Community Research and Development Information Service - CORDIS


5G-Crosshaul Report Summary

Project ID: 671598
Funded under: H2020-EU.

Periodic Reporting for period 1 - 5G-Crosshaul (5G-Crosshaul: The 5G Integrated fronthaul/backhaul)

Reporting period: 2015-07-01 to 2016-06-30

Summary of the context and overall objectives of the project

5G-Crosshaul: The Integrated fronthaul/backhaul is a 30-month collaborative project running under H2020, addressing the topic “ICT 14 – 2014: Advanced 5G Network Infrastructure for the Future Internet” of the Horizon 2020 Work Programme 2014 – 2015. The aim of the project is to develop an adaptive, sharable, cost-efficient 5G transport network solution integrating the fronthaul and backhaul segments of the network whilst supporting existing and new radio access protocol functional splits envisioned in 5G. This transport network will flexibly interconnect distributed 5G radio access and core network functions, hosted on in-network cloud nodes, through the implementation of two novel building blocks: i) a unified data plane encompassing innovative high-capacity transmission technologies and novel deterministic-latency switch architectures (5G-Crosshaul Forwarding Element, XFE); ii) a control infrastructure using a unified, abstract network model for control plane integration (5G-Crosshaul Control Infrastructure, XCI) enabling the operators to easily set up end-to-end services, transparently to all the underlying technologies in the data plane.

The 5G-Crosshaul consortium includes the following partners: (Coordinator) University Carlos III of Madrid, (Technical Manager) NEC Europe LTD, (Innovation Manager) Ericsson Telecomunicazioni SpA, Ericsson AB, Atos Spain SA, Nokia Solutions and Networks GMBH & CO KG, InterDigital Europe LTD, Telefónica Investigación y Desarrollo SA, Telecom Italia SpA, Orange SA, Visiona IP, EBlink, Nextworks, Core Network Dynamics, TELNET Redes Inteligentes, Fraunhofer-Gesellschaft zur Foerderung der angewandten Forschung e.V., Centre Tecnològic de Telecomunicacions de Catalunya, Center for research and telecommunication experimentation for networked communities, Politecnico di Torino, Lunds Universitet and Industrial Technology Research Institute (ITRI).

A. Project context

According to the latest predictions, mobile data traffic will increase 11-fold between 2013 and 2018. 5G radio access network (RAN) technologies serving this mobile data tsunami will require fronthaul and backhaul solutions inside the RAN and between the RAN and the packet core capable of dealing with this increased traffic load. Further, there will be a sizable growth in the capillarity of the network since data rate increase in the 5G RAN is expected to stem to a large extent from increasing the number of base stations and reducing their coverage, i.e., mobile network densification.

To support the increased capillarity of mobile networks (e.g., in terms of interference coordination), and in order to achieve the required 5G data rates, extensive support for novel air interface technologies such as Cooperative Multipoint (CoMP), Carrier Aggregation (CA) and Massive MIMO will be needed. Such technologies require processing of information from multiple base stations simultaneously at a common centralized entity and also tight synchronization of different radio sites. Hence, backhaul and fronthaul will have to meet the most stringent requirements not only in terms of data rates but also in terms of latency, jitter and bit error rates.

Given that the aforementioned challenges will need to be addressed by service providers in a business environment where a revenue increase proportional to the data volume increase is unrealistic, a cost-efficient network deployment, operation and evolution strategy is required. The preferred approach to address the cost-efficiency challenge by the industry, as can be observed in major standardization bodies, e.g., ETSI Network Functions Virtualization (NFV), is virtualization, which exploits multiplexing gains of softwarized network functions on top of commoditized hardware. On the RAN side, this has led to the Cloud RAN concept where cellular base station functions are hosted in cloud computing centers. Once virtualized, base station functions can be flexibly distributed and moved across data centers, providing another degree of freedom for load balancing.

