The potential impact of 5G will reach well beyond faster downloads and presents endless possibilities for the enterprise, carrier and consumer. The kind of services available in the 5G world offers us the opportunity to think about the transport network differently. The transport network can be so much more than a way to move bits from point A to point B. Instead, it can serve as a platform for service differentiation by creating and leveraging different tiers of bandwidth, latency, availability and duration. The main question to ask is, what is the best way to roll out the transport network to achieve this 5G future?
Today’s 5G network transport decisions will have a significant impact on tomorrow’s services as networks are upgraded from all-purpose IP connectivity to supporting the full potential of 5G’s future services. For service providers, the question is which strategy they should pursue to evolve the network infrastructure to support the unknown future potential of 5G services.
The answer primarily depends on two criteria: 1) How quickly carriers believe 5G will achieve its full multi-service potential. 2) Whether service providers understand that all traffic-engineering challenges are solvable at the IP layer. Based on this, we see three distinct service provider strategies emerging for 5G network transport.
Finding the right mix
There are three primary service classes 5G supports, and they all have a bearing on how we approach network transport. Enhanced Mobile Broadband (eMBB) is the faster connectivity we think about, delivering blazing-fast broadband to a home or business. The next is ultra-reliable and low latency communications (uRLCC) - examples include real-time video for gaming, remote surgery, etc. The final is massive machine-type communications, think of this as the Internet of Things (IoT) on steroids where you can support dramatically more IoT and sensor communications.
How exactly does network transport fit in with 5G and enabling these service classes? No matter where they are, all 5G devices connect to the mobile core via the transport network. While it is a given, there will be upgrades to the radio and mobile core, the future of the transport network is as yet undecided. Service providers must understand that the decisions they make today will impact the types of communications they can offer in the future, and this will affect tomorrow’s competitiveness.
Strategy 1: Business as usual
The first strategy we will explore is making a bet that the only 5G service deployed in the foreseeable future is enhanced bandwidth. In this case, service providers put off any other investment in the 5G transport network not needed for eMBB. The business as usual approach means operators will extend the network’s current IP-connectivity service to support higher bandwidths.
There are some benefits for this approach as it minimises the initial investment and puts operators on the fast-path to achieving ROI – if mobile broadband takes off. Besides investing in 5G radios, this limits additional capital investment for network infrastructure to expanding bandwidth capacity, which is a well-understood process. Operators then take a wait-and-see approach for other 5G services before committing to more significant transport network modernisation.
However, there are several downsides to this approach: 1) This often means expanding the IP and optical layers independently, 2) No optimisation of resources, and 3) Putting operators at a competitive disadvantage if other 5G services take off faster than predicted. These downsides will promptly erase any cost benefits gained by this approach.
Strategy 2: IP over dense wavelength division multiplexing (DWDM)
In this scenario, the operator believes the first 5G service will be enhanced mobile broadband (eMBB), but hedges its bets on URLLC and IoT in the future. In this case, offering 5G over the current 4G/LTE transport network infrastructure is a good start, but understanding the implications, operators start planning changes to their network infrastructure to prepare for future services.
In this case, the operator treats all traffic as IP-traffic, removing what appears to be unnecessary and simplifying network engineering and operations. Network slicing is used to handle the different 5G service categories, with all slicing mechanisms implemented exclusively at the IP layer.
Taking this proactive approach to network modernisation recognises that IP connectivity services predominate today. Overall this is a more cost-effective method, streamlining the network and bringing the IP service and DWDM transport layers closer to each other.
In theory, mechanisms at the IP layer will handle all 5G service variants, enabling fast deployment and added revenues as these emerge. But in practice, it is technically very challenging to engineer deterministic services like low latency plus high availability at the IP layer. This gap between theory and practice can lead to ineffective support for the more deterministic 5G services.
Strategy 3: Multilayer optimisation (MLO)
Taking a multilayer optimisation (MLO) approach is like the second scenario but concludes that deterministic 5G services cannot be supported adequately only at the IP layer. The goal of this approach is to maintain separate IP-connectivity and optical transport layers, but with one significant difference. They will optimise and manage these layers as an integrated whole, rather than independently as is done today.
An MLO approach accomplishes three things: 1) It dramatically reduces the capital and operating cost structure of the network by removing redundant resource utilisations, 2) It leaves in place TDM mechanisms such as OTN switching for creating network slices to handle deterministic 5G services, and 3) Once these multilayer mechanisms are in place, they provide a basis with which to offer the more sophisticated services leveraging bandwidth, latency, availability and duration.
MLO is not a new concept, but often discussed as the best approach to combine economic efficiency with multiservice transport effectiveness. But MLO can be technically challenging to achieve. The good news is that technology has evolved of late, and we now have many of the support systems of artificially intelligent, multi-gear and cognitive networks. MLO allows for much of the innovation needed for 5G, providing the opportunity to build vertical applications that will help realise the full potential of 5G and beyond.
What makes it promising is that it provides the most efficient use of CapEx by optimising overlapping capabilities among network layers, further streamlining network operations over an integrated network, hence reducing OpEx. MLO also maximises revenues via stratified services as it can provide both IP and TDM mechanisms to handle all types of 5G services effectively.
As one thinks about the kinds of services that will be available in a 5G world, it's worthwhile to think beyond business as usual. It is essential to consider how operators will get more capabilities from their network by better coordination between the IP and optical layers.
The bets service providers are making on the transport network today will have a significant impact on future 5G services they will be able to offer and whether they will be able to maximise 5G’s full potential. Of the three strategies described – business as usual, IP over DWDM and MLO – MLO has all the capabilities of realising the true potential of 5G and enabling service providers to tailor and profit from their investments, by offering different capabilities for different 5G service needs.
An intelligently controlled MLO network opens the network to an entirely new level of services – from all the excellent services hyped over the years as well as the unknown, innovative and disruptive services we haven’t even dreamed of yet. The bottom line comes down to how risk-averse the operator is. It will be interesting to see how today’s bets on network transport pay off 5 or 10 years down the road.
Jonathan Homa, Senior Director Portfolio Marketing, ECI Telecom