5G Transport requirements: SDN/NFV Platform, SR-MPLS,EVPN, Network Slicing & Edge Computing-A perfect match...

Introduction

5G is a game-changer, bringing in a host of services which were only in the confines of imagination; whether it is smart grid, remote surgeries, autonomous cars, or factory automation. 5G will represent a significant advance over previous mobile technology generations due to an explosion in the number of network-enabled IoT devices, greater fiberization and densification of cell sites and a “cloudified” RAN architecture. The magnitude of these changes is such that it is likely to have a transformational impact on the 5G optical network architecture extending right from the Access to the Metro and Core segments. The emerging cloud architecture with its software-centric network paradigm also presents opportunities for telecom vendors and service providers to evolve innovative products and services that can contribute to the revenue growth.?

What 5G will deliver in our day to day life?

Besides driving up the per-cell throughput by at least 10x when compared to 4G/LTE, 5G is also expected to lead to a massive 100x increase in the number of user devices through “Internet of Things (IoT)”, a significant reduction in network latencies by a factor of 10x to support real-time tactile Internet applications and an ultra-reliable network for a seamless service experience. 5G is introducing several new features that increases the complexity of radio access networks (RAN). Together these changes will put greater demands on the transport network and have a disruptive impact on the backhaul network architecture to better accommodate these requirements.

Radio is NOT the Only Key Ingredient in 5G —Transport backhaul is there

The radio tends to get all the attention in 5G discussions, but the underpinning transport networks that support next-gen mobile are a major focus for our customers and a major driver of our work with NEMs (Network Equipment Manufacturers). This is being driven by:

Densification of networks:?

The high volume of small cells that will be required in urban and city environments will be 200%-300% more than what is needed today, demanding substantial fiber deployments.

Need to feed the bandwidth beast:?

Once 5G networks are fulfilling heavy demand, data consumption will be off the charts and networks need to be prepped for this. In places like Korea, there is already double to triple data consumption versus 4G.

Networks need an edge:

As edge computing architectures are planned, the network is being disaggregated, with more components and functions hosted at varying edge locations. These edge sites must deliver reliability and high performance, and transport networks are a major part of assuring them.

Network Preparation

5G discussions typically focus on what’s happening in the air, but it’s the transport networks that will ultimately bear the heaviest burden of new 5G services. Network slicing, IoT, autonomous vehicles — take your pick. Each arrives with 5G performance requirements that typically need to be addressed by the same network architecture. All these ultra-fast speed, ultra-low latency, massive capacity, and high reliability, requirements demand a herculean effort to support.

Assuring the performance of 5G requires the complex transmission of video, data, and voice, from the core network to end devices. 5G services will not work as expected if the underlying transport network can’t provide connectivity that meets the requirements of each service.

In order to meet these challenges as well as widely varying requirements for performance, capacity and latency, the transport network must evolve and scale efficiently. This evolution needs to take place from the access network through to the core network.

Capacity –?Higher traffic demands in 5G will be presented through overall throughput increases and throughput per site increases. Peak user connection speeds in 5G will rise up to 20Gbps. In addition to this, a global average five-fold increase in traffic volume is predicted by 2025, 76% of which will be video traffic.

Connectivity –?5G will see the densification of cell sites which will drive the number of transport network access points that will be needed. It is forecast that this densification will mean a 3 to 4-fold increase in cell sites in new locations. In addition to this, new approaches to RAN deployments and virtualization of the 5G core will require more connection points, scalability and flexibility in the transport network.

Capability –?Transport networks will need new capabilities that can increase flexibility and agility, with an ability to provide the foundation for a distributed mobile network. In addition, techniques and technology used in 5G will facilitate increased coordination and?MIMO?(Multiple In Multiple Out) beamforming. This will see a requirement for low latency and improved synchronization, especially across the fronthaul network.

?Complexity –?The transport network needs to handle not only a huge increase in traffic but also a wide variety of network characteristics for each specific case. Some use cases will demand ultra-low latency connections, while others will have more relaxed requirements. Deployment of virtualized RAN, network slicing and distributed cloud networks will lead to more dynamic traffic patterns and more complex connectivity demands.

Cost –?The advances enabled by 5G cannot be managed cost effectively simply by adding more capacity to existing transport network infrastructure. The transport network needs to become more integrated end to end, from radio access, through the fronthaul and backhaul to the core network using solutions that will optimize TCO (Total Cost of Ownership).

5G Transport architecture

One main change that is noticeable when considering the evolution of technology from 4G to 5G is the split within the access network to incorporate centralization. This provides for greater flexibility and facilitates the virtualization of the RAN, but it does add to network complexity.