Besides, base station functions can be decomposed in multiple different ways, giving rise to the so-called flexible functional split, where the split between centralized and remote base station functions can be adjusted on a case-by-case basis. In this context, the division between fronthaul and backhaul transport networks will blur as varying portions of functionality of 5G Points of Attachment (5GPoA) might be moved towards the cloud network as required for cost-efficiency reasons. Also for cost reasons, the heterogeneity of transport network equipment must be tackled by unifying the data, control, and management planes across all technologies as much as possible.

To address the aforementioned challenges, the 5G-Crosshaul project aims at developing the next generation of 5G integrated backhaul and fronthaul in a common packet-based network, namely the Crosshaul, enabling a flexible and software-defined reconfiguration of all networking elements in a multi-tenant and service-oriented unified management environment. The 5G-Crosshaul transport network envisioned will consist of high-capacity switches and heterogeneous transmission links (e.g., fiber or wireless optics, high-capacity copper, mmWave) interconnecting Remote Radio Heads, 5GPoAs (e.g., macro and Small Cells), cloud-processing units (mini data centers), and points-of-presence of the core networks of one or multiple service providers.

B. Project Objectives

The 5G-Crosshaul project is a very ambitious initiative aiming at designing the transport network that will serve the 5G deployments. The next generation transport network needs to unify the way it manages the different traffic sources, with really diverse, and potentially extreme, requirements in terms of bandwidth, latency or number of users. Specifically, the project pursues the following eight key objectives:
• Design of the 5G-Crosshaul Control Infrastructure (XCI): Develop XCI by extending existing Software Defined Network (SDN) controllers to provide the services for novel Northbound (NBI) and Southbound (SBI) Interfaces and enable multi-tenancy support in trusted environments.
• Specification of the XCI’s northbound (NBI) and southbound (SBI) interfaces: Define interfaces to accelerate the integration of new data plane technologies (SBI) and the introduction of new services (NBI) via novel or extended interfaces.
• Unification of the 5G-Crosshaul data plane: Develop a flexible frame format to allow carrying fronthaul and backhaul on the same physical link to replace different technologies with a uniform transport technology for both fronthaul and backhaul.
• Development of physical and link-layer technologies to support 5G requirements: Exploit advanced physical layer technologies, not currently used in the 5G-Crosshaul network segment, as well as novel technologies, such as wireless optics, flexi-PON, etc. to increase coverage and aggregated capacity of integrated backhaul and fronthaul networks. Increase cost-effectiveness of transport technologies for ultra-dense access networks
• Design of scalable algorithms for efficient 5G-Crosshaul resource orchestration: Develop and evaluate management and control algorithms on top of the XCI NBI that ensure top-notch service delivery and optimal 5G-Crosshaul resource utilization.
• Design of essential 5G-Crosshaul-integrated (control/planning) applications: Develop an ecosystem of most essential XCI NBI applications, both to support (prediction, planning, monitoring) and to exploit (media distribution, energy management) the 5G-Crosshaul resource orchestration functions.
• 5G-Crosshaul key concept validation and proof of concept: Demonstration and validation of 5G-Crosshaul technology components which will be integrated into a set of 5G testbeds in Madrid, Berlin, Barcelona and Taiwan.

See Figure 1: 5G-Crosshaul System.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The first year of the project has been devoted mainly to the initial design and specification of the overall system that will become the integrated 5G transport infrastructure. Designing such a single infrastructure requires rethinking of all architectural aspects of the transport network conceived in 5G-Crosshaul. Work Package (WP) 1 is in charge of defining the set of use cases and scenarios to be used to challenge the architecture of the system. WP1 is also in charge of designing the baseline architecture of the 5G-Crosshaul, which will be refined in the second year based on the implementation experience gained. From a data plane perspective, WP2 evaluates to what extent each of the optical, copper-based, and wireless technologies can fulfil the requirements of 5G traffic flows and identifies what is needed to ensure all requirements are met. This not only includes the design of the links, but also of the nodes from a forwarding point of view. Such data plane is under the control of the 5G-Crosshaul Control Infrastructure (XCI), which brings the software-defined networking (SDN) and Network Functions Virtualization (NFV) paradigms to the project as part of the WP3 work. In turn, WP4 designs the network management applications that will orchestrate the resources required by the use cases by exploiting the services offered by the XCI through its Application Programming Interfaces (APIs). The goal of each of these WPs is to focus mostly on the specific architectural components that are their subject of study as well as the definition of interfaces towards architectural components dealt with in other WPs. The main goal of WP5 is to integrate the components designed in WP2, WP3 and WP4 and to validate experimentally that all the conceived building blocks can work together to fulfill the heterogeneous 5G traffic flow requirements. This will be done by building proof-of-concepts over the various testbeds that are available in the project (5G-Berlin, 5TONIC-Madrid, CTTC, and ITRI).

