Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - 5G-XHaul (Dynamically Reconfigurable Optical-Wireless Backhaul/Fronthaul with Cognitive Control Plane for Small Cells and Cloud-RANs)

Teaser

Wireless data traffic has drastically increased due to a change in the way today’s society creates, shares and consumes information. This current growth in demand from the Radio Access Network (RAN) makes necessary high performance and flexible topologies leveraging...

Summary

Wireless data traffic has drastically increased due to a change in the way today’s society creates, shares and consumes information. This current growth in demand from the Radio Access Network (RAN) makes necessary high performance and flexible topologies leveraging heterogeneous technologies for wired and wireless communications. The overarching goal of 5G to provide ubiquitous, high-speed, high-quality wireless broadband coverage beyond 2020 cannot be achieved without the development of reliable and flexible transport infrastructures, able to react to network-inherent events and demand variations.

Such infrastructure can support the society and people well-being by extending network connectivity with the highest possible energy efficiency, derived from the combination of future-proof technologies and the provision of flexibility of network resources. A consolidated network architecture along with suitable network management schemes and algorithms, which maximize the efficiency of network resources also contribute to CAPEX/OPEX reductions, benefiting both operators/providers and subscribers, providing e.g. improved Quality of Experience (QoE), and lower tariffs.

5G-XHaul aims at building up a novel transport network solution that relies on a flexible infrastructure able to tackle the challenges that future 5G RANs impose. This solution is exploiting the features and benefits of heterogeneous technologies. Optical and wireless technologies are efficiently combined and the associated interoperability challenges are addressed for a wide range of use cases. The architecture we propose is designed to be compliant with current mobile networks, as well as with upcoming 5G networks, addressing both end user and operational services.

The high level project objectives are the following:
- To design a flexible transport network for interconnecting future very dense RAN deployments, Cloud-RAN deployments, and support a defined set of functional split options in a dynamic, service oriented, and cost-effective way.
- 5G-XHaul will enable seamless integration of future-proof technologies in the wired (optical)/wireless (mmWave,Sub-6) metro/access domains by providing the required transport services.
- 5G-XHaul will provide a self-consistent transport network design able to operate in a RAN agnostic way. In addition, 5G-XHaul will make interfaces available to current and future mobile networks, in order to enable performance gains achieved by cooperation between the RAN and the transport networks.

Work performed

The 5G-XHaul work carried out in the framework of WP2 has focused on an analysis of the most relevant 5G use cases and their impact on the transport network in terms of throughput, delay and delay accuracy. A detailed analysis of candidate 5G RAN technologies was carried out considering the most promising ones, e.g. mmWave and massive MIMO, and how they impact the transport network. In this context, a combination of 5G RAT configurations with real downlink measurements from an operational LTE network, provided by COSMOTE, was analyzed. These measurements were collected from 10 sites (33 cells) residing in a typical urban area of a big city, and correspond to a period of 15 days, on a 15 minutes basis. The findings have been reported in the 5G-XHaul deliverable D2.1, as well as in a journal paper currently under review in the IEEE Wireless Communications Magazine. The findings from this study will be contributed to the IEEE P1914.1 (NGFI) working group.

This work represents the basis for our proposed converged wireless/optical transport network architecture, which has become one of the main outputs of the project so far. This architecture has been reported in detail in deliverable D2.2. The overall architecture of 5G-XHaul has been reported in a paper already presented at the 5G-architecture workshop in ICC’2016 together with some preliminary evaluation results. An extended version of this work has been also submitted for publication to the IEEE Communications Magazine and is currently under review. In addition, the physical transport infrastructure defined in 5G-XHaul has been included as one of the proposed deployment options for 5G transport networks by the 5G-PPP Architecture Working Group, as reported in the 5G-PPP “5G Architecture” White Paper (figure attached). The 5G-XHaul physical architecture incorporates a wide variety of technologies including mmWave wireless technologies supporting dense deployments of small cells, a supported by a high capacity, flexible optical transport technologies comprising both, passive (WDM-PON) and active (TSON) solutions.

