By Helen Xenos, director portfolio marketing, Ciena
Consumer demand can put considerable strain on a network, particularly around major events or when popular trends go viral. Take for example the streaming of a World Cup final, the introduction of augmented reality games like Pokémon GO or the release of a new Adele album. Moments that matter to users frequently translate into moments of heavy network traffic for carriers. While carriers should be prepared for these spikes in demand by having extra capacity at the ready, they don’t want to be sitting on costly spare capacity that goes unused and unsold most of the time.
The same is true for operators of any large public or private network. There are times when they would like to sell extra capacity, but they don’t want to invest in additional infrastructure that isn’t needed 90 per cent of the time. Their customers might need extra capacity to deal with the occasional distribution of large files, or even to cope with demand for an online retailer during a product launch or Black Friday. In these cases, it would be nice if existing telco infrastructure could be stretched to allow customers to overcome a short-term bandwidth crunch.
There is indeed a way to push the limits of the existing network to offer bursts of additional capacity and meet short-term demand without having to physically build out the network. Telcos simply need to unlock the full potential of their existing fibre connections.
Untapped potential
Modern fibre-optic cables are typically lit with equipment that operates at pre-determined speeds, such as 100G or 200G, and the speed is defined based on full capacity, end-of-life calculations. However, in reality the network is not operating under worst-case conditions, and the cables could accommodate more traffic capacity at any point in time. The available capacity over and above the calculated speeds will vary from cable to cable, and will also depend on the quality of the fibre or how long the cable has already been in use. Most networks naturally have a certain amount of additional capacity available to them, but because this spare bandwidth is dependent on network conditions, it isn’t viable to sell as a 24/7 service. But, for short-term capacity needs, it is perfect and can be tapped into using the latest optical transmission technologies.
Take, for example, a post-production company in Soho. At certain points in the week or month, there may be a need to move a colossal amount of digital data in or out of the business (sending an edited movie back to a studio or downloading raw rushes to be worked on). To do this quickly, a high-bandwidth connection is required. But, for the post-production company, it doesn’t make economic sense to have a massive broadband connection or leased line in place all year round if it is only being used at full capacity a fraction of the time.
A situation like this offers telecom operators the opportunity to sell short-term bursts of extra capacity to their customers by temporarily increasing the bandwidth for an additional cost, and turning it back down afterward. For telcos, this offers a way to extract additional revenue from existing fibre connections, but the challenge lies in identifying how much additional capacity they have sitting dormant in their network and where specifically it is located.
It obviously doesn’t make sense for telcos to have spare network capacity sitting there, unused, on the off chance that a customer might need extra bandwidth one day. New technologies offer an alternative. Rather than allocating a certain portion of existing capacity for these ad-hoc demands, telcos can use advanced software to help determine where spare capacity is located, so it can be tapped into. These technologies allow telcos to push their existing fibre cables to the absolute physical limit of the infrastructure.
Network optimisation technologies found in the latest optical transport equipment often work on the basis that there is a certain fluidity to how a network’s optical spectrum can be divided up. This flexibility can be exploited with programmable hardware, using APIs and standard interfaces that support customers’ requirements for an accessible, open architecture. Combining advanced machine learning software with this highly instrumented, programmable hardware allows telcos to monitor and mine all available network assets, instantly respond to new bandwidth demands and allocate capacity across any path in real time.
When programmable networks are implemented, the network operator can define the range, latency, efficiency and cost guidelines, and optimise them independently through the network. The network can then work efficiently at high loads, as well as avoid bottlenecks and stranded capacities. By providing services more efficiently and maintaining better service-level agreements (SLAs), these software capabilities will allow telcos to extract more value from their network – they can massively scale up capacity at the lowest cost per bit and fully monetise their network resources. Ultimately, telcos want to wring every ounce of performance out of an optical link. Only by using advanced software to push their existing hardware to the limit can telcos ensure they attain this maximum network profitability.