In WP1 (System Requirements, Scenarios and Economic Analysis) five use cases have been finally selected considering the potential benefits of the 5G-Crosshaul usage for the use cases, the compatibility with other 5G projects and the demonstrability through experimentation. Three of the five use cases are service-oriented, meaning that the use cases are related to specific applications. The service-oriented use cases are: (i) vehicle mobility, (ii) media distribution (CDN and TV broadcasting), and (iii) dense urban information society. In addition to them, two more functional-oriented use cases have been considered: (iv) multi-tenancy, and (v) mobile edge computing.
Due to the multi-tenancy nature of the 5G-Crosshaul, we have defined a classification of the different service use cases for the tenants. We have defined a taxonomy of three categories; i) Over-The-Top (OTT), ii) Mobile Virtual Network Operator (MVNO) and iii) virtual infrastructure provider. Each of them imposes different characteristics and requirements to the 5G-Crosshaul design. Regarding the architecture, in the first year of the project we have worked towards the definition of the 5G-Crosshaul architecture for the Single-Management and Orchestration (MANO) scenario, considering a network supporting multiple technologies through a multi-domain control. This design has been extended to the Multi-MANO architecture, in which the Multi-Tenancy Application (MTA) plays a central role to support recursive instantiation of XCIs. During this initial stage of the project we have also worked in defining a clear path of migration with the operators of the consortium. It is worth highlighting that we have also worked actively with the 5G-Exchange project to define mechanisms in order to include our transport network in their architecture and with the 5GPPP Architecture Working Group to push our architecture into the common architectural framework.
Finally, a preliminary version of a cost model in order to numerically evaluated the Total Cost of Ownership (TCO) of the solutions envisaged by the Project.

In WP2 (Physical and link layer of 5G-Crosshaul) the work has been focused first on identifying the technologies that are most suitable for the deployment of a 5G-Crosshaul network, envisaging a unified data plane encompassing innovative high-capacity transmission technologies and novel deterministic-latency framing protocols. Different data transmission technologies (wireless, fixed access based on both fiber and copper, optical technologies) are considered and their performance is discussed. Furthermore, we focus on understanding how these technologies are combined in the 5G-Crosshaul network. In some cases, it is required to broaden the application domain of existing technologies out of their current scope. The performance parameters of the identified technologies are analyzed in comparison with the requirements of the use cases identified in WP1. As the 5G-Crosshaul control plane is based on SDN and NFV, in a second stage, we have focused on providing guidelines for the development of a southbound interface (SBI) able to deal with the variety of technologies encompassed by the 5G-Crosshaul data plane. To do so, we defined a novel approach based on a protocol agnostic set of parameters to model network nodes and transmission technologies, in order to enable applications, such as optimization of resource allocation and energy, running over the whole network infrastructure. Then, we selected a careful choice of the parameters sets, neither too small to inhibit some applications nor too wide to negatively affect solution cost and scalability, and defined the protocol extensions required, taking as baseline the latest version of the Open Flow specification. 

