 액세스 망 구조 및 가상화 시나리오는 앞으로 무선 액세스 네트워크 가상화 알고리즘 개발을 위한 기틀 역할을 할 수 있을 것으로 기대된다. In accordance with evolution of next-generation mobile Internet, 2G, 3G, 4G, and B4G mobile communication wireless access networks will be co-existed and service providers will be merged as an integrated service provider. In addition, multiple virtual service operators will appear. In order to provide complicated unified-services, in the future Internet, wireless network virtualization where network resource is shared by various service operators is necessary. Therefore, in this paper, we investigate network architectures and virtualization scenarios for wireless access network virtualization where various wireless access technologies are flexibly operated by multiple service providers over next-generation wireless access networks. We expect that the virtualization scenario and network architecture yielded from this study can play a role as a basis for development of wireless access network virtualization algorithms.

Virtual access network embedding in wireless mesh networks

Network virtualization of a wireless mesh network (WMN) is an economical way for different subscribers to customize their exclusive access networks through a common network infrastructure. The most critical task of network virtualization is virtual network embedding, which can be divided into two sub-problems: node mapping and link mapping. Although there exist approaches to virtual network embedding in wired networks, the characteristics of WMNs make virtual network embedding become a unique and challenging problem. In this paper, virtual access network embedding is studied for WMNs. To support flexible resource allocation in virtual access network embedding, each access node is designed based on orthogonal frequency division multiple access (OFDMA) dual-radio architecture. Through subcarrier allocation on each link, virtual access networks are gracefully separated from each other. To coordinate channel assignment across different links under the constraint of a limited number of orthogonal channels, a novel channel allocation algorithm is proposed to exploit partially-overlapped channels to improve resource utilization. Since the virtual access network embedding problem is NP-hard, a heuristic algorithm is developed based on an enhanced genetic algorithm to obtain an approximate but effective solution. Simulation results illustrate that the virtual access network embedding framework developed in this paper works effectively in WMNs.