The C-RAN?(Centralized RAN)?approach introduces several advantages for the service provider:

Cost reduction –?centralizing processing capability means that the cost of a DU (Distributed Unit) reduces.

Energy efficiency and power cost reduction –?by reducing the hardware requirements of the cell site, general power consumption and also air conditioning costs can be reduced. This cost saving could be significant, especially with networks containing hundreds of thousands of cell sites.

Flexible hardware implementation –?this in turn will result in highly scalable and more cost-effective RAN solutions.

Improved coordination –?including performance optimization as a result of improved inter-cell interference coordination, as well as improved load management.

Improved offload and content delivery –?aggregation of processing capability at the CU (Centralized Unit) provides an optimal place in the network for data offload or hosting of content.

Deployment flexibility –?particularly with respect to where the functional split of the protocol stack lies, which in turn has an effect on the transport network.

Taking into account the implementation of C-RAN, the transport network can be broken down into three specific areas – fronthaul, midhaul and backhaul. Fronthaul transport is provided between the RRUs (Remote Radio Units)?and the DUs, midhaul transport is provided between the DUs and the CUs, and backhaul between the CUs and the?CN?(Core Network). The different transport network characteristics between these are summarized below.

When analyzing the capacity requirements in the 5G RAN fronthaul there are a number of considerations to take into account as they all contribute to the overall calculation. These include the number of antenna ports being used, the radio bandwidth and also the subcarrier spacing being used. The NG-RAN (Next Generation – Radio Access Network)?shall be able to support up to 1GHz system bandwidth, and up to 256 antennas. Calculations in relation to a possible transport deployment show that a theoretical maximum bitrate over the transport network of approximately 614Mbps per 10MHz per antenna port is needed.

Solutions with Fujitsu Packet Optical Networking Platform

Backhaul will provide the connectivity between small cells, macro cells, core networks and possibly numerous gateway nodes in between. The end points will be interconnected by a network of physical links, each with differing characteristics in terms of capacity, latency, availability, coverage, security, delay, synchronization, QoS, and physical design. Backhaul bandwidth capacity can be expected to be greater than 25Gbps and potentially up to 800Gbps. This will be dependent upon the amount of aggregation taking place and the deployment scenario (urban, rural, dense urban or indoor hotspot).

REDEFINING WIRELESS TRANSMISSION

Wireless backhaul will increase its position as the most flexible and cost-effective backhaul technology for mobile networks when 5G arrives. That status will not be achieved without serious technical advancement. The wireless transmission evolution will be driven by many new capabilities:

Ultra high-capacity wireless backhaul

High-capacity wireless backhaul will enable mobile operators to keep up with capacity demands and maintain excellent quality of experience for their customers. Meanwhile, operators will need to meet demanding operational efficiency targets by saving spectrum costs and avoiding costly and time-consuming fiber deployment.

Traditional microwave bands (4-42GHz) will leverage both wider channel spacing (such as 112MHz and 224MHz) and higher modulation schemes (4096 QAM and up), as well as ultra-high spectral efficiency techniques such as line-of-sight MIMO, to enable up to 10Gbps, and long and medium distance connectivity.

Short distance connectivity will heavily utilize higher frequency connectivity. E-Band and V-Band solutions will benefit from additional capacity-boosting techniques (currently more common in microwave solutions). Such techniques will include XPIC, LoS MIMO, and higher modulation schemes. This will enable rates of more than 20Gbps per link.

Combining multiple carriers in different frequency bands into a single link, sometimes referred to as Multiband or Carrier-Aggregation, will provide operators with the benefits of lower frequency ranges and availability alongside higher frequency band capacity. For example, a typical configuration might consist of a link created from an E-Band carrier planned for medium availability (99.9%, for instance, enabling longer range) and a MW carrier that increases the link’s capacity while providing a “safety-net” whenever the E-Band link is unavailable, thus creating an ultra-high capacity link with the range and availability of a standard MW link.

Because mmW spectrum will be heavily used for 5G RAN, additional, higher frequency ranges will also be allocated for wireless transmission. Above-100GHz bands such as W-Band and D-Band,

Though not yet regulated, are already the subject of R&D efforts to create power-efficient, small form-factor, ultra-high capacity wireless transmission solutions.

New frequency reuse scheme - deploying cellular base stations and small cells exactly where needed

Increasing the re-use of wireless backhaul spectrum will let operators meet operational efficiency targets by saving spectrum fees. This will also improve subscribers’ quality of experience by placing their cell sites at the optimal locations, with no real constraints posed by wireless backhaul frequency allocation and planning.