The work of WP3 (5G-Crosshaul Control and Data planes) focuses on designing the 5G-Crosshaul Control Infrastructure (XCI) and the 5G-Crosshaul Forwarding Element (XFE). During the first 6 months of the project we have focused on understanding the different requirements for each of these elements and sketching a first design of the architecture of each element. The initial design of the data plane architecture, including the concept of 5G-Crosshaul Common Frame (XCF) and its use across the data path, was done in collaboration with WP2. First we have worked towards the initial design of the 5G-Crosshaul Forwarding Element (XFE). Its design has been refined to consider a packet forwarding element (XPFE) and a circuit switch all optical element (XCSE). This design enables the use of packet based technologies while having the possibility of offloading to a pass-through all optical path for traffic with extreme delay and jitter requirements such as traditional fronthaul (e.g., CPRI). 

At this stage, we have also defined the 5G-Crosshaul Common Frame (XCF). This frame will serve as a common transport encapsulation for different technologies, enabling several desired characteristics, such as multi-tenancy and isolation of traffic flowing through the transport network. We have adopted as XCF the MAC-in-MAC encapsulation (Provider Backbone Bridges), and adaptation functions (AFs) for the different media considered. To also support networks already deployed, we also consider a secondary profile for the XCF, based on MPLS-TE. The selection of the technology of choice for the XCF was done after a gap analysis based on the requirements established in WP1.
The work on the XCI has been focused on the initial description of the information model and APIs exposed by a set of XCI services towards the application-plane through a Northbound Interface (NBI). Such services include Topology and Inventory, Provisioning and Flow actions, IT infrastructure and Inventory, Statistics, NFV Orchestration, VNF Management, Analytics for Monitoring, Local Management Service, and Multi-tenancy. 

WP4 (Enabled innovations through 5G-Crosshaul) work during this period has focused on determining initial set of requirements for the Northbound Interface (NBI) of the 5G-Crosshaul Control Infrastructure (XCI) and the design of the applications that can provide optimization and reconfiguration of 5G-Crosshaul resources through the NBI. A set of applications have been defined: Multi-tenancy Application (MTA), Mobility Management Application (MMA), Energy Management and Monitoring Application (EMMA), Resource Management Application (RMA), Virtual Infrastructure Manager and Planner Application (VIMaP), Content Delivery Network Management Application (CDNMA), and TV Broadcasting Application (TVBA). Following this analysis, WP4 has provided an in-depth description of the applications in a clear and schematic way of the main algorithms that the applications will implement as well as the main modules involved for the application design and the scenarios considered. The interaction between the different applications and the main workflows and interfaces to other applications to be taken into account has been a major piece of work during this time, this is a key point that will allow harmonizing the whole system performance. In addition, an initial high-level software design of the applications and their implementation roadmap are defined to provide guidelines for the implementation.

Finally, the work performed in WP5 (Validation and proof of concept) has been devoted to identify all the components provided by partners that will be integrated. This includes components already available that enable the construction of the experimental frameworks as well as components that will be developed during the project according to the design decisions taken at the application, control infrastructure, and data plane levels. In addition to preparing the experimental frameworks required by the project, the above analysis already enabled some early integration efforts among a reduced number of components, which have already been demonstrated at various events (e.g., preliminary multi-domain wireless and optical control plane). We have also defined an initial list of the demonstrations that are planned to be carried out over each of the testbeds. These demonstrations are tightly linked with the use cases defined by the project as well as the key performance indicators (KPIs) that the project targets. In this way, we try to clearly state the purpose of each of the validations undertaken towards meeting the project goal. In any case, it is expected that the demonstrations will be adapted during the project as a function of project decisions to adequately tackle the R&D objectives of the project.

See Figure 2: Architecture of 5G-Crosshaul.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The 5G-Crosshaul Project targets innovations around three pillars of the future 5G transport network. These include: (1) Innovations for data-plane integration across the heterogeneous transmission technologies; (2) Innovations for a unified programmable control; and (3) Novel network applications running on top for optimizing the overall system performance. All these innovations are glued together into an innovative architecture framework that takes into account both technical and techno-economical requirements from the stakeholders of the value chain, namely operators, vendors and service operators.

