We’re going to need a bigger slice
Around about the year 2020, 5G networks will start to supercharge the world of data services and the IoT. It sounds like IoT nirvana – data rates 100x compared to 4G, and latency reduced from 50 to just one millisecond. But it’s not just the turbo effect that makes 5G so different – it is also the ability of a 5G network to be many different networks at the same time.
The concept of network slicing is going to be implemented in 5G Networks. This enables networks to allocate different resources to different services all at the same time. One slice could be reserved for connected power meters or machinery in offices and factories, while another might be dedicated to healthcare services including remote monitoring and diagnostic equipment.
The slice looking after the connected car market obviously has different needs to the others – for a start, cars tend to move a lot faster than factories. Add in the concept of self-driving cars needing to talk to one another when sharing the same roadway, and the need for fast data speeds and low latency becomes even more important. For example, at 60mph a 4G connected car would cover four feet of road before receiving and reacting to a message while a 5G one would travel little more than one inch.
But there’s more to the 5G network slicing than the ability to offer differing service speeds to different slices. There’s also the opportunity for the operator marketing teams to create different service models and pricing configurations. And that’s what led to our involvement in a TMForum Catalyst group looking at service assurance across different network slices. After all, if an enterprise signs a contract for a network service at a premium pricing point, it will also certainly want an SLA on service quality.
In the TMF Catalyst group we’ve been working alongside partners such as AT&T, Vodafone and Orange to look at 5G Network slice assurance and the concept of 5G Networks As A Service (NaaS). We unveiled the first results of our work at the TMF Live event in Nice this month.
The Catalyst Group looked at what are termed ‘closed loop operations’. These enable elements of the network slice to optimize themselves, reacting to changing conditions in multiple industry verticals at the same time in a NaaS scenario.
We led the Catalyst looking at the requirements for the connected car market while another group looked in parallel at machine communications in an Industrial IoT environment.
Particularly challenging for the connected car environment are the related issues of road and network congestion. When a motorway is flowing, the transmission of data by connected cars will be sparse and across a broad geographic area. But in the event of an incident that causes a major tailback on the road – the connected cars are concentrated into a denser configuration in a restricted area.
The people in the cars then start using their mobile broadband apps, searching for information on the delay, looking to plot a different route, sending messages, and using email and other business applications. At the same time, the cars themselves start to transmit data more frequently to each other, to the network and to their own supporting infrastructure – especially as their control systems cope with the need to transmit and receive more data. The combination of these two factors puts enormous stress on both the mobile broadband and the connected car elements of the network slices.
For the services enabled by the connected car market, operators will need to guarantee network assurance and deliver the ultra-reliable, low latency communications required for the connected cars continuously regardless of network load. This means the network slice needs to be able to automatically grow with demand if the KPIs are showing that it is in danger of failing to meet its SLA. What’s more, as demands begins to tail off, the network slice needs to shrink back intelligently to ensure that service assurance isn’t put at risk by scaling back too quickly.
Enabled by the roll-out of SDN and NFV, the catalyst team recognised that each 5G network slice becomes its own mini-network with its own Quality of Service parameters represented by SLAs and OLAs. It is configured with its own architecture, engineering and provisioning rules and requires many of the same OSS/BSS support systems as the main network. The network orchestration system needs to be automated not only within the slice, but also co-ordinated across the entire network. This ensures that the network load is balanced and that service assurance is maintained for all verticals and their different requirements
The ability to slice the network into service segments opens-up the potential for operators to develop and market new value-added services. The technologists in the TMF Catalyst team are now delivering the next piece of the puzzle – the service assurance to back up the service potential.