Based on the data plane architecture, the support of heterogeneous technologies has been addressed, in the design of the 5G-XHaul control plane reported in deliverables D2.2 and D3.1. To enable integration of heterogeneous technologies, 5G-XHaul has defined the concept of control plane areas, whereby each control plane area embodies a single type of technology. Forwarding elements of each control plane area are controlled by a separate controller, and a hierarchy of controllers is used to enable services provisioned across areas.

The project has recognized the benefits derived from a close integration between the mobile network (including the RAN) and the transport network. A set of use cases for the RAN and the transport network has been identified and described in D3.1. Based on the requirements derived from these use cases a preliminary set of interfaces between the mobile network and the transport network have been defined.

Final results

5G-XHaul has defined an overarching transport architecture to support both, operational and end-user services. In this context the 5G-XHaul data-plane considers an integrated optical and wireless network infrastructure for transport and access that can support a variety of functional split options. The wireless domain comprises a dense layer of small cells. Small cells are complemented by a macro cell layer to ensure ubiquitous coverage. Backhauling can be supported using a combination of mm-Wave and sub-6 wireless technologies. Alternatively, the 5G-XHaul architecture allows small cells to be directly connected to a CU using a hybrid optical network platform. This adopts a dynamic and flexible/elastic frame based optical network solution (TSON) combined with enhanced capacity WDM PONs. This platform can support demanding capacity and flexibility requirements for traffic aggregation and transport. Details of this work can be found on “D2.2 System Architecture Definition”. The 5G-XHaul data plane architecture has been included as part of the 5G physical architecture recommended in the 5G-PPP architecture white paper.

As part of the 5G-XHaul data plane architecture, specific extensions to current TSON and WDM–PON implementations have been identified and relevant designs are being studied. TSON extensions focus mainly on providing the ability to support CPRI type implementations as well as improved granularity and elasticity features that will allow flexible resource allocation not only in the time (current implementation) but also in the frequency domain to better match the 5G service requirements. 5G-XHaul will design a low-cost flexible WDM-PON for wireline X-haul based on autonomous wavelength tuning technique, with sufficient bit-rate and distance performance
5G-XHaul will design a programmable 60GHz transceiver, which will enable an SDN controller to control, among others, beamforming weights, hence the physical topology of the backhaul network, rate allocation policies, and energy management parameters.

The project has evaluated the compatibility of IEEE 802.11ad wireless technology against the requirements of different transport classes. Whilst CPRI support is not possible, new functional splits and traditional backhaul are. Future work will explore enhancements to the MAC and PHY, including MIMO, to support CPRI.

Regarding the SDN control plane for mobile networks, the work carried out in 5G-XHaul during the first year has contributed to theprogress beyond the state of the art in the following ways. First, a concrete definition for 5G Transport Slice has been proposed in D3.1, which is supported by a companion control plane architecture that maintains virtualization at the edge of the network. Second, the problem of scaling SDN control planes in transport networks, where control channels are often in-band has been identified as a key issue, and Dynamic Flow Rules have been proposed as a mechanism to reduce signaling load between the network elements and the controller. Finally, also reported in D3.1, a set of interfaces between the controller of the transport network, and certain elements of the mobile network, have been proposed to enable information exchange between the mobile and the transport network. A set of supporting use cases have been identified that benefit from having a tight coupling between the transport and mobile networks.

The approach to re-distribute processing between remote access points and central processing unit – so-called functional splits – has been the focus of several works in the context of 4G networks. However, the impact that the introducing of new air interface technologies in 5G will have on this concept has been largely unclear. With our work on 5G functional splits we were able to shed some light on this. Furthermore, these functional splits are expected to enable statistical multiplexing gains, yet these gains have scarcely been analyzed in realistic settings. In D2.1 a

Website & more info

More info: http://www.5g-xhaul-project.eu/.