In the first year, significant progress has been made to bring in early innovations that the project can build on and nurture further in the following years. These are presented briefly below:

1. An innovative design for the data plane that allows for the multiplexing of various 5G backhaul and fronthaul data across heterogeneous transmission and switching technologies. The innovative design featured novel framing procedures, under the umbrella of so-called XCF – 5G-Crosshaul Common Frame, both for packet-switching and circuit-switching, raising the potential for exploitation in two new products, namely XPFE – 5G-Crosshaul Packet Forwarding Element, and XCSE – 5G-Crosshaul Circuit Switching Element. The concept of an adaptation unit (XAU – 5G-Crosshaul Adaptation Unit) has also been proposed, to allow for integration between XCSE and XPFE as well as interoperation with other non-XCF technology domains. This might lead to new features that can be exploited in new products. Alongside these innovations, the project has also researched and developed new transmission technologies for high capacity both optical and wireless, and analyzed their respective domain of suitability for a cost-efficient deployment in the service coverage-area.

2. A novel method for common abstraction of the data-plane technologies towards the control infrastructure (so-called XCI), which will help in the design of the south-bound interface of the controller, and in providing a common forwarding model to build upon in the design of the controller. This method aims at exposing in the most common way possible parameters and resources of the underlying data-plane technologies, so that one can facilitate the integration of all these technologies under common control and orchestration. In order to assess the proposed method, a design based on OpenFlow has been provided. This design results in novel extensions proposed to current implementation of OpenFlow Standard, and would hence lead to adoption in future releases of the OpenFlow standard.

3. First novel network applications that can optimize the overall system performance. As an example, the EMMA - Energy Management and Monitoring Application has been designed to consistently optimize the use of the energy consumption of the network infrastructure. This enables a sustainable approach to dynamically activate/deactivate networking resources to reduce energy and hence save costs. Importantly, the framework for the development of novel network applications has been laid down and more applications are under development and hence expected to yield future innovations that can be packaged into new product or service offering.

All the above innovations are driven primarily by the need to make the future 5G transport network more flexible in order to ease and hence accelerate the deployment of new services, whilst guaranteeing cost-efficient use of all the resources in play. This obviously results into a direct socio-economic impact, through lower cost and higher efficiency for the networking infrastructure stakeholders (operators, vendors, and service providers), and the end user customer in terms of better service in terms of quality and ubiquitous access, and lower bills. The overall society will also see the benefit of driving the future transport network towards more flexibility and cost-efficiency, whilst supporting effectively the various services envisioned in future 5G system. In addition, the innovations from 5G-Crosshaul project are expected to give the industrial companies (large, medium and small) in 5G-Crosshhaul and the extended European 5G-PPP community a privileged position and competitive advantage in the European and global markets through new generations of flexible and innovative access and core networks solutions. An exploitation plan is being defined to assess the possible impact on the product roadmaps of the main vendors involved in the project.
In order to ensure wide-reach of the innovations developed in the project, the consortium members have been very active in disseminating the project concept and early results to the European (inside and outside the 5G-PPP community) and wide international research and industrial community. Several (over 30) talks were delivered at key industrial and research events (e.g. NGMN, IWPC, etc.), over 20 scientific articles published or accepted for publications and many others submitted, over 5 workshops organized or co-organized, and over 15 presentations and input contributions to relevant standardization activities (e.g. ITU-T, ETSI MEC, IEEE 802.11ay, IETF) provided. A standardization roadmap has also been developed and presented at ETSI, leading to the identification by ETSI of Xhauling as an important area where there might be a standardization gap to be addressed by ETSI (this is ongoing). Despite the start of the project in July 2015, the project has also been proactive in ensuring a visible presence through early demonstrations (1) at the Mobile World Congress 2016 flagship event, with 4 demonstrations provided by four partners, namely InterDigital, CTTC, Fraunhofer HHI, and CND, and (2) at EUCNC 2016 conference, with 4 demonstrations from Ericsson, Fraunhofer HHI, CND, and ITRI.

Related information

Record Number: 192807 / Last updated on: 2016-12-14
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top