Such higher frequency re-use is available, enabling inter-link interference cancelation to the level of re-using wireless backhaul frequency with links close together to the level of 15 degrees of angular separation. This technology enables massive re-use of wireless backhaul spectrum, which will be critical in the ultra-dense 5G cell site grid. It also enables wide channel adoption (112MHz/224MHz) by freeing adjacent channel spectrum in existing networks. This enables the combination of several narrow channels (56MHz or below) into a single side channel.

High capacity Non Line-of-Sight (NLoS) solutions - enabling fast deployment

High capacity NLoS point-to-point wireless transmission solutions will enable true street-level mass deployment to accommodate capacity and coverage requirements in 5G dense urban environments.

While current sub-6GHz can overcome the limited spectrum channel widths to provide a fair backhaul solution for 4G/4.5G street-level deployments, 5G deployment demands capacities that are far beyond the scope of such solutions. Instead, 5G will call for high capacity, microwave and millimeterwave NLoS solutions.

Moreover, Vendors are pursuing a combination of NLoS adaptive channel estimation with MIMO implementation. This can increase link robustness, which is highly required in a NLOS environment.

In addition to NLoS operation mode, street-level backhaul will also feature low footprint, low power consumption, zero touch provisioning, and enhanced security.

Highly Accurate Synchronization?

The new 5G frame structure requires ±390 ns synchronization accuracy for the air interface. 5G inter-site CA and JT technologies requires ±130 ns synchronization accuracy (±5 ns for a single node of the transport network). Moreover, the high-precision positioning service with different positioning accuracy levels also has some special requirements for the synchronization.?

5G network clock synchronization by Anritsu External clock source

Wireless backhaul virtualization will serve two aspects of network virtualization:

1. SDN integration:?Wireless backhaul will integrate, via open interfaces, with the end-to-end SDN and NFV infrastructure and enable SDN applications to achieve network resource optimization (spectrum, power), higher service availability (with smart re-route mechanisms), and faster introduction of services and technologies. All of these are applicable in the wireless transmission domain, as well as in multi-domain, multi-vendor environments (assuming vendor alignment to standard-based interfaces and applications).

One application that will increase operational efficiency in the wireless transmission domain is the adaptive adjustment of power consumption at each site, according to the traffic running through the site in any given instance. Meanwhile, dynamic frequency allocation will be performed throughout the network based on required capacity weather conditions. This will offer considerable savings on spectrum and costs.

2. Cloud-RAN support:?Separating baseband units (into BBU hotels at data centers) and remote radio heads will create significant benefits to mobile operators. However, such a model heavily depends on what is today a highly inefficient I/Q interface between the two elements (CPRI, for instance). This interface should be transported via wireless transmission (and not only over fiber) in order to create a cost-effective transition to C-RAN. This will be enabled by higher capacity wireless fronthaul, as well as highly efficient compression mechanisms incorporated at the wireless fronthaul nodes.

?NOKIA’s Wireless and Optical Backhaul Solution

Mission critical backhaul

High-availability, low-latency, and highly secured wireless backhaul is common practice in today’s public safety and utilities networks. By using multiple layers of redundancy, low-latency transmission technology, and physical and virtual security and encryption mechanisms, wireless transmission has been proven, in many cases, to be a more secured and reliable transmission method than fiber (plus it is inaccessible to vandals and hackers).

These standards will be widely adopted in 5G networks as mission-critical applications are introduced, demanding a high level of QoS, very good quality of experience, and five-nines of reliability.

Edge computing for mission critical networks

5G Xhaul Transport Technologies:

At the crux of the 5G story is its ability to support time-sensitive applications that require processing of massive amounts of data with minimal delay.

TSN-Enabled Ethernet

Enhanced Ethernet for deterministic time-sensitive communications can guarantee end to end latency of 1ms or less required by uRLLC. IEEE 802.1CM developed to meet timing and synchronization requirements of 5G Xhaul Incorporates centralized configuration, path control and reservation capabilities. Ethernet is extensively used in access networks and enterprise connectivity.

Optical Transport Network (OTN)

OTN is technology based on SDH/SONET adapting to the great increase of data traffic in transport infrastructures. It is described in in ITU-T G.872 and the network interface is specified in ITU-T G.709. OTN is a highly scalable core network technology capable of time multiplexing the existing SDH stream with packet-switched data over the same frame. SDH/SONET, Ethernet, Synchronous Ethernet (SyncE), IP-MPLS, and MPLS-TP can all be accommodated by OTN. This technology provides a deterministic behavior similar to SDH/SONET. OTN is more important than ever before given their high transmission capacities forming the basis of many transport networks.

