The global Coronavirus pandemic has seen more people working from home than ever before. Everything from national government to neighbourhood coffee mornings and international business sales meetings have been conducted online over Zoom, Google, Skype or Microsoft Teams, really putting home internet connections to the test. Frozen images, broken and corrupt audio may have been commonplace but – in the main – they have also been tolerated. It’s the new normal.
It’s also the consequence of the ‘best effort’ internet connectivity that makes up the vast majority of domestic connections. In these unprecedented circumstances people largely seem willing to forgive the lapses for now, but it is far from certain that they will continue to do so when the current crisis is over. On top of that, this reliance on home internet connectivity for business also means that consumers and business executives now fully realise the drawbacks of ‘best effort’ connectivity compared to guaranteed connections.
This new normal also highlights an upcoming 5G challenge that operators will face. While many businesses pay for guaranteed service level agreements, very few consumers operate with such specific contracts. But the 5G promise is all about guaranteed speeds and the low latency requirements associated with specific business and consumer services.
Exposure to first-hand experience of what ‘best-effort’ really means, at a time of heavy network load, is likely to result in an uptick in consumers wanting a more guaranteed level of service. Managing that demand for guaranteed performance alongside a higher standard of general performance will be a significant part of the 5G challenge facing the operator community.
Guaranteed service levels
In a 5G World, consumers will be asked to pay for, and come to rely on, guaranteed service levels for initiatives such as cloud gaming or even high-quality broadband connectivity. These agreements will likely cover speed, availability and latency. Once those metrics are being paid for in a service contract, best effort quality will no longer be tolerated.
Of course, at the same time, business expectations from mobile connectivity over 5G will also increase. Currently, operators that fail to deliver on enterprise service level agreements – maybe only hitting the agreed target 98 percent of the time – usually simply suffer from a financial penalty clause in the contract. However, 5G-based services such as remote medical diagnostics and field support, national infrastructure monitoring, or motorway traffic control systems will not tolerate 98 percent effectiveness. A guaranteed service will need to be precisely that – guaranteed – as no amount of financial penalty compensates for failure that could involve loss of life, not just loss of service.
5G networks will introduce a number of new transport techniques into the operator’s network toolbox. Of these, network slicing has captured the most attention because it is the feature that enables slices to be created across the network with each one potentially having different transport characteristics. And while the initial slicing focus has been on new 5G networks, it is a technique that could be equally applied and supported in fixed access connectivity.
Priority lanes via network slicing
To deliver ‘Best Effort’ internet connectivity, networks use a technique termed ‘weighted fair queueing’ (WFQ). It is relatively simple to implement, it is ‘fair’, and it simply prioritizes traffic queues based, for example, on their order in a round robin. It does not distinguish between individual frames of packets – it treats them all equally. This technique underpins virtually all internet browsing and video streaming traffic.
But network slicing effectively creates priority lanes. It says not all network traffic has the same needs; some traffic is paying for different priority treatment – especially in terms of latency. Think of it like airline passengers with silver and gold passes that can leap the queue.
You can therefore augment WFQ and best effort traffic management by adding different levels of priority forwarding. But it is not quite so simple as reserving lanes for ‘expedited traffic’. There would still be an element of ‘jockeying’ for position unless you have full isolation all the time between the different degrees of standard, silver and gold traffic flow. And that doesn’t make for maximum network efficiency.
Time Division Multiplexing
Adding another path into the mix, using Time Division Multiplexing (TDM) is now being actively explored as a route to give the best of all worlds – best effort services most of the time, supplemented by degrees of priority for some applications, and topped by dedicated paths when absolutely required. This would mean time slots on the network being reserved for the highest priority, 100 percent guaranteed traffic. But equally, like first class airline seats, these could be given up to other passengers if not required.
Those operators able to apply these techniques to the radio path, to the backhaul connection, and to the core transport network to deliver end-to-end guaranteed service levels, have the chance to get first mover advantage in the new normal world. And in a world where consumer and business expectations of service levels have been heightened, that could be a very significant advantage indeed.