I feel the need... the need for speed

It’s around 30 years since Top Gun was showing in cinemas around the country. Fax machines had reached critical mass, though most offices still had a telex machine lurking in the corner. A V.32 modem was state of the art and 9600 bit/s was the most one could expect without deploying the heavyweights of Kilostream or Megastream and, despite Mercury’s appearance, BT really was the only game in town.

Thirty years on the industry is still obsessed with speed, and we are constantly being bombarded by carriers’ messages of high-speed, fibre, and super-fast broadband.

The latest OFCOM market report says that the number of UK superfast broadband connections (which they define as those providing 30Mbit/s or more) rose by two million in 2015. While there has clearly been huge growth in available bandwidth, the cynics can be heard saying that this represents an unnecessary increase, and that there is no underlying need for more capacity.

However further examination of the OFCOM report shows that these increases may not actually be enough on their own – for example there is a reference to a Nokia Bell Labs report predicting that mobile data will be 21-31 times the 2005 levels for Western Europe by 2020. This massive increase in a short time-frame reflects a seismic shift in how we interact and use devices and applications.

There are three obvious drivers for these predicted increases in data usage, particularly when it comes to mobile data. These are:

  • The influence of millennials joining the workforce with their penchant for social networking, collaboration and multi-tasking
  • The move to cloud services and the new compute paradigm
  • The Internet of Things (IoT), and the increase in machine-to-machine communications

So if demand for data continues to increase how will we satisfy the demand, given that networks cannot be redesigned in such a short timeframe?

In a business context, we should see three trends emerge to meet this growth in demand:

  • More mobile data – we will see a proliferation of dedicated SIMs for specific applications
  • Software Defined Networking (SDN) – no more active:passive connections where one sits idle in case there is a fault – all access and connections will be utilised, all the time.
  • More direct internet access – relieving pressure on corporate backbones, WANs and centralised internet breakout, providing more, local, high-speed access connections.

Clearly, it’s not just about feeds and speeds – we also need to look at latency, packet loss, jitter and all the other elements that contribute to a great end-user experience. It is here that deploying SDN can play its greatest part in ensuring traffic is routed according to its priority, application type and the relative available speeds, feeds, traffic and congestion.

By employing application acceleration services and local caching we can actually make a massive impact on the latency and usability of applications – whilst barely using any of the available bandwidth. The best news? This leaves the speed and the feed for those who really need it! 

Given the capabilities of SDNs and pronouncements from various carriers around their SDN capabilities, does this mean we can simply sit back and let the carriers provide the answers to all the above challenges?

To use the catchphrase of another film from thirty years ago “there can be only one!”

For those customers with one access provider and a common deployment of access connectivity and technology, it may indeed suit them to adopt the Highlander approach. However for those using a mix of Ethernet services, broadband and mobile data with without back-up circuits to optimise resilience and minimise cost, adopting a one-carrier approach is likely to involve a large shift in services and a steep hike in running costs. So adopting a carrier SDN solution is not going to be a panacea for higher speed, unlimited bandwidth for all.

For the real world, a blended approach will continue to offer the best results for cost management, resilience of routing and access. What’s more, deploying SDN as an overlay to a core network, gives you a highly cost-effective way to get all the benefits without seismic network shifts.