MPLS-TP

MPLS-TP is designed to meet transport network operational requirements. It borrows critical elements from IP/ MPLS such as its forwarding mechanisms, while including additional functionality such as performance monitoring, OAM, Tandem Connection Monitoring (TCM) and protection switching. MPLS-TP feature set is implemented as per RFC 56541. These are divided into general, layering, control plane, and protection and recovery.

How MPLS-TP works?

Three key characteristics of MPLS-TP that are relevant:

? Reduced dependency on “routable” IP protocols thus lowering vulnerability to network layer cyber-attacks

? Provides superior OAM capabilities with pro-active and reactive fault management and performance monitoring.

? Uses LSPs (Label Switched Paths) and PW (Pseudo-wires) to deliver connection-oriented services. \

Network provisioning via centralized Network Management System (NMS) is possible without using a control plane. MPLS-TP provides support for static traffic-engineered “pinned-down” service from a centralized network management system (NMS). This is perfectly compatible with traditional transport-style operations that assure enhanced reliability, predictability and determinism.?

IP/MPLS

To enable a 5G network, virtualization and control of IP routes is necessary. Control of MPLS using MPLS-TP was done in 4G. IP is needed for virtualization and PTN/CE for statistical multiplexing and for cost reduction. Carriers need to converge their networks to a single infrastructure to reduce Opex and support IP-based networking services as well as traditional layer 2 transport services. As an example as shown in the below illustration – in a shared nx1G/100G network, a combination of PTN and IP for access traffic aggregation and deploying the OTN deeper in the network is effective.

Network Slicing

Network slicing created using SDN/NFV and a network orchestrator helps to dynamically create new slices based on consumer needs with varying pricing while achieving the necessary isolation. A 5G end-to-end network with network slicing technology improves the time to market, accelerates the revenue growth and reduces total cost of ownership. 4G networks had only one network slice. In 5G networks, there are three slices, the IoT (mMTC) Slice, Broadband (eMBB) Slice and Low latency (uRLLC) Slice. Network slicing will maximize the flexibility of 5G networks, optimizing both the utilization of the infrastructure and the allocation of resources.

5G slicing with NOKIA E2E network slicing platform

Disaggregated routing

5G requires a major upgrade of transport infrastructure in order to provide up to 10 Gb/s to users and a 1,000-fold increase in capacity per unit area, as well as lower latency, higher availability, and sliceability.

Multi-access edge computing (MEC) is necessary to meet the low latency requirements of 5G’s ultra-reliable low-latency communications (urLLC), scale for massive machine-type communications (mMTC), and support emerging applications such as augmented reality.?While the scalability of IP routing has long made it the technology of choice for mobile backhaul, it will have an even stronger role to play in 5G as traffic patterns become more meshed and less predictable with the adoption of MEC, as shown in Figure below:

Therefore,?5G?represents a key use case for?disaggregated routing.

Segment Routing

As the RAN densifies with 5G deployments, increasing levels of network complexity become a concern for Service Providers. Segment Routing (SR) is one technology being discussed to help with simplification in the transport network. The concept of Segment Routing provides a very good way to “bridge” between older routing methods and 5G. It’s a much needed enabler for carriers who are migrating. However, Segment Routing is missing a crucial element for successful implementation – accurate and complete network inventory and connectivity data.

EVPN

Ethernet virtual private network (EVPN) is a next-generation full-service bearer VPN solution.

EVPN unifies the control plane for various VPN services and uses Multiprotocol Extensions for BGP (MP-BGP) to transmit Layer 2 or Layer 3 reachability information, separating the forwarding plane from the control plane. This makes EVPN suitable for use on SDN networks. In addition, EVPN can be used together with VXLAN, whereby the data plane uses VXLAN encapsulation to forward packets.

EVPN introduces the advantages of IP VPN traffic balancing and flexible deployment to Ethernet, solving traditional L2VPN's problems such as high network resource consumption and the inability to implement load balancing. In addition, EVPN can carry L3VPN services, reducing protocol complexity.

SR-MPLS or SRv6 is used for tunnels on a 5G transport network, and EVPN is used for VPNs, simplifying protocols in the overall transport solution.

Vendor Innovation

Blue Planet Unified Management Solution

Summary

Compared with conventional mobile communications systems, 5G puts more stringent requirements on the transport network. As the cornerstone of the 5G network, the transport network needs to introduce new transport interfaces, technologies and network control capabilities, adapt to diverse network architectures, meet demands for bandwidth, differentiated latency, highly accurate synchronization, network slicing and enhanced network openness and coordination, facilitating the continuous network evolution.

Thank you!!

Monowar Hossain

HOD, Microwave Unit (Planning+Operation)

VEON, Bangladesh

E-mail: [email protected]

+8801962424691