End-to-end transport assurance is key to virtualisation success
Virtualisation is sweeping across the mobile world. It has already taken over the core networks of the operators, and now aims to take baby steps into the radio access network (RAN) network, starting with C-RAN in LTE networks, writes Sandeep Raina, a product marketing manager at InfoVista.
Virtualisation brings clear technical and commercial benefits of new technologies to the telecoms network. But have we figured out what will make this transformation a smooth success? Is it the new telco cloud infrastructure, the reduced infrastructure footprint, the reduced operations through a management and orchestration (MANO) system and/or an orchestrator or a combination of all these? The answer is yes, however we need to consider some critical factors of success that need attention now, as the network functions virtualisation/software defined network (NFV/SDN) roll-outs gain momentum.
Not everything has virtualised
Let us begin with underlining the fact that the network functions have not disappeared, they have only been virtualised– that is relocated, centralised, downsized, and run on non-proprietary hardware. The virtualised core network entities are still connected to the end-user mobile through the physical mast via a physical/virtualised RAN (C-RAN). Which means that it still requires thousands of miles of transport network to transfer voice, video and data to the virtualised core via the virtualised RAN. And the transport network – the backbone of the mobile network – continues to be physical: microwave or fibre. While its IP-isation (IP/MPLS or carrier ethernet) have made the physical transport infrastructure more efficient, it is this understated element of the network that requires the most attention if virtualisation is to deliver.
With C-RAN introduced in LTE networks, and introduction of a fronthaul, the result is the transformation of transport networks to the new mobile xhaul (fronthaul and backhaul), which will serve as the backbone for current and new access technologies for many years. It is essential to assure QoS over this highly IP-ized transport infrastructure in an integrated way to support current LTE-Advanced (LTE-A) traffic, small cell growth, and the upcoming 5G network slices.
Transport QoS is critical to virtualisation
A stable, error-free, highly available and highly reliable transport network will greatly contribute to – and guarantee – the success of the virtualised network and its promised services. To support this need, optical fibre is the transport network of choice. But with low availability of fibre globally, especially in the under-developed or emerging economies, this may not happen so easily, so fast.
Carrier ethernet offers a solution on microwave and optical fibre, making the transport network more reliable. However, there need to be better protocols to make the virtualised network deliver on stringent requirements of 5G. Introduction of new protocols on the fronthaul will solve the problem to a large extent. Such protocols will establish low bit error rates (BER), low jitter/wander and high availability. However, as all new technologies need to be vetted and policed, especially in their earlier days, QoS and service level agreements (SLAs) will come into play. Delivering such high-QoS services over NFV/SDN networks also means a high dependency on the transport network as virtual network entities are dynamically provisioned to keep up with the SLAs.
New transport protocols to achieve ultra-high speed and capacity
Transport networks have always been the cornerstone on which new technologies were rolled out. 3G would not have been as data-centric if carrier ethernet was not simultaneously invented. 4G would not have been as fast if optical fibre had not been introduced. Finally, the always-available 5G will not be possible if fronthaul protocols such as CPRI (Common Public Radio Interface), eCPRI (enhanced CPRI) and OBSAI (Open Base Station Architecture Initiative) are not implemented.
While CPRI is the interface of choice for carrying fronthaul traffic on fibre, eCPRI is being designed to support stringent 5G performance requirements – especially latency and jitter, and to offer ethernet as a transport channel with the promise of a significant reduction in the required bandwidth. Here is the promise of CPRI and eCPRI: Ultra-low BERs (10-12), low latency (return trip time (RTT) of 5 microseconds), low jitter and wander, and high data rates (up to 20 Gbps). But the promise is not enough; close monitoring is required to ensure that these new technologies and protocols are delivering.
In the virtualised world, the IP transport management is thus a different ballgame than traditional transport management. It cannot be relegated to the background any more. Only a robust transport network will ensure that the new telco cloud, and its web-scale traffic is efficiently handled, and that ultra-high quality is continuously guaranteed.
Assuring the transport as network entities virtualise
No transport system is immune to failures, degradations, faults and poor performance. Despite real-time orchestration introduced into fronthaul, IP/MPLS, NFV and SDN to counter this, prevention of such undesirable network situations is highly recommended. For this, extension of service assurance to the transport network has become as critical as service assurance for the RAN and core networks.
As fronthaul traffic throughput grows from current levels of between 1 Gbps to more than 20 Gbps, careful and continuous monitoring will be required. Management of backhaul and fronthaul in an integrated manner is key to an end-to-end view of how data quality is maintained or how it suffers, as it passes through the entire network.
Mobile operators should look for performance assurance solutions that offer real-time, multi-domain, automated root cause analysis (RCA) based troubleshooting to respond quickly to developing transport network problems. A system that not only provides performance data of the transport network, but also shows the dependencies of the RAN and C-RAN on it. A system that uses adaptive threshold capabilities to identify abnormal performance and provides usage forecasts to help network planning teams as well as orchestration systems to scale-up or scale-down in real-time. A service assurance system that offers transport SLA capabilities, together with real-time analytics.
Transport assurance, thus, will not just enable but accelerate the success of the transformed virtualised RAN and core networks, and optimise spend.